Servo Driven 4th Axis for CNC Mill

[awemawson]

All I had to do was mill a matching 3/16" slot on the other side and I could assemble the bits the other way round !

[Manxmodder]

Hi Andrew, the plastic knob is is very likely to be made of a glass filled plastic as many structural plastic components are(PA 66 maybe?).

It is very abrasive and dulls most tooling very efficiently....OZ.

[dsquire]

As far as I know I've never set anything up for longtitude on the forum other than put the country as UK. I've just had a poke around my profile and found a setting to have an offset from 'forum time' with  :Dan auto-detect feature that says it is to display local time. Clicking this gave a four hour offset so maybe this post will bear UK time.

Later edit - Nope shows 4 am ish !!!!! unless the 24 hour clock is wrongly implemented and it means 16:00, but it shows as 04:.. which is early hours in the dark  even though it SAYS PM !

Andrew

Local Time:March 18, 2014, 05:57:21 PM (copied from your profile)
Local Time:March 18, 2014, 01:58:21 PM (copied from my profile)

Note the difference of 4 hours between your profile and mine. If you right click on my profile and compare it to yours you will see that there is a four hour difference which will be the reverse of that shown above.

Hope this helps Andrew, now back to those servo drives and sheep and whatever else may come your way. I know all about this retirement thing. I don't know how I ever had time to work.  

Cheers 

Don

[Hazel]

Hi Andrew,
I am very impressed with your 4th axis upgrade. I know this is an old thread, but I am about to do something quite similar. I have the choke, the servo card and a J&S 4th axis. Before I start putting all the bits inside my BP Interact 1Mk2 I thought I'd try to get the encoder at least working first . I can get the IVth axis to display on the TNC151, and I can tell it to display encoder outputs for inhibited axes, but my spare working encoder does not work in the 4th axis. It works if I swap it into the X encoder plug if the X axis is inhibited. Is there a set of Machine Parameters someplace that I can set to check the IV axis encoder?  Currently using MP72=9, MP75=0, MP76=1. Not really sure what I am doing to be honest.

Regards, Hazel

[awemawson]

Hazel,

I'll pop over to the workshop later tonight and see how those parameters correspond with mine

[Hazel]

Thank you very kindly. If you can think of a way to test the 4th axis encoder without the rest of the 4th axis attached, that would also help me out a bunch, I've spent all day today trying to get this encoder to run.
-H

[awemawson]

Hazel,

I assume that you have a manual with a list of the parameters and their make up? If not it's freely downloadable from the heidenhain.de web site.

I've just checked my parameters against those you list:

MP72=9, Mine = 0
MP75=0, Mine = 0
MP76=1, Mine = 0

I've just dumped all my parameters to file and I attach a copy. Remember - mine is a TNC355 not a TNC151. I think that the lower numbered parameters are the same but best to check

[Hazel]

Hi Andrew,
thank you, thank you, thank you! I do have all the 151 manuals from Heidenhain, which are kind of sparse. Installation/Interface, Service Instructions, operators book, plus all the Bridgeport stuff. Your MPs being zero makes sense, as none of your axes are inhibited. I'm right in the guts of my TNC151 analogue board, tracing through the 11uA buffers to a bunch of comparators, which I can only guess is some sort of ADC to do the phase-quadrature interpolation.  Next step is to power up the analogue board with an encoder and check the comparator outputs to compare Z and IV.  Maybe I should save up for a 355 control...
Regards, Hazel

[Hazel]

Some photos of my Mad modding:

[awemawson]

I think that logically this bit of the two controls is the same, although the implementation is different. IIRC the encoder inputs on the 355 wire directly to custom lsi modules just onboard from the connectors - when I had a 151 I never pulled  it apart

On my 355 in it's Beaver Partsmaster there is a user parameter (165 I think) that enables / disables the 4th axis - let me have a poke around in my hard - drive ....

... no I was wrong - it's parameter 158 - copy of decode attached

[Hazel]

Ooo that is interesting. All the 151 blurb says is "MP158=Setting of 16 markers to binary number" which isn't so very helpful. The User manual simply says to activate the 4th axis from the user parameters. I've sent an email out to a repair company to see if they do an exchange on the board, as it may save a month of my life
H

[awemawson]

They are probably machine specific - Heidenhain just provides the facility for manufacturers to have 'flags' that the user can set that are (usually) used as inputs to the PLC ladder logic

Somewhere buried in the parameters is one that sets the count per rev of the encoder - iirc I harp on about it in the pages above when I'm making the belt drive for my encoder to change from 125 to 100 pulses per motor rev. If that hasn't been set and left blank it may explain your problem. There are only two permissable values and in my case it was 100 x 4 x 90 = 36000 pulses per table rev. I think 18000 is the other one

[Hazel]

Hi Andrew,
Sadly nothing in the TNC151 documentation directly says how many counts the rotary table should be. There is an option (MP15)  to change the interpolation of the 4th encoder between 10 and 20 divisions , and another (MP90) that says to display the 4th axis as 0 to 359.999 for options A,B & C . So I guess that would be the 36000 and 18000 options. ? One of the documents also mentions that the 4th axis reference end position be tied to 24V, yet my rotary table does have a normally closed index switch - how is this supposed to be wired then?

regards, hazel

[awemawson]

Hazel,

I've pm'd you some info I got when I was struggling.

