MadModder
The Shop => Tools => Topic started by: spuddevans on August 27, 2010, 03:32:39 PM
-
For the last year or so I have been gradually gathering up various bits and pieces to do a CNC conversion on my X2 mini-mill, just as and when I have had the spare cash and when I see things on offer.
Now I have got my Paddleducks engine running my attention has turned to working on my machines, and the first thing I wanted to so was to get started on my CNC conversion.
So for the last few days I started working on the electronics side of things.
A long time ago I played Bass in a little band, and I also did the PA duties. Anyway, the point of this seemingly pointless story is that I built myself a stereo 150Watt per channel PA amplifier for the band and mounted it all in a nice 19" rack case. Now this amplifier has been serving very usefully as a footstool under my desk for the last few years, so I decided to recycle it.
I stripped out the 2 150watt amplifier modules and all the wiring and pre-amp section, and this left me with a nice big 6U 19" rack case with a toridal transformer rated at 500VA, 35V. Once rectified and smoothed that will give me approx 40-45V at about 14amps, and that will provide that power for the stepper drivers.
Here is a pic of the transformer fixed in place, along with the stepper motor drivers placed roughly in the position that they will be fixed. Also pictured are a couple of small transformers that will be used for powering low voltage stuff.
(http://velvet-art.smugmug.com/photos/984449446_gT5BH-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#984449446_gT5BH-A-LB)
The reason I have 2 little transformers is to give me 4 power supplies that are completely isolated from each other. Why do I need this? Well, to protect the controlling PC from any voltage spikes and surges from the steppers, I have a isolated interface that goes between the PC and the stepper motor drivers, but this requires a 5Volt supply that is seperate ( ie, isolated ) from the PC's power supply. ( that's one isolated power supply )
I also have got one LCD voltmeter and one LCD Ammeter for monitoring the main power supply voltage and current demands ( not really needed, but I like dials and displays ), and each of these displays needs its own power supply that is exclusive to each LCD meter. ( so thats 2 more isolated supplies )
Then I have one more that I'm not totally sure if I need it at the moment, but it may well come in handy.
Anyway, I have also done some metalwork on the front and back panels of the case. It has been interesting trying to cover up all the holes originally made for the amplifier with the new holes for the new application.
Here you can see that I have installed a fan, this is positioned so that it will draw the heat away from the stepper motor drivers as seen in the 1st pic.
Also you can see the cutout for the 25way "D" connector for the connection to the PC, and the 4 4way XLR sockets that will connest to the stepper motors ( X,Y,Z and A (for the rotary table) ), these are of a locking type because if you unplug a stepper motor from its driver while it is powered then most likely you will have to replace the blue smoke in that motor driver, they really dont seem to like that. So hopefully having locking connectors will prevent disconnection thru vibration or pulling of cables.
(http://velvet-art.smugmug.com/photos/984450095_SPCbW-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#984450095_SPCbW-A-LB)
Moving on to the front panel, I have got the 2 lcd meters to mount, and also I have 4 moving coil ammeters and 5 switches to mount.
This is as far as I had got up to yesterday.
(http://velvet-art.smugmug.com/photos/984447840_CpHHn-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#984447840_CpHHn-A-LB)
Then today I milled out the holes for the LCD's and also another hole for a switch.
And this is what I ended up at the end of today.
(http://velvet-art.smugmug.com/photos/984448608_crSEq-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#984448608_crSEq-A-LB)
So on the front panel I will have a master Voltage and current readout on the 2 LCD's, and individual readouts for each of the stepper motor drivers. I will also have a switch for each stepper driver to easily isolate any motor ( especially helpful for swapping the Rotab for say another axis without having to power down the whole system.
There is also a master power switch ( the green switch )
Anyway, that's as far as I have got. That is pretty much all of the metalwork done on the electronics box, now I will be working on wiring it all up.
Tim
-
I stripped out the 2 150watt amplifier modules and all the wiring and pre-amp section, and this left me with a nice big 6U 19" rack case with a toridal transformer rated at 500VA, 35V. Once rectified and smoothed that will give me approx 40-45V at about 14amps, and that will provide that power for the stepper drivers.
Tim
I make that 49 volts output, too close to the max of those drivers, think you need a bit of buck boost winding to get the voltage closer to 30 v AC.
dare say someone else will chip in with more accurate figures.
John S.
-
I make that 49 volts output, too close to the max of those drivers, think you need a bit of buck boost winding to get the voltage closer to 30 v AC.
dare say someone else will chip in with more accurate figures.
John S.
That's a good point John, would you have a recomendation as to a safer voltage to run with?
I could just drop a couple volts pretty easily using a couple or 4-6 diodes in series.
I will have to fire up the transformer with caps and see what the unloaded voltage is.
Tim
-
I'm running these exact same drivers with an RS transformer rated at 30v AC and according to my reckoning that should put 42 volts out with a 10,000uf cap and under test it does exactly that.
I know you have all your bits so more of a heads up for the UK guys.
This guy on Ebay located at Swindon has rakes of these power supplies.
http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=380071301709
These power supplies are rated at 50.5 to 56.5 volts which is a bit too high but to get these voltages you need to feed a signal to a sensor pin on the PS.
If you don't connect the sensor up just feeds mains in , they put 42 volts out all day dead stable. If you can collect at the door he sells them for £10 each, which is an absolute snip, I bought 50 :thumbup:
John S.
-
Well that really is a great price for those psu's John, if I didnt already have this transformer ( and another 300VA one, same voltage, in case I want to expand to more axis's for a future lathe conversion ) I would def get a couple of those.
I just did the calculation for rms to peak voltage dc and it seems to work out at 35Vac gives 49.49Vdc, a bit too close for comfort. So if I can drop about 3-4 volts I will be ok ( I hope )
Anyway, I can drop the voltage by just using 4-5 diodes in series, a bit wasteful I know, but it is the easiest way that I can think of to reliably drop 3-4 volts without losing any amps. ( but I am open to suggestion if anyone else can give any other ideas )
Tim
EDIT:
I've just found some high power diodes that drop 1.5volts, so 2 maybe 3 of these will drop the voltage to a safer level for the drivers.
-
A while ago I bought what i thought was a 50v xformer from RS, wired it all in and just did a quick test, wanted 72 volts, and finished up with 78 volts which was a bit close to the 80 volt max on the large drives.
Turned out I had ordered a 55v xformer :hammer:
Not knowing about the diode trick I wound some heavy flex hookup wire round the torroid, 10 turns, switched on and that gave me 1.6 volts, so wound another 20 turns and got 5 volt.
Connected one the end of new coil to the existing one and took a reading over the two windings and got 60 volt. Reversed the termination and that got me back to 50 volts. Wrapped some of that white packing tape round it, the type with like a glass fibre weave and it looked just the part. It's still working like that today.
John S.
-
Not knowing about the diode trick I wound some heavy flex hookup wire round the torroid, 10 turns, switched on and that gave me 1.6 volts, so wound another 20 turns and got 5 volt.
Neat trick, I would never have thought of that :clap: :clap:
I dont have a whole lot to show for todays work, I spent most of this morning arranging and then re-arranging all the components to try and get them fitting with enough space around them, and with enough airflow thru to the exhaust fan to ( hopefully ) keep the drivers cool.
Then it was just a case of bolting it all down and then this afternoon I set about wiring up the drivers to the 4pin XLR sockets. And that's all I got done.
But I did finish off by probably doing the most important step, I took a photo of the progress.
So here it is, I've added a few labels to hopefully make clear what some of the bits are for.
(http://velvet-art.smugmug.com/photos/985368896_nnchE-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#985368896_nnchE-A-LB)
I still have to mount the smoothing capacitors, mount the bridge rectifier and wire them, wire the breakout board, make up a rectifier and regulator for the breakout board, and a whole bunch more too!!
Stay tuned for more updates,
Tim
-
.......with enough airflow thru to the exhaust fan to ( hopefully ) keep the drivers cool.....
Tim,
Just an observation - you may be better rigging it as a slightly pressurised case by having an inlet fan, with a filter, rather than an exhaust fan. Most cases "leak" (and I guess your metal pegboard base plate will let most of the air in) and you can't have those multifarious apertures filtered. Even if you don't have, say, any wood-dust, general environmental dust will gradually accumulate. Many consumer PCs rely on exhaust fans, and when you open up a PC that's been in an office environment for a couple of years you'll usually find lots of gunge & dust around all the gaps in the front of the case.
(good re-use of the 19" case though :dremel: )
Dave
-
Thanks Dave for your thoughts and suggestion :thumbup:
I have thought about reversing the fan, and I may well give it a try and your suggestion of the filter is a great one thanks, I hadnt thought about that.
One reason I have been leaning towards an exhaust fan is that with a pressurised system I would be trying to force the warm (hopefully not hot) downwards thru those little holes. That seems a little bit kinda the wrong way to me :scratch:
That is why I positioned the stepper motor drivers ( the main source of heat ) next to the fan so that the heat could be extracted from the case instead of being blown into the case. ( the stepper driver that is positioned further in the case is the "A" axis driver which will be driving at a lower current and thereby run a bit cooler )
The lid does have a series of vent slots in it, but I want to cover these up as with life in a metal-munching workshop there could well be tiny metal fragments that could settle on it, or even in it, and that could spell disaster in blue smoke.
I appreciate what you are saying though, I may well try running it in either direction and see how it works out.
Tim
-
Tim,
Who's breakout board is that ?
John S.
-
Tim,
Who's breakout board is that ?
John, it's justs a generic chinese made unit like this one linky (http://cgi.ebay.co.uk/CNC-DB25-Breakout-Board-MACH3-KCAM4-EMC2-6-axis-New-/250676385911?pt=LH_DefaultDomain_0) (I didnt buy this particular one, but it looks identical to mine )
I dont know if it is brillient or not, I guess I'll find out :zap:
As this is all a new field for me I am completely open to all suggestions and hints/tips and help offered, so if anyone spots something that could be done better, please speak up :thumbup:
Tim
-
Tim, One thing to note with this board is there is no charge pump for Mach fitted.
The Charge pump circuit prevents the machine from moving / glitching whilst the computer starting up. As a computer starts it polls the various ports etc to check if they are all active, this can cause an active machine to move off zero or even start the spindle, if so controlled.
No big deal if it's your own machine and not for resale, just make sure the computer is up and running with an open copy of Mach on the screen BEFORE switching the machine on.
John S.
