DIY bluetooth DRO for my mill

[vtsteam]

A few years ago I started a project building a low cost bluetooth DRO using a mix of inexpensive digital calipers and an IGaging  scale using plans and instructions found on Yuriy's Toys website. It was a pretty complicated project, mainly because the calipers needed a big hand wired adapter board to interface with a small Texas Instruments controller board.

I built and tested the system, but never got around to mounting it on a machine. One of the calipers was unreliable, and I didn't have much more money to invest in the project.

A couple years ago, when the prices came down, I bought 3 IGaging capacitative scale units, intending to use them as-is, but again, never mounted them on a machine. Fast forward to yesterday, while working on the mill I got irritated trying to read the dials in low light and fighting backlash, I decided then and there to build and add a DRO, but this time, to use both the new scales and parts from the old controller project, but without the caliper adapter board.

So I put together this unit. It has the TI processor board, 4 micro USB ports for X, Y, and Z scales and a rotary encoder. The scales have a backup battery of 2 AA cells, which should provide 2500 hours before needing replacement.

The bluetooth connection is provided by an HC-06 breakout board on the right side of the unit. There are two Shotky diodes between the battery and the onboard USB power supply input to prevent battery charging or backfeeding the battery power to the processor board. Batteries only power the scales which draw 17 uamps. A 2200 ufd capacitor maintains the power during transition between usb power source and battery power to the scales. All of this is detailed on the Yuriy's Toys site

For my own preference, I mounted everything on a vectorboard and cut and wired  traces by hand to reduce the number of jumpers. I did keep some jumpers in order to make the system flexible in case I want to change the port assignments and inputs.

I tested it out tonight and it works great with the Igaging scales and an old tablet I have. The Android software for the display is called TouchDRO. It also works with my cell phone.

The first photo below shows the older mixed caliper and scales project while being worked on.

The second photo is the new DRO project which works with the Igaging scales.

[PekkaNF]

Pretty nifty

You have this arragement: "The scales have a backup battery of 2 AA cells". I supose it is because scales send absolute position serial signal instead of incremental position?

Good work,
Pekka

[vtsteam]

Yes Pekka, these IGaging or Shahe type capacitative (inexpensive) DRO scales store absolute position and transmit that. They also have extremely low power consumption.

Apparently the system can also work with magnetic and glass scales, and quadrature encoders and also with Hall effect tach sensors (for lathe speed), but I believe glass scales only provide incremental position changes, like a typical encoder. You would not be able to store a position using these scales with this system during a power outtage without battery backup for the whole processor board.

Here are some screen shots of the TouchDRO application on both a tablet and a phone. The various hole pattern functions are at the bottom of the screen on a tablet and can be scrolled horizontally. On a phone they are arranged vertically, and can be scrolled through for selection that way

My setup uses an older version of the MSP430 board firmware (1.3). I found that the later version, 2.0 did not work with my scales because, I believe that version update stopped providing a clock signal from the microprocesor board (formerly available board on pin 1.7) that the scales need to receive in order to send data back.

Maybe a separate hardware clock pulse could be arranged. I don't know much about that -- it is above my electronics knowledge level.

Screenshots of a few of the various functions available in the TouchDRO application on tablet and phone:

[vtsteam]

I just made up a small plywood box to fit the bluetooth DRO.

I was thinking about lining it with aluminum foil as a shield (except at the end where the bluetooth transmitter is), but I think I'll just try it out without first and see how stable it is.

[PekkaNF]

Yes Pekka, these IGaging or Shahe type capacitative (inexpensive) DRO scales store absolute position and transmit that. They also have extremely low power consumption.

Apparently the system can also work with magnetic and glass scales, and quadrature encoders and also with Hall effect tach sensors (for lathe speed), but I believe glass scales only provide incremental position changes, like a typical encoder. You would not be able to store a position using these scales with this system during a power outtage without battery backup for the whole processor board.

I have traditional DRO with magnetic scales. They are the normal incremental position transducers (A/B and index). I never found it difficult to "zero" on manual machine. CNC needs to home or retain the position. On milling machine I center the part and zero to center of the hole or part..or..zero to vice back jaw (touch jaw and move half diameter, then zero). I am not sure if this is right way, but works for me. I'll learn something else when this stops working.

That big screen interface looks really good. My 7-segment display is pretty unintuitive to use on PCD and such. Zeroing, centering and setting cordinates works well.