Also picture below is of my home switch wiring - note it does ref pulse inhibit and home on one c/o switch

[Hazel]

Hi Andrew, again, many thanks. My BP Interact 1.2 only uses the reference end positions for X,Y. Z and I think it uses e-stop if you go over the end of the ref point.
I've given up trying to get the Heidenhain repair places to return my emails. Instead I squandered my time wiring up the Analogue PCB to the appropriate Voltages, and stuck a SEM motor into each axis to see the voltages. I managed to trace it down to a little red module, which I think is a capacitor array or maybe a transorb input protector array?

[Hazel]

Okay, in for a penny so to speak, I decided to repair the module, seems to be SM capacitors. Reassembled and all the front end 4th encoder amps now work the same as the others!
So, now to reassemble the mill...

[awemawson]

Well done that girl 

[Hazel]

So now I've put the TNC151 back in the mill, connected an encoder to the 4th axis , and after the usual referencing I now get counts in the 4th axis, and sure enough they go from 0 to 359.999
Yay. Of course I still have to figure out the axis home signal, and put the choke, servo card and amphenol connectors into the cabinet, oh and repair the encoder in the 4th axis, which has a dead bulb.  I also have to put the spindle bearings back in as I had a large amount of play while I was trying to calibrate my home made touch probe.  Seems like a long time since I had any chips on the garage floor and in my hair...

[awemawson]

Hazel, do you have a particular 'thing' that you need to make, or is it, (like me) the journey that you enjoy ?

[Hazel]

I do have a project in mind - A 5 cylinder radial engine, which will require some accurate placement of lifters to get the sound I want. 

[Hazel]

Hi Andrew,
I'm just installing my Z15 Servo card, and just discovered that the card does not have screw terminals on it, but takes another connector with screw terminals on it. I found something in Farnell, but the pins are much thinner than the other ones in by axis cabinet, which are more like 1.5mm thick. The farnell ones are only 0.9mm thick according to the datasheet. I don't suppose you recall where you got yours from?



I Just realised the same connector system is used on the rear of the TNC151. Which is anooying, as I removed the interface PCB and ribbon cable to avoid disturbing the cabling - and all I needed to do was pull the connectors out - D'oh... Oh well I only did that about a dozen times while I was repairing the 4th axis encoder interface.

[awemawson]

Hazel,

I used these from Rapid:

http://www.rapidonline.com/Cables-Connectors/CamdenBoss-CTB1050-12-12-Way-24A-90-Interlocking-Terminal-Block-5mm-Pitch-21-4625

They are not a perfect match, and infact I have put a dab of 'hot melt' glue bridging the male and female housings to ensure that they don't vibrate loose. They've proved ok in practise.

[Hazel]

Ah, those are the same as I discovered  in Farnell. I think my plan is to get some of the 5mm pin header stuff which has 1.5mm diameter pins and solder it to one of the Camden thingies.  Many thanks !  Now I have to wade through all the din terminal blocks to find some 35mm rail ones to fit the 3R3 motor brake resistor. Fun Fun Fun!

[awemawson]

Well it's been a LONG time in the gestation, but at last, as it hasn't stopped raining for days, I've found a bit of time to sort out a suitable 'Post Processor' to allow me to engrave text on the outside of a cylinder mounted in the 4th Axis. 

So what's the issue 

Well, if you use a CAD/CAM set of programmes to generate the bits you want to make on a CNC mill or lathe, the program has to spit out code that meets the specification of the controller that drives the machine. So most CAD/CAM set ups use an intervening set of definitions and code that take in the symbolic actions, and translate them to actual code that the controller can understand. Once defined and installed, from then on the CAD/CAM should (hopefully) produce the right stuff for the controller.

Now there wasn't a 'standard' Featurecam Post Processor suitable for the TNC355 controller in my mill that understood the 4th axis. There were ones that drove 4th axes. And there were ones that were close to the TNC355 (as the TNC150/151/155 is very similar, so it was a case of making a Frankenstein version, with a few tweaks of my own.

It didn't help that some of the 'standard' Post Processors made mathematical errors when wrapping stuff round a cylinder  All sorts of Gotcha's appeared - like the one that made moves all the way round the bar backwards to get to somewhere close to where it started, cutting a nice groove as it did it  Anyway much  and rather a lot of  got me there in the end.

So what are we talking about: well the first picture below is a screen shot of a very small fraction of the program like structure of the Post Processor, and the second picture is code suitable for my Heidenhain TNC355 to obey

[awemawson]

Now there is still quite a bit of tidying up to do to generate more compact code - but hey - it works

The 32 mm bar in the picture below has been engraved with my initials 'AWEM' using a diamond 'drag engraver' - quite hard to photograph curved surfaces but I'm sure that you get the general idea.

If you look at the over view picture, you'll see several failed attempts along the way