-
Thanks for that John :thumbup:
I had read about using a charge pump, but like you say this particular breakout board doesnt have that feature. I was considering using one output pin to power a relay.
This is where it gets a little confusing, but stick with me ( otherwise it'll just be me who's confused here :lol: )
This relay ( controlled by Mach ) feeds power (the 45V supply from the transformer ) into a second relay, this 2nd relay is set up to be a latching relay (it will latch on until power is removed ) and it is controlled by the little green switch on the front panel. ( This is a momentary switch )
So the circuit works something like this, When 1st switched on, the power is not applied beyond the 1st relay ( the one controlled by the pc ), then when Mach is run and it sends out the signal to the 1st relay to turn on, then and only then can the 2nd relay be turned on by pressing the green button. Any time that the signal is inturrupted from mach then obviously the 1st relay will cut power to the 2nd relay which will also cut power to the drivers. The power will remain cut off until the signal from Mach is re-established AND then the green button is pressed again.
Hopefully this may make a heath-robinson charge pump replacement.
Tim
-
Well I left you all hanging on the edge of your seats, ( you were all on the edge of your seats, weren't you? if you weren't on the edge of your seats, please could you do so now or the whole effect will be ruined :lol: :lol: :lol: )
Anyway, I have not been completely idle since the last post, I have been working on the wiring and fitting a few more little bits into the box.
Here's a pic of the whole progress as of today.
(http://velvet-art.smugmug.com/photos/993085677_CA4DS-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#993085677_CA4DS-A-LB)
Here's a few closer shots.
This is a closer shot of the bridge rectifier. I mounted it on a piece of Ali heatsink that I had left over. I drilled and tapped it M8 so that it would be able to fit it onto the bolt that secures the transformer to the case. Then I drilled and tapped a M4 hole to mount the bridge rectifier to it. That should keep it cool even when under full load.
I have yet to hook up the positive to the smoothing caps, this is where I will be putting 3-4 diodes to drop the voltage to keep it within the limits of the stepper drivers.
(http://velvet-art.smugmug.com/photos/993086482_kkqZH-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#993086482_kkqZH-A-LB)
This is the next part of the power supply chain, the smoothing capacitors, 2x 10,000uF. Also beside it you can see the power relay that will switch the transformer on and off as described above in my post about my substitute for a charge pump.
(http://velvet-art.smugmug.com/photos/993087033_SECsQ-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#993087033_SECsQ-A-LB)
Here is the bank of 4 relays that are switched by the front panel switches, these feed the 45V dc from the smoothing caps to the actual stepper drivers. The reason for using relays for switching instead of just using the front panel switches are, firstly, the switches are illuminated, but they need to be switching 240Vac in order to light up (ok, not a really important reason, but I like lights and dials), and secondly when the main transformer is switched off the relays will immediately isolate the stepper drivers without waiting for the caps to discharge.
Also you can see a ferrite choke to hopefully suppress any interference.
(http://velvet-art.smugmug.com/photos/993087611_EoHaK-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#993087611_EoHaK-A-LB)
A closeup of the stepper driver wiring.
(http://velvet-art.smugmug.com/photos/993088159_Qwppe-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#993088159_Qwppe-A-LB)
Showing the wiring of the ampmeters and switches. Each meter is wired across a shunt ( a short, extremely low resistence link, in this case about 0.015 ohms ) You cant see the shunt directly, but in the picture above of the stepper driver wiring you may have noticed a single terminal block sticking up. The reason it is sticking up is that the shunt is inbetween the terminal block and the terminal of the driver. ( the actual shunt I had to make myself out of some resistence wire I had, it turned out to be about 14mm long was the needed length )
(http://velvet-art.smugmug.com/photos/993089103_DYRw7-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#993089103_DYRw7-A-LB)
A closeup of the back of the LCD displays.
(http://velvet-art.smugmug.com/photos/993084815_2TiVK-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#993084815_2TiVK-A-LB)
So that's as far as I have got, I have just been taking my time trying to keep the wiring as tidy as I can, it makes troubleshooting a bit easier ( I hope )
Stay tuned for more updates ( you dont have to stay on the edge of your seat this time :lol: )
Tim
-
So here's a little update on this.
I fitted a bigger transformer for powering the breakout board and for powering a speed controller board that I just got from the states, and possibly a spindle speed monitor circuit that I also just got. The original transformer only gave me 0.3Amp output, but this one gives me 1.3Amp, plenty to play with.
I also made up 2 sets of Bridge-rectifier and 5volt regulators along with electrolytic capacitors. I just hard-wired all this lot together, mainly as I had mislaid my vero-board ( which will no doubt appear as if by magic when I have finished doing all things electronic ).
You also might note a little yellow blob attached to the regulators in the pic below, these are Tantalam (woeful spelling) Bead capacitors, like electrolytic caps they are polarized and will go bang if connected the wrong way. A little cap mounted close to the regulators will help to supress interference.
(http://velvet-art.smugmug.com/photos/1002144305_EpGzZ-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1002144305_EpGzZ-A-LB)
I also modified the wiring on the back of the ammeters. Because I am using home-made current shunts which may differ slighty from one to another, I need some way of adjusting and setting each meter to read a true value. So I got some 100 Ohm 10-turn trim pots and fitted them in-line with one lead of each of the meters. (100 Ohm is a lot higher than I need, probably 30 ohm would do, but I only got 100 Ohm ones, but being 10-turn should give me the range of adjustment I need )
(http://velvet-art.smugmug.com/photos/1002143237_XoQJu-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1002143237_XoQJu-A-LB)
Then I had to create a little relay driver circuit to interface between the breakout board and the main power relay (my poor-mans charge pump system). The problem I faced was this, the breakout board's output ports are active low. What that means is that when the computer sends a signal to turn a particular pin on, the breakout board will actually connect that output to ground. This would not create a problem if the breakout board could handle the current to operate a relay, but alas it cant so I had to come up with a little circuit.
It is a very simple circuit, just a PNP transistor, a 1k resistor, a diode and a relay. The PNP transistor will turn on when the base pin is brought low, ie ground. Then the current can flow through it and into the relay, turning it on ( which will then allow the main psu relay to turn on ). The diode is there to suppress the back EMF when the relay is de-energised.
Again, not having any stripboard or veroboard to mount the circuit to, I mounted the diode to the relay, secured it with a dab of hot-melt-glue. then I hot-glued the transistor to the side of the relay, and then soldered the whole lot up. By mounting it all together it meant that I only had to bring 3 wires to the relay from the breakout board, then I could position this relay close to the main power relay, thus saving having to run 240Vac lines all over the place.
Here's a couple of views of the relay and associated circuit.
(http://velvet-art.smugmug.com/photos/1002142926_CPc3S-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1002142926_CPc3S-A-LB) (http://velvet-art.smugmug.com/photos/1002142677_pFGBP-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1002142677_pFGBP-A-LB) (http://velvet-art.smugmug.com/photos/1002142332_RzRq3-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1002142332_RzRq3-A-LB)
I then ( very tentitively ) plugged the power in, but before switching it on I unplugged the power supply wiring that goes into each stepper driver ( handily these stepper drivers have easily unpluggable connectors for all their wiring ) Then I switched on, and ....... well nothing went bang. Then I realised that my poor-mans charge pump circuit was doing its job, preventing the main transformer from switching on until Mach sends the signal to the right pin.
Not wanting to ( potentially ) fry my pc, I bravely decided to just over-ride the need for connection to a pc and just used a little piece of wire to link the output pin to ground, thereby energising the little relay. Then I was able to press the power switch on the front panel and ....... again, no bang or flash, just the Lcd's lit up showing 0.00v and 0.00A ( because I havent connected up the main Positive from the bridge-rectifier yet ).
So, the only thing left to do before testing with actual stepper motors and pc attached, is to work out how many of these power diodes are needed to bring the voltage down to a safe enough level for the stepper drivers. Then I will be able to do initial testing and setting up of the ammeters, and then the electronics side will be complete. ( barring a complete meltdown and explosion !!! )
Tim
-
Looks pretty damn good to me! Look forward to seeing it run some motors!!
-
Thanks Chris :thumbup:
I hope it'll work ok, if it does work then the real work starts. I have to get some ballscrews, Angular Contact bearings, oldham couplings, couple of timing belt pulleys and belt, a bunch of Ali plate to attach it all together, plus I am toying with the idea of a One-Shot lubrication system for the mill.
That's not to mention making a set of brass gibs for the mill too, and the possibility of making an enclosure for the mill to keep the chips and coolant managable ( did I forget to mention the thought of flood coolant, but that will def' be a later thing, I dont intend going into full-scale production with this, just for my own enjoyment )
I think I've caught it :ddb: :proj: :ddb:
Tim
-
Yep, definitely a nice control box. I often wished I had an ammeter on my spindle driver. It would have saved quite a few end mills over the last few years. :hammer:
What do you have in mind for spindle control? That's one thing my X2 has never had. I have a servo driven spindle on my gantry milling machine and I'll tell you, its nice to have the spindle fully controlled. This guy sells some speed controls that work with the X2. http://www.cnc4pc.com/Store/osc/index.php?cPath=25&osCsid=ce2a646b9ab3911fd1210250cff9b211 (http://www.cnc4pc.com/Store/osc/index.php?cPath=25&osCsid=ce2a646b9ab3911fd1210250cff9b211) I should get one, but I don't use the X2 enough now to justify it.
So what do you have in mind for the mechanics? Are you upgrading to ball screws all around? I can take some pics of my machine if you want some general ideas as far a motor mounting / coupling to the screws, ect. What size motors are you using?
You won't regret the conversion... :beer:
Paul
-
Thanks Paul :thumbup:
Yep, definitely a nice control box. I often wished I had an ammeter on my spindle driver. It would have saved quite a few end mills over the last few years. :hammer:
Actually, you have given me a great idea. All the ammeters on my control box are just for the steppers. But adding another ammeter to the mill itself to monitor the current going to the spindle motor. 1st class idea, thank you :clap: :clap:
What do you have in mind for spindle control?
I got the C6 from CNC4pc (http://www.cnc4pc.com/Store/osc/product_info.php?cPath=25&products_id=58&osCsid=ce2a646b9ab3911fd1210250cff9b211)
So what do you have in mind for the mechanics? Are you upgrading to ball screws all around?
Yea, I'm planning on converting all the axis's to ballscrews. I intend to use oldham couplings for the X and Y, and then use a timing pulley setup like Hoss on cnczone has for the Z axis.