[vtsteam]

I once fixed a defunct mill with an Anilam Crusader DRO at the place where I worked. It kept losing position and doing weird stuff. I asked if I could have a look at it. I took the boards out of the control head and they turned out to use some unusual static memory IC's with internal batteries. Since these had aged, the batts stopped maintaining memory contents, and the DRO started having issues. I located some replacements, which had an explicit spec for ten year life, and the machine was usable again.

I later fixed a thin film deposition machine that had similar memory problems, though more easily figured out because the batts were external to the memory chips. I couldn't find replacements for those specialized batts, so I added new batt holders for a more common cell type with similar specs.

[vtsteam]

Here's the new box, painted.

[vtsteam]

I've finally started mounting the scales to the mill -- a typical heavy Asian round column mill drill, this one happens to be an Enco branded version, which I bought second-hand a dozen years ago from a repair facility. They also had a Bridgeport, and said they hardly ever used the mill drill -- which seemed likely from its like-new condition. The only problem with it seemed to be that the R8 collet retaining bar had broken at the top end and had been brazed together again. Not sure how that could have happened unless it was off the machine when it broke. Anyway, I considered it a good deal at the time, and I got a mill vise and a box of useful accessories into the bargain.

So, back to the scale mounting, today ..... I debated for awhile whether I wanted to mount the X scale to the front, or the back of the carriage, and finally decided on the front. That way I get full Y travel, because otherwise the scale reduces the free distance to the column.

One method of mounting a scale on the front seems to be to re-purpose the stops, which are on the front face of the table. These fit into a dovetail groove that runs the full length of the table. There are two stops, retained by Allen screws into dovetail nuts, and they can be set anywhere. I think most people use them merely to prevent running the mill table too far off the end.

I happen to like those stops a lot, and use them frequently when slotting on the mill. They allow easy milling of slots with clean ends without having to keep an eye on the dial. While I imagine that with a DRO it's easier to know when to stop at either end of a slot, you can do it by number. Nevertheless, I'd rather have a physical stop, and not have to constantly check the readings at all. It definitely speeds up the work when it takes say 20 passes to fully mill a deep slot.

So, though replacing or repurposing those stops to support the scale would make installation easy, I didn't want to do that. Instead, I decided today to make a second set of dovetail nuts (kind of like tee-nuts), and some solid stand-offs to support the scale and use the same dovetail slot that the stops slide in. That way I can retain the stops, and still support the scale by that simple method, and there would be no need to drill a hole in the table. The standoffs would be long enough to space the scale far enough away from the stops, so I can get an Allen wrench in there to tighten or loosen them when I need to. The standoffs would be at the extreme ends of the table -- the stops are free to slide and position in the rest of the length of the table.

I decided therefore that the standoffs should be 1.25" long, and made from some 1"x1" solid aluminum bar stock cutoffs  from a scrap barrel that I won at a machine shop auction.

I measured the dovetail nuts that the stops used, and they turned out to be 1/2" wide, and had 15 degree tapers to fit the dovetail slot. I milled these dimensions into some hot-rolled steel strip and sawed off two inch-long nuts, and tapped them 1/4" -20.  I also drilled the 1-1/4" aluminum stand-offs 1/4" and countersunk the ends. By evening I was able to screw the standoffs onto the table using the dovetail nuts, and it looked quite sturdy, clean and simple. Happy to get that particular slide mount figured out.

 

[PekkaNF]

I want to see pictures of that!

I have probably same mill/drill on green colour and I was wondering the same things. I put the scale on the front:
https://www.madmodder.net/index.php/topic,12813.msg153772.html#msg153772

There are ramblings about the location of X-linear scale and factors starting at the end of page#1. I noticed that I had my nose hard at the machine and didn't notice that thread became more of a build log than actual structured content on OP.

My main drivers were:
1: To maintain full travel of the axis
2: To protect the magnetic scale
3: To put magnetic scale onto moving part to minimize reader head cable movement.

Only compromise was with X-axis stops.....I though it hard and there has been very few ocacions, where one (and once even two) stops would have somewhat usefull, but:
1: Hard stops are way more usefull with dials, DRO actually helps a lot
2: Plan "B" is to build stops on the bottom of table dowetail ....sort of clamps that clamp on the table dovetail and bump into a saddle.  Haven't needed it yet / Haven't build it yet.

Pekka

[vtsteam]

Hi Pekka, very nice looking mill and photos compared to mine! 

I initially thought the same as you about maintaining full travel, but after playing around with the parts and positions of the stops and mounts, I realized I don't actually use or even want full travel on the X axis, end to end. Why? I normally position my stops to prevent over-travel.