I can take some pics of my machine if you want some general ideas as far a motor mounting / coupling to the screws, ect.
Yes please, the more info I can look at before hacking into my mill the better.
What size motors are you using?
They are the "350Ncm - 10mm Shaft - Hybrid Stepper Motor" that are part way down this page (http://www.arceurotrade.co.uk/Catalogue/Stepper-Motors/Stepper-Motors) ( hopefully these are strong enough, if not I guess I will use a 2:1 reduction via timing pulleys as I already have these motors )
On the Z axis I'm going to get a 2:1 reduction using the timing pulleys, and I will still have my air-spring to support the weight.
Tim
-
Just a little update, I dont have anything by way of pictures to show, but I have had it powered up tonight without letting any of the blue smoke out, so that is a bonus.
I ended up using 3 diodes to bring the voltage down to a safer level which turned out to be 47 volts with no load attatched. Now that the power supply was complete I set about calibrating the analogue ammeters, an easy job as I was able to apply a load to one channel at a time and then compare to the Lcd ammeter.
Then, just for laughs, I plugged the power leads back into the stepper drivers, hooked up all 4 steppers, donned an upturned saucepan on me head, stepped away and bravely switched it on with a wooden broom handle.
Well actually I didnt use a broom handle, but I did switch on and nothing went Booom.
Buoyed by this tremendous and overwhelming success I progressed on to switching on each individual channel, and with no smoke emerging I switched all the channel switches off.
It was at this point that I realised that the main Lcd ammeter was still giving a small reading :scratch: I pulled the cover off the box and saw that 2 of the stepper drivers Power LED's was still lit.
The problem is that a couple of the relays that are for isolating the power to each stepper driver are tending to stick in the on state. Not good!!
I should have known that they could be a problem area, the inrush of current at switch-on tends to weld the contacts shut. So I will just do a little re-wiring and run the power straight to the front panel switches.
Oh well, only a minor setback that will be easily fixed.
Tim
-
Great job so far Chris. I read the whole thread this morning and took the determination to finish my CNC controller board. Thanks for the inspiration!
Eric
-
Great job so far Chris.
Thanks Fred :lol:
I took the camera out to the workshop today, so here's a pic of where I put the 3 diodes to drop the voltage for the main Stepper drivers.
(http://velvet-art.smugmug.com/photos/1006265362_oPTMg-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1006265362_oPTMg-A-LB)
I got a bit more done today, I started on taking out the faulty relay and it's 3 companions and just wired the power feeds direct to the front panel switches. That done I turned to mounting the speed controller board into the case. I fabricated a few brackets out of Ali.
(http://velvet-art.smugmug.com/photos/1006266568_yLouo-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1006266568_yLouo-A-LB)
And I also made up a couple of ali brackets to mount the little spindle speed detector board in the case.
(http://velvet-art.smugmug.com/photos/1006268010_gF4Tb-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1006268010_gF4Tb-A-LB)
I then wired the 2 boards up as seen in the pics above.
I was thinking of mounting both the speed controller board, and the speed detector ( tach ) into the mill's own electronics enclosure, but then I realised that if I convert my lathe over to cnc too, then I'd need to buy or make a 2nd speed controller and Tach. But if I had the 2 boards mounted in the stepper driver case, then I would only need to have a switch to change over between lathe and mill.
Anyway, I think I am now at the point of bringing a pc out to the workshop to hook up to it, and then hopefully make the steppers turn according to my command. Then I will have to have a tidy up ( as the workshop is looking like a bomb has gone off in an electronics dump ) before starting on the modifying of the mill.
Tim
-
Just a little update on this. I brought out a PC and hooked it all up, installed Mach3 and then set about working out how to configure Mach3.
What is a little overwhelming is the shear amount of options and customisability ( I think I just made up a new word ). At 1st when I tried to move any of the axis's all I got was a quiet squealing noise from the steppers, but a quick trip into a connections menu to change the "active low" setting on each axis to being on solved that. And so I have motors spinning on my command. I have to say that I was really pleased that I had managed to wire all the stepper motor phases up correctly for each axis, I was half expecting to have to re-wire some of my leads.
Anyway, I spent the rest of that afternoon playing with Mach3 and it's settings.
That being done and found to be working ok, I have now turned my attention to the task of working on the mill itself. Over the last few days I have been gradually breaking it down into it's componant pieces with a view to cleaning them, setting them up better, and also seeing just how I am going to mount the Ballscrews and nuts when they arrive.
I have spent this afternoon working on aligning the head of the mill with its upright column so that the spindle axis is parallel with the dovetails. I ended up having to shim the bottom of the head casting by 0.16mm to get it close.( I say close because I have still got a runout of about 0.08mm over 12 inches of travel, but I got fed up with taking the head off the dovetail to adjust it. The error when I started was over 0.3mm over the 12 inches of travel)
I have a number of parts coming in the next week or so for this, some oldham couplings, some Angular contact bearings a few other bits that escape my memory at the moment. So until they come I am a little stalled. I have to re-assemble the mill though, so I'll be getting on with that, as well as working out what sequence I need to follow to make the various brackets and fittings for the conversion.
Tim
-
I havent done much today ( well nothing actually ), but Krymis' post asking about tramming a X2 and so I went out with my camera and took some pics of what I did yesteday.
As I "almost" described above, I have currently got my X2 mill in bits and spent the afternoon tramming the head to the column.
This procedure is much easier if you have the column laying flat on your bench, not that it would be impossible to do while the column is attached to the base, but it involves taking the head off and on again a number of times.
This is a photo showing the initial setup of the column lying on it's back with the head in place ( I have removed the pinion from the head so that it can just slide up and down unhindered ) at the top of the column. I have adjusted the jibs until I cannot feel any slop or side to side play, but not so tight that the head cannot be moved. Then I chucked a 12" 10mm piece of silver-steel into the chuck, mounted a Dti on the bottom end of the column and aligned the dti with the bar.
(http://velvet-art.smugmug.com/photos/1013579096_gpqLL-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1013579096_gpqLL-A-LB)
Now just before you think that you dont have a perfectly straight "test" bar to use, there is a work-around. Just rotate the bar thru 360 degrees, note the high and low readings, find the middle of the 2 readings and then make a note of that reading.
Then slide the head as far down the column as you can towards the Dti to take a 2nd reading, and repeat the above "averaging" reading.
(http://velvet-art.smugmug.com/photos/1013580000_uaexV-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1013580000_uaexV-A-LB)
If (without any adjustment) you have an identical reading as the top then double check it. If it is still identical then quickly think of 7 numbers for next weeks lotto!!!
But most likely it will not be the same, on mine I had an average difference of 0.30mm lower at the bottom than the top reading.
So now we have to shim the head to correct it. So loosen the jib screws a little to make it easier to slip off the column, and slide the head off the column and place it upside down and loosen the 4 screws that hold the 2 castings together that make up the head.
(http://velvet-art.smugmug.com/photos/1013582566_xM96d-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1013582566_xM96d-A-LB)
If, like mine, yours is painted over you might have to remove the 4 screws completely and then give the top casting a little tap with the handle of a screwdriver to break the paint seal. I then discovered that the paint was also covering some of the mating surfaces, so I very gently used a file to get the worst of the paint off, but I didnt file down to expose bare metal, then I put a sheet of wet-and-dry sandpaper on my surface plate and with some light oil for lubrication I rubbed the mating surfaces of both halves of the head until I got rid of all the offending paint.
Having done that I wondered if I had made it any better, so I put it all back together and mounted it back on the column, adjusted the jibs again, and went through the same checking procedure as I described above. Amazingly there wasnt much difference, just a bit worse than before :doh:
Anyway, I took the head off again, loosened the 4 bolts shown above, and got my very expensive shim material (tinfoil), and folded up 4 thicknesses of it and inserted the shim at the bottom edge of the casting, tightened the bolts, remounted the head onto the column and adjusted the jibs again, and re-checked it.
On mine it reduced the difference to 0.20ish mm, so I removed the head and folded up 12thicknesses of foil and re-mounted and re-checked... difference was down to 0.08. So it was "Off with it's Head!!" again and this time tried 14 thicknesses of foil.
This time it was close enough for me, actually a little over, but at a later point I plan some other mods that will give me an easier way of adjustment, so I am happy to be within 0.04mm over 12 inches.
Here's a pic of the shim in place, sorry the pic isnt great quality, this is a really well cropped fragment of a closeup, you can just see the shim at the bottom.
(http://velvet-art.smugmug.com/photos/1013582698_KMcNs-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1013582698_KMcNs-A-LB)
Now, once you have got the shim to the right thickness, this time only tighten the 4 bolts slightly ( loose enough to make rotational adjustment ) Then remount the head, adjust jibs, and then move the Dti round to measure the side-ways movement of the bar.
(http://velvet-art.smugmug.com/photos/1013581002_hR7iA-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1013581002_hR7iA-A-LB)
Make a note of the reading as before, and then
(http://velvet-art.smugmug.com/photos/1013582026_FFhG7-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1013582026_FFhG7-A-LB)
Note reading as before.
This adjustment is much easier than the 1st one. Very importantly, retract the Dti before making any adjustments. Then just use a deadblow hammer to GENTLY tap the relevent top head casting corner in the relevent direction, re-introduce the Dti and then check again.
Once you have it aligned to your satisfaction then loosen the jibs and very gently remove the head, and really really gently put the head upside down again and then tighten the 4 mounting bolts up fully. You have to be really gentle with it so as not to disturb the settings before you tighten the bolts.
Then it is just a case of re-mounting the head onto the column and checking it one (hopefully) last time to make sure it is still ok in both ways.
And there you go, the head should now be aligned with the column.
Some may wonder just why go to all this bother, what difference does it make? Well imagine that you have to drill a hole, maybe 1st you chuck up a spotting drill or centre drill to start the hole. Then you raise the head up and chuck a full length drillbit and then you notice that the longer drillbit is not perfectly lined up with the spotting drill's hole. This is because the head is further up the column with the drillbit than the short centre drill or spotting drill, and so if the head is not totally in line with the dovetail ways of the column this will introduce this error.
There may well be other implications of not having it fully aligned, but I'll leave it for other to add them. Suffice to say that a properly alligned and trammed mill should work a lot better than one that is not.
This is just the 1st part of tramming the X2, but there is plenty of info on the 2nd step, especially Bog's tramming tool build thread, well worth checking out. That's on my To-Make list :thumbup:
Tim
-
Good info on tramming the head to the column. Many thanks Tim! I am going to be doing this within the next 4 weeks or so. Because of this thread, I have been encouraged to start my CNC conversion again. I just finished soldering up my controller and order a power supply (don't feel like making one).
Eric
-
That's great Eric, I'm glad you're carrying on with your conversion, I look forward to reading all about it.
I'm at a bit of a standstill at the moment, just waiting for some parts to arrive, I have to re-assemble the mill though so hopefully that will fill in the time til the parts show up.
Tim
-
Great job so far Chris.
Thanks Fred :lol:
Doh!! :doh: :hammer: :hammer: TIM I meant Tim!! :lol: :lol: Teach me to read Chris's post on your thread before I reply!!
Eric
-
Well, after ordering them on the 13th of this month, the Ballscrews arrived today from China. Here's a shot of the longest one sitting on my mill table. They may not be the best in the world, but for me they will do just nicely. If needs be they can be upgraded at a later point if the need arises, but I think they will do just fine for my home use.
(http://www.velvet-art.co.uk/photos/1026176813_WM6RJ-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1026176813_WM6RJ-A-LB)
I also have got most of the ali' that I need for the stepper mounts, and I also have the bearings, pulleys and couplings that I need ( I think :scratch:)
Here's a shot of the ballscrew with the Ballnut removed and all the (48) ballbearings and the white plastic "wiper" for keeping debris out of the ballnut. The little orange recessed bits on the ballnut are re-circulation passages for the ballbearings to run through.
(http://www.velvet-art.co.uk/photos/1026177898_LGddX-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1026177898_LGddX-A-LB)
I have seen that other folks have used a homemade split collet to protect the ballscrew while machining the ends, but as I have an ER32 collet setup on my lathe that I have used to clamp (relatively) soft brass without causing damage any to it, I decided to use it to hold the ballscrew in the lathe, and it worked ok.
(http://www.velvet-art.co.uk/photos/1026179599_2V4cs-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1026179599_2V4cs-A-LB)
These ballscrews are hardened, but only on the outside, so I used carbide tooling. It worked really well, much easier than I was expecting, I was able to take 0.2mm cuts even on the hardened sections. So I turned down the end to 10mm diameter for a length of about 10mm ish (thats a metric "ish" by the way).
Then I extended the ballscre out a bit more until I had about 40mm sticking out of the chuck in total. I then turned down the rest to 11.95mm-11.98mm. This is to fit a 12mm id bearing. Once I got a nice close sliding fit I marked off the 10mm or so that was left between the bearing and the 10mm section and set about threading it M12.
(http://www.velvet-art.co.uk/photos/1026178307_Q87f4-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1026178307_Q87f4-A-LB)
The end result is here:-
(http://www.velvet-art.co.uk/photos/1026179197_9p9P8-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1026179197_9p9P8-A-LB)
Now the less-than-completely-blind of you may well have noticed that the thread does not appear to be that clean. I would like to say that it is completely an anomaly caused by the phases of the 3rd moon of jupiter, but I cant. I made the mistake of using a threading tool that had 0 degrees top relief. This kinda shredded its way through the thread. I only realised this about half-way to depth, and so after I realised it I carefully removed the toolbit from the QC holder and ground some top relief on it and replaced it. Threading went much better after this. Anyway, even though it looks kinda rough, it will be ok for this.
But before I cut the thread to full depth I hunted high and low through the workshop to try and find even one M12 nut, but I didnt have anything M12. So I dug a little 20mm round steel bar out of the "scrap" pile and set about drilling and tapping it M12. Handily it had a central hole already there.
I then clamped my 3jaw onto the mill table to hold and drill the steel "nut".
(http://www.velvet-art.co.uk/photos/1026178743_vttkm-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1026178743_vttkm-A-LB)
Once this was done I was able to cut the thread to full depth, testing with my "nut" to make sure it fit ok. (having a M12 die would have been really helpful)
Here's a shot of the completed machined-end with double row angular contact bearing, small space for locknuts, and then the oldham coupling in place.
(http://www.velvet-art.co.uk/photos/1026180167_4eoZB-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1026180167_4eoZB-A-LB)
The one thing I was kinda dreading was haveing to repack the ballnut with the 48 ballbearings, so much so I had been contemplating turning the ballscrews without removing the ballnuts, but that would not have been a safe thing for me to do. However, it turned out to be a fairly simple process with these ballnuts. In fact the easiest way seemed to be to just start the ballscrew in the empty ballnut ( they have a plastic wiper on each end that are threaded to match the ballscrew) and just pop in about a third of the balls and "jiggle" the screw and nut while gently turning the screw into the nut a couple of rotations, then add another third of the ballbearings, jiggle and then repeat for the last third. They all went in and the screw turns without any binding or feelable backlash.
So, that is the Y-axis ballscrew machined on one end ( I dont think I need to machine the other end ), next up will be the long X-axis ballscrew which will get an identical end machined on it, and then the Z ballscrew which will be a little simpler.
Tim
-
Tim,
Don't know if it's too late or you have any more to do but if you get a bit of 14mm bar and turn about 0.5mm off the diameter and push this up to the end of the screw you can unwind the nut onto it without loosing all the balls.
A bit of tape at both ends after stops it sliding off.
To refit you just push the bar up to the screw again and wind on.
John S.
-
Tim,
Don't know if it's too late or you have any more to do but if you get a bit of 14mm bar and turn about 0.5mm off the diameter and push this up to the end of the screw you can unwind the nut onto it without loosing all the balls.
A bit of tape at both ends after stops it sliding off.
To refit you just push the bar up to the screw again and wind on.
John S.
Thanks John :thumbup: :thumbup: That's a really useful tip. I have already removed the X-axis nut, but I will try your tip with the Z-axis, and with the others if I need to remove the nuts for any reason.
Tim
-
Help required, please . . .
I have acquired the same boards as you and have no documentation to suit.
I've been searching the Web for days, found this Page and have joined this Forum expressly to ask these questions.
I have figured out from your excellent Photos, that a Red Wire goes from Breakout Board P2, and a White Wire from P3.
Looking at the Photo of the Stepper Controller Board, I see the Red Wire going to the Pulse Pin, and the White to the Dir Pin.
Query 1;
Can you confirm that each of the even Numbered Output Pins, ie P2, P4 etc are Pulse Pins, and that the odd Numbers are Dir Pins??
Query 2;
Can you tell me which DB25 Pins control which Breakout Board Output Pins?
I will be running a 4-Axis Mill, (Sieg SX3), with these Boards under Mach3. (X. Y, Z & A)
Query 3;
I have a 24Volt 100Amp PSU which I intend using for this Project. Is it feasable/advisable to take the 24V and drop it to 5V for the Breakout & Stepper Boards as well as feeding the Steppers, or is it better to have a seperate 5V supply?
Looking forward to a reply
Murray McKenzie
New Zealand.
.
-
Help required, please . . .
I have acquired the same boards as you and have no documentation to suit.
I've been searching the Web for days, found this Page and have joined this Forum expressly to ask these questions.
Let me welcome you to the forum, its a great place to hang out :mmr:
I have figured out from your excellent Photos, that a Red Wire goes from Breakout Board P2, and a White Wire from P3.
Looking at the Photo of the Stepper Controller Board, I see the Red Wire going to the Pulse Pin, and the White to the Dir Pin.
Query 1;
Can you confirm that each of the even Numbered Output Pins, ie P2, P4 etc are Pulse Pins, and that the odd Numbers are Dir Pins??
Yes that's how I have wired it, P2,4,6,8 are pulse to the steppers and P3,5,7,9 are the direction controls. P14,1,16,17 are used for my charge-pump circuit, spindle speed control and 2 relays for coolant and possibly spindle direction.
Query 2;
Can you tell me which DB25 Pins control which Breakout Board Output Pins?
I will be running a 4-Axis Mill, (Sieg SX3), with these Boards under Mach3. (X. Y, Z & A)
As far as I know they are the same as they are numbered on the outputs on the breakout board, ie P2 (which goes to the stepper driver pulse input) is actually pin 2 on the DB25 connector (although it is opto-isolated on my board)
Query 3;
I have a 24Volt 100Amp PSU which I intend using for this Project. Is it feasable/advisable to take the 24V and drop it to 5V for the Breakout & Stepper Boards as well as feeding the Steppers, or is it better to have a seperate 5V supply?
It would be a better idea to have seperate power supplies for the breakout board, and if you have the same C6 spindle speed control from CNC4PC that I have, it is very important that it is powered from its own dedicated power supply, otherwise it can blow your breakout board and stepper drivers.
That's why I have put a couple of extra transformers in my box to give me plenty of totally isolated supplies.
You might find that your steppers will not have their full strength/speed on just 24V. A really good place for info is the cnczone forum, I've learnt a huge amount from reading on there.
Tim
-
Thank you Tim, this is the information I've been seeking for days.
The Boards that I'm using are totally identical to yours, so I'll pick your Brains again if I have problems, (if you don't mind).
I wont be using a spindle Controller, as the SX3 has Electronic Spindle Control on the Head Unit.
As I usually switch on the CPU, then the Control Electronics, is a Charge Pump needed?
Murray
-
The Boards that I'm using are totally identical to yours, so I'll pick your Brains again if I have problems, (if you don't mind).
You are more than welcome, pick away, but I warn you that there is "slim pickings" in my brains :lol: :lol:
I wont be using a spindle Controller, as the SX3 has Electronic Spindle Control on the Head Unit.
As I usually switch on the CPU, then the Control Electronics, is a Charge Pump needed?
In a perfect world you wouldnt need one, and if you are careful to always do just as plan to then it should be ok. My own charge-pump type circuit is nowhere as good as a proper charge-pump, something I may upgrade to in the future. The reason is, what if the unthinkable happens and windows crashes, even my system would most likely keep on working until I hit the emergancy stop button myself. But with the proper charge pump circuit then if windows crashes then the charge pump signal would also cease and thereby cut the steppers and spindle dead.
Feel free to ask anything else, I may not know the answer, but there are plenty of others who are more knowledgable on here.
Tim
-
Which Forum on CNCZone did you go to for Charge Pump info? (preferably "Charge Pump 101 for Newbies")
You didn't happen to document all your Circuit Wiring did you??:D
.
-
Which Forum on CNCZone did you go to for Charge Pump info? (preferably "Charge Pump 101 for Newbies")
You didn't happen to document all your Circuit Wiring did you??:D
.
There wasnt a specific forum about Charge pump info, I was meaning that I have learnt a lot in general about CNC by reading a lot in the Benchtop Machine forum, but I dont think there is a "Charge Pump 101 for Newbies" thread, I just have picked up a lot from reading the various build threads.
I havent documented the wiring any more than in described here, but it is fairly straightfoward stuff. If you have any specific questions just ask.
Tim
-
I've no idea when I found the charge pump circuits , but this may give you an idea
-
Hmmm, thanks Tim & John.
What I really need is to understand what a Charge Pump is, what it does, and how it relates to the CNC Conversion.
Murray.
-
Hi Murray ,
the charge pump circuit is a basic safety interlock so that the motors can only run if the control pc is
generating the 12.5 khz signal
this stops random outputs from the pc while booting up or when it crashes
being a danger to you or your cnc machine
if you look at figure 1 in the PDF
the 12.5 khz signal is ac coupled to the two diodes and the fets gate ( the charge pump bit, note in this circuit the charge is stored on the gate its self )
its a simple circuit that provides the bias to switch on the fet and the relay
the relay contacts are then connected to your control circuit to enable the motors
when the pc stops working ,a logic 1 (+5v)or logic 0 ( 0v) from the printer port can not enable the motors
the second diagram is a watch dog timer
the 12.5 khz pulses reset the timer just before it times out ,to keep the motor control enabled
(think of the train driver pushing the dead man switch when the buzzer sounds so the train
is'nt stopped automatically)
not all charge pump circuits are actually charge pump circuits !!!
the first circuit used was and the name stuck
like "hoover" is used for all vacuum cleaners
John
-
This is what I've achieved so far, thanks to you guys.
Murray
-
I've no idea when I found the charge pump circuits , but this may give you an idea
This is great, thanks very much John :thumbup: :thumbup: I will most likely build one of these circuits to upgrade my very simplified system.
This is what I've achieved so far, thanks to you guys.
Murray
That is a very neat looking box of wires, well done :clap: :clap:
Although I dont have anything to show yet, I have been able to get a little more work done. I have machined the X-axis ballscrew on both ends, and have just finished grinding off some the X-axis ballnut to make it fit into the mill. I will take some pics when I next go out to show what I've done.
Tim
-
Hi Tim and Murray
when I decide to convert a machine to CNC I'l have to build a control box like yours
keep up the good work
its taken some time to find them but I have 2 more circuits for you
I'l have to collect all my cnc info into one folder instead of several across three drives
at the moment my desktop that fitted my 22" wide screen monitor is hard to read
on the old 14" ,4 x 3 monitor i am using for now -
new backlight driver in the post i am told ,so back to normal in a week or so
the second diagram looks like the power is disconnected from the stepper drivers when a limit switch or stop button is pushed
your not supposed to disconnect the motor or power supply , when the powers on , are you ???
if thats the case I would connect a large diode across the contacts so the back emf can flow back into the
power supply capacitor(s)
also if you put a fuse in the supply to the drivers a normally reverse biased diode across the fuse as well
what do you think ?
John
-
your not supposed to disconnect the motor or power supply , when the powers on , are you ???
I cant speak for all stepper drivers, but on mine it is ok to disconnect the PSU at any time (otherwise you could never switch the driver off)
However, I'm pretty sure that most stepper drivers will let the magic blue smoke out if you disconnect the motor from the driver while the power is switched on, so make sure that if you use connectors in your motor wiring try and use locking style connectors in order to cut down on the chance of an accidental disconnection.
Thanks for the circuit diagrams :thumbup: :thumbup:
Tim
-
Hi Tim
I'll have to think about that
the diagram shows RLB switching the DC supply to the gecko drivers and not the mains to the power supply !!
I will have to look at the gecko website
found some info on gecko's site
-
the diagram shows RLB switching the DC supply to the gecko drivers and not the mains to the power supply !!
I had mine wired like that, relays switching the 45V DC to the stepper drivers, but I had problems with relays sticking from the inrush of current at switchon. So I moved the relay to switch the AC on the input to the power transformer where it had a smaller current to deal with (but with a higher voltage) and that seems to be working better for me.
Of course, you have to make sure that the relay will only switch off the higher voltage just for the stepper drivers, and not the low voltage 5V supplies for the breakout board, otherwise you will not be able to switch the steppers back on.
Tim
-
Of course, you have to make sure that the relay will only switch off the higher voltage just for the stepper drivers, and not the low voltage 5V supplies for the breakout board, otherwise you will not be able to switch the steppers back on.
Tim
Hmm . . . Do you Guys always answer my Questions before I've asked them ::)
Murray
-
Hi Murray
I'll take it as a good sign we are all asking the same questions
-
What Switch settings would you recommend to start with?
I'll be running 24VDC @ 100Amp PSU.
Stepper Motors are NEMA 24 rated at 4.2Amps 4.5Nm, (650 OzIn).
Running a Sieg SX3 under Mach3.
Steppers are driving the LeadScrews at 2:1 ratio. LeadScrews are the original Acme at present. BallScrews will happen after Financial Recovery :ddb:
.
-
Earlier I promised some pics on what progress I have made, well hear they are.
A couple of days ago I stripped down the mill table to do some preparation work on the table. I had to dig out some of the casting in order to make room for the BallNut for the X-axis to fit in. Not having access to an angle grinder I turned to less gainly methods. I started out by drilling as much away as I could, and then it hit me that I could fix it to my lathe and mill it that way, so I finished the job off in that manner.
I dont have any in-progress pics as it was a fairly dirty job and I value the working status of my camera :D But I do have an "After" pic.
(http://www.velvet-art.co.uk/photos/1055130864_p5aKp-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1055130864_p5aKp-A-LB)
(http://www.velvet-art.co.uk/photos/1055131642_Rtqw6-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1055131642_Rtqw6-A-LB)
The slot is to allow the body of the Ballnut to sit low enough in the casting. It was after this that another problem raised it's head. After milling out the casting for the ballnut I tried the ballscrew complete with ballnut in-situ to see how it fitted. This photo shows what the problem was.
(http://www.velvet-art.co.uk/photos/1055132387_pwe2K-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1055132387_pwe2K-A-LB)
(http://www.velvet-art.co.uk/photos/1055133126_uMLZD-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1055133126_uMLZD-A-LB)
Ack!!! the top half of the ballnut's flange would not fit under the mill table. Then double Ack!!! the whole of the ballnut was hardened, my hacksaw would just skate over it without hardly marking it. Still not having an angle grinder I turned to my trusty bench grinder. A very dusty hour ( or so ) later I had this;
(http://www.velvet-art.co.uk/photos/1055133657_yk7a7-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1055133657_yk7a7-A-LB)
And this is how it looks in place ( minus the ballscrew as I havent had chance to re-pack the ballnut )
(http://www.velvet-art.co.uk/photos/1055134732_3E29W-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1055134732_3E29W-A-LB)
After that I started cutting some Ali for making the mountings for the steppers, but then I was rudely interrupted by the call of the wild (my dinner), no pics as I'm pretty sure that most of you have seen 2 pieces of rough cut ali before :scratch:
Tim
-
What Switch settings would you recommend to start with?
I'll be running 24VDC @ 100Amp PSU.
Stepper Motors are NEMA 24 rated at 4.2Amps 4.5Nm, (650 OzIn).
Running a Sieg SX3 under Mach3.
Steppers are driving the LeadScrews at 2:1 ratio. LeadScrews are the original Acme at present. BallScrews will happen after Financial Recovery :ddb:
.
I'm not too sure what you mean by switch settings, the dip switches on the Stepper drivers? or switch relay ratings? or switches like I have for individually switching the power to each axis?
I forgot to say earlier, but if you want even bigger versions of any of my pics you can just click on them and it will take you to my photo site where larger versions are available. ( just so you can see my mistakes and cockups in even greater detail )
Tim
-
What Switch settings would you recommend to start with?
I'll be running 24VDC @ 100Amp PSU.
Stepper Motors are NEMA 24 rated at 4.2Amps 4.5Nm, (650 OzIn).
Running a Sieg SX3 under Mach3.
Steppers are driving the LeadScrews at 2:1 ratio. LeadScrews are the original Acme at present. BallScrews will happen after Financial Recovery :ddb:
.
I'm not too sure what you mean by switch settings, the dip switches on the Stepper drivers? or switch relay ratings? or switches like I have for individually switching the power to each axis?
I forgot to say earlier, but if you want even bigger versions of any of my pics you can just click on them and it will take you to my photo site where larger versions are available. ( just so you can see my mistakes and cockups in even greater detail )
Tim
Sorry, I meant the DIP Switches on the Stepper Driver Boards.
.
-
I would start off by setting the current for 4.2Amps. Because you only have 24Volts available, your steppers will not be generating their maximum strength, so keep the amps to the max acceptable for the motors, especially as you have plenty of amps to spare.
As you are gearing down the stepper motors, and are using the stock screws which have a finer pitch than ballscrews have, I would stick to 400 steps per revolution, unless you find that that is not giving you enough resolution when running.
If you find that the motors are running too hot, or the drivers are getting too hot then switch the auto half-current mode on.
I think that covers it all, but these are just my semi-deranged musings, all the standard disclaimers apply and remember that your home is at risk if you leave the doors open.
Tim
-
Power (torque) of stepping motors is related to current not voltage.
Speed is related to Voltage not current.
So if powering your stepper at 12Volts and 4.2A Its torque will be the same as at 48V at 4.2A. In terms of supply current, as the voltage increases the supply current tends to reduce although the current in the motor winding remains constant (subject to how good the controller is).
In order to maintain the 4.2A current at higher speeds the controller needs to switch the current from one winding to another (or reverse the winding current direction). The speed with which it can do this is limited by the inductance of the winding (its resistance to current change) and the voltage applied to it. So higher voltage, higher speed of switching, faster motor.
Heating of the motor is, for the most part, related to the electrical resistance of the winding. Power (heating effect) is equal to the square of the current x resistance. Additional heating caused by induced switching losses in the magnetic material are normally far lower than this and can be largely ignored.
Hope this helps a bit.
Thanks for the pictures, they are very informative. I first used a similar chargepump design 48 years ago on a pdp mini computer, memory lane again .. .
picclock
-
Power (torque) of stepping motors is related to current not voltage.
Speed is related to Voltage not current.
So if powering your stepper at 12Volts and 4.2A Its torque will be the same as at 48V at 4.2A. In terms of supply current, as the voltage increases the supply current tends to reduce although the current in the motor winding remains constant (subject to how good the controller is).
In order to maintain the 4.2A current at higher speeds the controller needs to switch the current from one winding to another (or reverse the winding current direction). The speed with which it can do this is limited by the inductance of the winding (its resistance to current change) and the voltage applied to it. So higher voltage, higher speed of switching, faster motor.
You are right, the voltage effects the speed and the current effects the power. But the thing is, with at least 200 steps per revolution, you are really going to need a good supply of voltage in order to turn at any sort of useful speed. With a lower voltage you can have the full power at very slow speeds, but as you stated the higher speed of switching needs a higher voltage in order to drive the needed current.
It's not like we are talking about some super speeds of table movement. If you think that 1 revolution takes at least 200 steps or switches, then think that a ballscrew may have a pitch of 5 TPI, and a standard acme leadscrew may be 14TPI. That means for a direct driven Ballscrew you would need at least 1000 steps to move an inch and 2800 steps with a acme leadscrew. That is without any micro-stepping or any gearing down which could easily take you to over 10,000 steps or switches to move just one inch.
Just some food for thought.
I got some more done today on the stepper motor mounting hardware. The 1st problem was the raised ring on the stepper motor.
(http://www.velvet-art.co.uk/photos/1056100718_mRdsE-M.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1056100718_mRdsE-A-LB)
(http://www.velvet-art.co.uk/photos/1056100901_onTQ2-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1056100901_onTQ2-A-LB)
So I set up my boring bar on my mill and cut out the needed part. I didnt really calculate it out but just marked out where the edge of the ring should be, then marked the centre and then just edged the boring bar very slightly into the work and adjusted it by eye to be centered.
(http://www.velvet-art.co.uk/photos/1056101005_bYLed-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1056101005_bYLed-A-LB)
Then I did it again for the other support.
(http://www.velvet-art.co.uk/photos/1056101219_XJTBB-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1056101219_XJTBB-A-LB)
(http://www.velvet-art.co.uk/photos/1056101405_JucTV-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1056101405_JucTV-A-LB)
Next will be drilling and tapping for fixings, boring out the thicker plate for the bearing and mounting holes to attach to the mill table.
Tim
-
Tim, a question for you.
I've hooked the BOB to Stepper Controllers as such;
P2 X - Step
P3 X - Dir
P4 Y - Step
P5 Y - Dir
P6 Z - Step
P7 Z - Dir
P8 A - Step
P9 A - Dir
I have setup Mach3 Output Pins as above, but only my Y - Axis moves.
I've checked all My Wiring, which is identical for all four 'BOB - Controller - Cords - Stepper' setups.
Instead of playing around trying to find which Pins are which, ( I don't like seeing Expensive Smoke) :zap:, I thought that I would get lazy and ask you what your Mach3 'Pin Selection' is?
Murray
.
-
Tim, a question for you.
Hi Murray, I havent got my controller pc hooked up at the moment, but I do remember that when I first applied power I couldnt get any of my axis' working. The reason turned out to be that I needed to check (or uncheck, I cant remember which, but it had to be changed from it's initial setting) the option in the "ports and pins" menu for "Active Low" on each axis. That sorted it for me, it may be worth trying to change the setting and seeing if it works for you. Also make sure that each axis is also Enabled in the same "ports and pins" menu.
I'll try and take a screengrab of my settings when I get back into the workshop, but that may be tomorrow or friday though :(
Tim
-
Murray,
One problem that I had that may be similar is that I had to increase the pulse width significantly - I am not near my system at the moment but I think I increased it to Mach3's maximum (is that 15ms??) and that cured many problems. So did esuring that the 25 way D connector was mounted the correct way up on the breakout board - but that is another story... :bang:
Tim,a question since I am now going through the upgrade path from leadscrew to ballscrews - when you described how you replaced the balls into the ballscrew nut, were they replaced through the removed PTFE end cover? I assume that this was the case but your description seems at odds with many other accounts of the difficulties of returning the balls into a ballscrew. Also I recall that there is a 'magic diameter' for a keeper that replaces the ballscrew itself - does anyone know what this is for a 16mm thread please?
Simon (generally a lurker...)
-
Tim,a question since I am now going through the upgrade path from leadscrew to ballscrews - when you described how you replaced the balls into the ballscrew nut, were they replaced through the removed PTFE end cover?
......................
Simon (generally a lurker...)
Hi Simon, yes you have it spot on, I removed the ptfe end cover thing on the non-flanged end in order to re-pack the balls. If you try packing from the other end on these ballscrews it is an exercise in futility that could teach you to utter words that your mother would not approve of :bang:
On these ballnuts that I have, they have split the nut into 3 re-circulating sections, so if you empty out all the balls and want to re-pack them, you just divide the balls into 3 piles and then ( with the PTFE wiper removed from the unflanged end only ) screw the ballscrew into the other end a little bit, then put a third of the balls in and gently rotate the ballscrew back and forth, gradually moving it further into the nut until it has got past the 1st recirculation point, then add the next set of balls and repeat, then the remaining balls.
If you find the the ballscrew will screw in to the flanged end but is pretty sticky when trying to go the opposite direction, this may well mean that there is one or more balls that have got into the section of the ballnut that is between the PTFE wiper at the flanged end and the 1st recirculation loop.
From what I have seen, other ( probably more expensive and higher quality ) ballnuts are so much easier to re-load, with external re-circulation tubes that are removable, making it easy to feed in the ball bearings.
Another tip with reloading the type of ballnuts that I have is to chamfer the end of the ballscrew that first goes into the ballnut, this helps the balls to naturally go into the channels in the ballnut.
Hope this helps
Tim
-
Tim, a question for you.
Hi Murray, I havent got my controller pc hooked up at the moment, but I do remember that when I first applied power I couldnt get any of my axis' working. The reason turned out to be that I needed to check (or uncheck, I cant remember which, but it had to be changed from it's initial setting) the option in the "ports and pins" menu for "Active Low" on each axis. That sorted it for me, it may be worth trying to change the setting and seeing if it works for you. Also make sure that each axis is also Enabled in the same "ports and pins" menu.
I'll try and take a screengrab of my settings when I get back into the workshop, but that may be tomorrow or friday though :(
Tim
Thanks Tim but all my settings are the same for all Steppers, (except the Pin Assignations), Y-Axis is working but the other three are not.
This indicates to me that all my settings are correct except my Output Pin Assignations for X, Z & A.
My BOB came fully built, so the DB25 is correctly mounted.
Your Pin Assignments may be helpfull.
Murray.
.
-
Your Pin Assignments may be helpfull.
Ok, here they are,
(http://velvet-art.smugmug.com/Engineering/madmodder-stuff/ports3/1066152846_q7hAH-L.jpg) (http://velvet-art.smugmug.com/Engineering/madmodder-stuff/8250908_Wivwj#1066152846_q7hAH-A-LB)
(http://velvet-art.smugmug.com/Engineering/madmodder-stuff/ports1/1066152754_MpZse-L.jpg) (http://velvet-art.smugmug.com/Engineering/madmodder-stuff/8250908_Wivwj#1066152754_MpZse-A-LB)
(http://velvet-art.smugmug.com/Engineering/madmodder-stuff/ports2/1066152638_HNB3q-L.jpg) (http://velvet-art.smugmug.com/Engineering/madmodder-stuff/8250908_Wivwj#1066152638_HNB3q-A-LB)
(http://velvet-art.smugmug.com/Engineering/madmodder-stuff/ports4/1066152546_jBqzM-L.jpg) (http://velvet-art.smugmug.com/Engineering/madmodder-stuff/8250908_Wivwj#1066152546_jBqzM-A-LB)
Tim
-
:mmr: WOW, thanks Tim, I'll try these settings later on & let you know. :mmr:
-
Thanks Tim,
Y, Z & A are now all working.
I've just got to chase down why the X-Axis isn't. :scratch:
Is there anything missing on your Screen Shots that might be hiding below the Scroll-Bar?
Murray
.
Edit; X-Axis is now up & running - the 'Resident Workshop Gremlin' wired a Stepper Motor the wrong way. :wack:
.
-
A little update, I havent forgotten this one, just not had much workshop time lately. But I have managed to get out there for a couple of hours this week.
I got started on the thicker mounting plate, boring out the pocket for the Angular Contact bearing, and wonder of wonders, I managed to get a really nice snug, "finger-press" fit. I bored it to the depth of the bearing minus about 0.5mm.
I then made up a bracket to hold the bearing into the pocket out of some 1/8" steel plate, drilled it for 4 mounting screws and drilled and tapped the corresponding holes in the mounting plate.
I then marked up, drilled and counterbored the mounting plate for the side support plates. Then the corresponding holes were drilled and tapped in the side plates.
Then I drilled and counterbored the mounting holes for attaching the whole assembly to the mill table.
Here's a pic of all the parts as they stand today.
(http://velvet-art.smugmug.com/photos/1078628972_4UbK6-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1078628972_4UbK6-A-LB)
And a mockup of them
(http://velvet-art.smugmug.com/photos/1078627960_HNZAC-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1078627960_HNZAC-A-LB)
Closer view of the coupling (minus the lock nuts on the ballscrew)
(http://velvet-art.smugmug.com/photos/1078628384_wgSJi-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1078628384_wgSJi-A-LB)
Next will be a trial fit on the mill itself, and if that goes well it will be on to the Y-axis.
Tim
-
Very nice Tim.
I have actually started on mine again... Thanks for the inspiration!
Eric
-
:offtopic: but we have the same work bench :lol:
yes I have idiot stickers on my mill
Excellent work :thumbup: I'm watching with great interest :thumbup:
(http://i602.photobucket.com/albums/tt102/ja2onx/5cd92dbc.jpg)
-
This is my next step once I've financially recovered from the CNC Conversion. Thanks for all the updates.
Murray.
-
Thanks Eric :thumbup:
:offtopic: but we have the same work bench :lol:
It's a good bench innit :thumbup: But I have a custom bench on the other side of the workshop, it is half beech and half black-leftover-kitchen-worktop with an attractive white joining stripe of silicone sealant.
This is my next step once I've financially recovered from the CNC Conversion. Thanks for all the updates.
Murray.
You're welcome Murray. It is surprising just where the costs mount up, not on the obvious parts like the steppers and drivers, but all the smaller fixings and fittings, bearings and couplings. I've been collecting bits and pieces for over a year before starting on this conversion, I figured that getting the bits gradually would soften the financial blow.
Tim
-
Crumbs, dont time fly by, hard to believe that it is not far off a month since my last update.
I have been working on this, but it has been a bit stop-start (more stop than start) as I have been just grabbing a bit of time where I can to work on it, so this post covers a few weeks of stop-starts.
I mounted the x-axis on the mill and it worked great, all it needed was a couple of 1-2mm thick packing pieces to go either side of the ballnut flange to fit neatly into the pocket from the old acme nut, then jsut tighten up the 2 grub screws and it was fine. Testing went exceedingly well and I am very happy with it.
That being done I replaced the X-axis back to the original manual acme nut and screw, and set about making the Y-axis stepper mounting, which is basically identical to the X-axis, so see the earlier pics for the details. Then I mounted the Y-axis stepper and mount onto the mill.
(http://www.velvet-art.co.uk/photos/1108579484_ryeWU-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1108579484_ryeWU-A-LB)
As you can see, there was a slight design flaw in that the gib screws of the x-axis would crash into the stepper mount, therefore limiting the x-axis by about 4-5mm. So as a quick and easy solution I just cut a little ledge into the mount with the bandsaw.
(http://www.velvet-art.co.uk/photos/1108580198_5euQu-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1108580198_5euQu-A-LB)
And here's a shot of both axis's
(http://www.velvet-art.co.uk/photos/1108580855_PZA8F-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1108580855_PZA8F-A-LB)
So now I can use the X and Y axis as computer controlled to help make the Z-Axis mount. It certainly is easier to just type in a command to move one direction by 72.3mm than counting turns of the handwheel. And being able to spot-drill all the holes at once and then go back to drill all the holes without worrying that you are drilling in the wrong place is just great. I guess it is like having a DRO, but instead of reading off the coordinates of where you want to be, you just type in where you want it to be.
So, onwards and ummm ....... sideways?!?!?
Tim
-
Very good Tim!
Getting there. How's the backlash?
Eric
-
Very good Tim!
Getting there. How's the backlash?
Eric
Thanks Eric :thumbup:
Backlash seems to be pretty low, in my initial tests and setting of steps-per-inch/mm in Mach3, I found that there was about 0.02-0.03mm when changing direction. I dont know if that is final or whether it can be reduced further by proper adjustment of the gibs and making sure that everything is properly tightened up. And if that's not good enough, I believe that Mach3 has backlash compensation that is quite effective.
All I can say is that it is a darn sight better than the backlash under the acme screws!!!
Tim
-
Yup it is. Unfortunately, I am going to be stuck with the Acme Screws for a little while. Making some anti-backlash nuts to take it up some. Mach3 will have to take the rest by adjustment.
I hope to have X and Y done in a week or so.
Exciting isn't it?
Eric
-
Tim,
Have you given thought to upgrading the motor controller for the spindle? This THREAD (http://madmodder.net/index.php?topic=3866.0) Has a link to what I am talking about. I am going to do this as well as a spacer for the column, Will get a little more range out of the Y axis.
Eric
-
Tim,
Have you given thought to upgrading the motor controller for the spindle? This THREAD (http://madmodder.net/index.php?topic=3866.0) Has a link to what I am talking about. I am going to do this as well as a spacer for the column, Will get a little more range out of the Y axis.
Eric
To be honest, I havent thought about it, I am pretty happy with the standard controller, and I have the C6 speed control board from CNC4PC that will give speed control via Mach3. Until it goes :zap: I'll stick with it I think. If I need more speed I might just make a new pulley for the motor.
I have had an idea about getting more Y range, it involves changing the tilting base part of the column with a fixed block of steel/CI with a dovetail in it that the column fits into, thus allowing the whole assembly to be moved back on the base a bit further. That coupled with getting a little more travel to the front should mean getting away without spacing the head out more and gains about 1" of travel.
Tim
-
Hi Tim,
What are the two holes for, that mysteriously appeared when you did the Bandsawing?
Are they for a Swarf Cover to protect the Stepper Coupling?
Murray.
-
Hi Tim,
What are the two holes for, that mysteriously appeared when you did the Bandsawing?
Are they for a Swarf Cover to protect the Stepper Coupling?
Murray.
Ah ha!! we have an eagle eyed member here :D Thanks for pointing that out, I had completely forgot to mention it. I drilled the holes with just a cordless drill handheld ( yea I know, not the "proper" way, but it saved having to replace the acme screws just to drill 2 non-critical holes ) and then tapped them M6, also using the cordless drill.
They are for fitting the original folding "bellows" cover for protecting the ways of the mill, but will also be used to attach a cover to also protect the stepper coupling.
Tim
-
Tim
I've finally figured out from your last photos, what you've achieved.
I was confused to hell & back with the previous posts, but it's now fallen into place :bang:
I was trying to work out where you placed your 'Y-Axis' LeadNut Mounts in relation to 'My' Mill.
The confusing bit? You've got an X2, I've got an SX3.
The SX3 has the Y-Axis Lead Screw mounted to the Moving Axis, yours is mounted to the Base - Duh . . . . :loco:
Murray
-
Murray
If it helps at all, just send me your SX3 and I'll send you my X2, It'd be no bother, honest :lol: :lol: :lol:
Tim
-
Murray
If it helps at all, just send me your SX3 and I'll send you my X2, It'd be no bother, honest :lol: :lol: :lol:
Tim
Wanna pay the Freight from New Zealand??
.
-
Wanna pay the Freight from New Zealand??
Darn it!!! I hadnt figured on the shipping :lol: :lol:
-
Well, two updates in one month, I think I may have to slow down and pace myself on this one :lol: :lol:
So I have been able to make some progress, I have managed to grab a few mins here and there to escape to the workshop.
I have been working on the Z-axis stepper mount. The Z-axis ballscrew is powered via a timing belt reduction of 3:1, so 1st of all I made up a baseplate that had a pocket milled in it for a ballrace so that the stepper-motor's shaft and it's own bearings are not taking the strain by themselves. Then I made up a couple of stand-off's to support the stepper motor itself.
holes drilled and tapped and some partial assembly,
(http://www.velvet-art.co.uk/photos/1138447036_5dqdj-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1138447036_5dqdj-A-LB)
And with the stepper on board,
(http://www.velvet-art.co.uk/photos/1138447043_8MJjt-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1138447043_8MJjt-A-LB)
I then turned to what I thought was going to be one of the easier machining operations, drilling the steel timing belt pulley 10mm for the shaft of the stepper. In fact the drilling was really easy, the problem was I was holding the pulley by it's little spigot in an ER32 collet, and I failed to make sure that it was running true before drilling, and so the hole turned out to be a little askew. When I put it on a stepper and spun it, the pulley wobbled more than my uncle Joe after a night on the black soup ( guinness )
When I then bored it true and parallel it was 0.66mm too big. So I turned up a piece of brass to a sliding fit, mixed up some flux and silver soldered it in place. Then I re-bored the hole, this time I used my 3jaw chuck and used a Dti to make sure that the pulley was held true. Then I drilled it out to 8.5mm and then very gently bored it out to match the spindle of the stepper with a boring bar.
(http://www.velvet-art.co.uk/photos/1138446668_YGifh-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1138446668_YGifh-A-LB)
(http://www.velvet-art.co.uk/photos/1138446509_Am4Uq-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1138446509_Am4Uq-A-LB)
In hindsight I should've bored out the pulley a bit more before soldering in the plug, I had to go pretty gently while I bored out the remainder after drilling to 8.5mm for fear of breaking through the brass or if it collapsed any. I say this because the brass was only 0.6mm bigger than the stepper shaft, so the wall thickness would be only 0.3mm thick and I wasnt sure if the solder had worked its way all the down the brass plug, the last thing I wanted was to go too rough and then end up having to redo it all again.
Actually, in hindsight I should have made sure to drill/bore the pulley in the first place :doh:
So next I will be working on the base that will attach the stepper mount to the top of the column.
I also want to chase down some backlash that I have noticed (about 0.2 - 0.3mm) on the X and Y axis's's's's's's's, I think it is coming from where each of the ball-nuts attach to the mill table.
Tim
-
Well here's the latest update on this one, actually this is the combination of about 3 updates rolled into one ( to save the postage you know :lol: )
I worked on making a top hat adaptor for the Z-axis, I turned it out of a lump of Ali', and bored out the centre to 17mm ( so that the 16mm ballscrew will slide through without binding ). Then I drilled and tapped 6 holes M5 to match the holes in the ball-nut.
Here it is with the Angular Contact bearing that it will sit in.
(http://www.velvet-art.co.uk/photos/1145657815_fzgSg-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1145657815_fzgSg-A-LB)
(http://www.velvet-art.co.uk/photos/1145657869_4dxsQ-M.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1145657869_4dxsQ-A-LB)
and with ballscrew and nut posed
(http://www.velvet-art.co.uk/photos/1145656995_baniF-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1145656995_baniF-A-LB)
I then clamped the large timing pulley onto the mill, and using the PCD hole wizard in Mach3 that I used for drilling the holes in the TopHat above, I drilled the same pattern of holes in it. Then it was over to the lathe to bore out the ID large enough to allow the ballscrew to go through it easily.
(http://www.velvet-art.co.uk/photos/1145656963_mfbuh-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1145656963_mfbuh-A-LB)
and the obligitory posed shot of how all these bits will stack together.
(http://www.velvet-art.co.uk/photos/1145657158_FuTTH-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1145657158_FuTTH-A-LB)
Then I turned back to the stepper mount, or rather the base that the stepper mount will sit on. I needed to have some means of adjusting/tightening the timing belt, so I made an arced slot in the base and used one of the mounting screws that the stepper is held on with. The screw on the left is the pivot point, and you can see the M5 socket head screw on the right in its arced slot. ( again using mach3 wizard really helped in making that )
(http://www.velvet-art.co.uk/photos/1145657903_AvnRB-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1145657903_AvnRB-A-LB)
And that's as far as I've got today. Next will be boring a hole in the base for the TopHat assembly to go through, along with a recess for the big AC bearing, then on to some mounting options for fixing the stepper mount and base to the column.
Tim
-
Is the angular contact bearing dual or single row?
With the top hat adaptor, does it have a second part to hold it on to the bearing or does it rely on being an interference fit?
The Z axis option (http://www.hossmachine.info/images/cnc%20z%20axis%20mount%20color%20coded_800x461.jpg) that I'd been looking at for my X2 conversion has multiple bearings and a two part flange assembly to hold them together. Your option looks a lot easier (and seems to be using the same ballscrews as I have)
-
Is the angular contact bearing dual or single row?
It is just single row, the price of a dual row at this size is unreal.
With the top hat adaptor, does it have a second part to hold it on to the bearing or does it rely on being an interference fit?
Yes it does have another part, I neglected to say that I have to make a steel washer to cap the bottom of the tophat and then drill and tap to hold it in place. There is also a needle roller thrust bearing.
The Z axis option (http://www.hossmachine.info/images/cnc%20z%20axis%20mount%20color%20coded_800x461.jpg) that I'd been looking at for my X2 conversion has multiple bearings and a two part flange assembly to hold them together. Your option looks a lot easier (and seems to be using the same ballscrews as I have)
Actually I am basing my assembly (loosely) on Hoss's plans :doh: I was going to thread the bottom of the TopHat and then make a threaded washer and locknut, but I decided to use hoss's design of using a number of cap screws to hold the washer on to the tophat.
Tim
-
It is just single row, the price of a dual row at this size is unreal.
What's the inner diameter of the bearing you're using? VXB (http://www.vxb.com/) have dual row angular contact bearings but only up to 25mm inner.
Actually I am basing my assembly (loosely) on Hoss's plans :doh: I was going to thread the bottom of the TopHat and then make a threaded washer and locknut, but I decided to use hoss's design of using a number of cap screws to hold the washer on to the tophat.
Tim
Look forward to seeing your final design.
-
What's the inner diameter of the bearing you're using? VXB (http://www.vxb.com/) have dual row angular contact bearings but only up to 25mm inner.
The ID is 25mm, I guess I could have got a dual row AC bearing, but I dont have a lot of depth to spare with this design, so I settled on using a single row AC bearing and using a needle thrust bearing on the underneath to hold the whole assembly firm to the column-mounted stepper base.
Tim
-
Just to do things a little differently, I'm making todays update in reverse!!!
So to show you the assembly of all that I've got done on the Z-Axis,
(http://www.velvet-art.co.uk/photos/1147721043_GgYG9-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1147721043_GgYG9-A-LB)
Then to break it down a bit, here's the ballnut removed showing the 6 holding bars that are threaded M5 on each end.
(http://www.velvet-art.co.uk/photos/1147720345_fao8P-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1147720345_fao8P-A-LB)
And with the timing pulley removed.
(http://www.velvet-art.co.uk/photos/1147720310_G46r5-M.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1147720310_G46r5-A-LB)
Then whipping the remaining assembly over we have the washer that holds the thrust bearing on to the TopHat.
I made the washer out of some 3mm steel plate, firstly drilling the 8 fixing holes and then holding it on the tophat and then turning down the outside diameter, then I held it in the 3jaw and bored out the inner hole, and then bored a little recess.
(http://www.velvet-art.co.uk/photos/1147720065_SdBbC-M.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1147720065_SdBbC-A-LB)
(http://www.velvet-art.co.uk/photos/1147720654_TyG3Y-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1147720654_TyG3Y-A-LB)
And the obligitory shot of everything in bits.
(http://www.velvet-art.co.uk/photos/1147721101_dmLwV-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1147721101_dmLwV-A-LB)
Tim
-
Well I have been a naughty boy, I have been chipping away at this one without taking any in-progress pics. So now I have a glut of photos to show today. As of today the mill has all three axis' cnc'd and working.
So here is a pic of the Z-axis in place. As you can see I have used a couple of side "cheeks" to hold the Z-axis stepper mount in place.
(http://velvet-art.smugmug.com/photos/1173276578_BSmpd-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173276578_BSmpd-A-LB)
To attach the Z screw to the top of the mill head casting I chose to turn a M8 thread onto the end of the screw and then use a locknut to grip it tight.
When I drilled and tapped the mill head by hand, some idiot ( me ) drilled the hole slightly off centre, by about 1.5mm :doh:
So the solution that I came to was to make a little off-set adaptor,
(http://velvet-art.smugmug.com/photos/1173276471_RPkuc-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173276471_RPkuc-A-LB)
and secured both adaptor and screw with locknuts.
(http://velvet-art.smugmug.com/photos/1173275543_yzRwp-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173275543_yzRwp-A-LB)
With that done and the whole mill now working on computer control, I am now turning to looking at converting my Rotary Table.
My RoTab is just a real cheap one, so what I do may not be the way I would do it if it were a more pricey model, but I am just working with what I have.
I started by breaking the rotab down into its seperate pieces, and then I took the shaft that the worm is fixed to and shortened it and turned down a section to 6mm (to fit the bearing and oldham connector).
(http://velvet-art.smugmug.com/photos/1173276665_hHqh5-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173276665_hHqh5-A-LB)
As you can see here I have left just a smidgen of the 10mm diameter sticking out.
(http://velvet-art.smugmug.com/photos/1173272593_n25ks-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173272593_n25ks-A-LB)
This is for the bearing to rest up against,
(http://velvet-art.smugmug.com/photos/1173272691_ogk5u-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173272691_ogk5u-A-LB)
Then the oldham connector is attached ( when final assembly is made I will clamp the oldham on the shaft while it is under some pressure, thereby gripping the inner race of the ballbearing which is in turn held up against the 10mm part of the worm-shaft, clear as mud eh? This is hopefully engineered so that when everything is tightened up in place, the worm will be held away from the inner face of it's guide just enough so that it will not bind up (as it had been doing at a certain point each revolution under manual activation) but the bearing will take up any lateral movement, or so the theory goes )
(http://velvet-art.smugmug.com/photos/1173272534_2dJ95-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173272534_2dJ95-A-LB)
Then I made a steel washer that slides easily over the oldham coupling but butts up against the outer race of the bearing.
(http://velvet-art.smugmug.com/photos/1173273301_9fXow-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173273301_9fXow-A-LB)
Then I turned up an Ali' spacer
(http://velvet-art.smugmug.com/photos/1173273605_ZFP4r-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173273605_ZFP4r-A-LB)
The reason for the hole will become clear,
(http://velvet-art.smugmug.com/photos/1173273605_ZFP4r-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173273605_ZFP4r-A-LB)
Spacer in place, showing that the hole allows you to tighten and release the oldham coupling that mounts to the stepper shaft.
(http://velvet-art.smugmug.com/photos/1173274339_tSRHE-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173274339_tSRHE-A-LB)
Then I made a stepper mount out of some 10mm ali plate, This was done via semi-cnc using some of the wizards in Mach3 and also by just using single G-Code commands.
This is the side that the Ali spacer fits into. The outer 4 holes just need to be tapped M5 for the screws to hold the stepper, the inner 4 holes are for threaded rods to hold all the stepper mount pieces together and to the RoTab itself.
(http://velvet-art.smugmug.com/photos/1173274604_PfByM-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173274604_PfByM-A-LB)
This is the side the stepper mounts to, you can see the circular recess for the boss on the stepper, and the 4 large counterbores are to allow clearance for the nuts to tighten the whole stepper mount onto the rotab.
(http://velvet-art.smugmug.com/photos/1173274966_Hm9So-M.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173274966_Hm9So-A-LB)
And here is the obligitory mock assembly of the whole lot. I just need to make the threaded rods to hold it all together.
(http://velvet-art.smugmug.com/photos/1173275370_n9Vvo-L.jpg) (http://velvet-art.smugmug.com/Engineering/CNC-conversion/13513205_C5ZYg#1173275370_n9Vvo-A-LB)
Tim
-
I had a little time this afternoon and so I took some 4mm stainless rod that I had and cut 4 lengths and threaded both ends. Popped on a few nuts and then assembled everything together.
(http://www.velvet-art.co.uk/photos/1174618098_VPzJh-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1174618098_VPzJh-A-LB)
I am very pleasently surprised how this has improved this little table. It no longer rocks on its axis, and the backlash is much reduced too. Even if you dont want to convert one of these to cnc, putting in a angular contact bearing will really reduce the backlash even in a manual machine. A reasonably cheap way to make a big improvement on a cheap Rotary Table.
Tim
-
Contrary to popular belief, I havent been doing absolutely nothing since the last update, just doing very little :lol: :lol: :lol: :lol:
I have been using the mill in a semi-cnc mode, mainly using manual commands, and even have used some of the wizards in Mach3 which are very handy.
For the last few days I have been making a couple of mounts to make the connection between the X and Y ballnuts and the mill table.
This was where mach3's wizards came in very useful. I used one that cuts an arc of whatever radius you choose and from whatever start point for the desired degree of cut. Then I used a circular bolt hole drilling wizard that does exactly what it says on the tin.
The result:-
(http://www.velvet-art.co.uk/photos/1205664070_n5hpB-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1205664070_n5hpB-A-LB)
and it fits in here for the X-axis
(http://www.velvet-art.co.uk/photos/1205664477_sdJ72-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1205664477_sdJ72-A-LB)
Then using the above mentioned wizards as well as a hole boring wizard, to make a larger mount for the Y-axis.
(http://www.velvet-art.co.uk/photos/1205664186_rdYSR-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1205664186_rdYSR-A-LB)
Then I had a bit of whittling to do as the slot in the mill base casting was slightly narrower, so once I had broken down the mill to get at the Y-axis ballscrew I put the mount on the "bandsaw" diet and trimmed down the sides. A few mins with a file had it fitting ok. A bit more filing to get the fit right and it was time to fit the ballnut into the mount.
This presented another problem, the little space in the casting of the mill that this mount will fit into is only about 2-3mm bigger that the mount, so how was I to fix the ballnut firmly to the mount. I am able to use 3 cap-screws at the bottom, but for the top I trimmed down some cap-screws and then filed a couple of flats to use a spanner. And whatya know, it worked!!
(http://www.velvet-art.co.uk/photos/1205664207_Ed4pr-L.jpg) (http://www.velvet-art.co.uk/Engineering/CNC-conversion/13513205_C5ZYg#1205664207_Ed4pr-A-LB)
The result of doing all this is that my backlash has reduced from 0.22mm to 0.1mm on the X-axis and from 0.34mm to 0.15mm on the Y-axis. I think that it was worth it. I still use the Backlash Comp in Mach3, but when I want to do some small scale engraving with the mill it should make it much better.
Tim