There is not much support of the table as it nears the maximum possible travel. The table has 14" of support normally. If cranked over to the bitter end, that is reduced to only 7". If one has a 70 pound vise near the far end, plus the table weight itself, that is putting heck of a lot of stress on that cantilever for only 7" of support. The table is cast iron, not a great material in tension.

Also, because it is a round column mill drill (which has its well know disadvantages) it also has one advantage in that the head can be swung over to either side a couple inches with little loss of Y travel, to compensate for a far off-center part location on the table, thus compensating for any loss in travel caused by the scale or reader. And in fact that allows more table support -- for two inches of head swing you get 9" of table support instead of 7"

The only possible disadvantage for me personally would be if I were to try to surface a very long piece -- like the head of my tractor engine -- which I did a few years ago (some of you may remember the thread here), But at that time I actually wasn't aware that my table stops had been set conservatively, and I still managed to skim the head true by use of a large diameter homemade fly cutter (also documented there). Another possibility which didn't occur to me then was that I could have made up a column stop, so that swinging the head did not lose vertical position.

Anyway, I don't want to give up the advantage of the stops (or the carriage clamps) so simply spacing the scale out from the table is a very simple solution. The spacers are very simple to make. See photos below, and one where I'm holding a short scale up to give the relative positions of the spacer, an Allen wrench for adjusting a stop, and the table clamps.

All fit, and I believe actual table travel with the stops unscrewed from their dovetail nuts will be about the same as it would have been if there was no scale and the table stops were positioned at the ends of the dovetails. Well minus the length of the reader head, which is unavoidable in any DRO setup, unless the scales are longer than the table.

Although for my setup when you think about it, the reader head isn't much wider than the central stop bracket on the mill already.

In sum, I think I will see very little difference in useful table length with this setup, and any loss of extreme travel can be compensated for.

Phew.....sorry, pretty long winded for a very simple setup! 

[PekkaNF]

Very nice to read your though process. You clearly have it squared out. You mentioned tractor head...that is big boy on this size of mill. Last week my brother bought a big hunk of aluminium....He wants to mary Jaguar engine and Ford gearbox. I showed him how to operate the mill and 6 hours later enough metal was removed.

The hunk of metal was way too big to mill on one setting. Because bottom side of the blank and mill table is flat enough it was possible to mill one part at the time....took four different mounting position to surface spacer one side once.

Most was cut with a metal circular saw mounted on arbor, then 50 mm four insert mill was used to surface. Pretty much max travel and max distance to column was used. Large vacuum cleaner bag and most of the floor was filled with swarf and even when table tops we covered with sheet, plenty of aluminium swarf was flung all over the garage.

[vtsteam]

Haha, well that sounds familiar -- getting more out of a machine than it was ever intended for!  If I remember correctly my tractor head was longer than the tee slot part of the table, and I super-glued square tube bearers on top of the table so I could fasten the head down to something. 

Well for another crazy head project, I could always remove the two screws that hold the DRO scale on, so again, the DRO is no detriment to max travel. No need for even an X measurement when surfacing a head, and the dials still work anyway. Z is the only thing you need for decking a head -- if even that.

Back to the present project -- today I wasted time deciding that 1.25" was too much length for the standoffs -- I realized that the supplied scale brackets added 0.4" additional that I didn't need. So I decided to mill the standoffs back to 1.000" and that looks a little better. Then I located the positions I needed to drill the standoffs to accept the supplied scale brackets, drilled and tapped those, and assembled the scale to the mill with 10-24 scvrews. All worked well.

Next task, attach the reader to the mill's stop bracket. 

[efrench]

The Shahe scales I have don't have very good accuracy (about .1mm) and one of them would randomly jump 5mm.  They now reside in a junk drawer and were replaced by glass scales.  That made a world of difference.

[vtsteam]

Well I guess we'll find out what my setup does. The scales are the later stainless steel types, and the Igaging readers have been eliminated in exchange for the TI  MSP430 launchpad board and firmware. Also filter caps have been added in the reader heads. There are a lot of variables.

BTW if you have scrapped your old scales, don't want them, and postage is reasonable, i wouldn't mind seeing what i can do with them.

[vtsteam]

This morning I measured the distance between the reader head and the fixed stop bracket on the mill. I then cut off another piece of the 1x1 aluminum bar that length and cut a channel in it to form a U bracket.

The reader head takes some very short, supplied 3mm screws, so I milled that wall .064" then drilled for the screws which needed a 20mm spacing. (heh, lots of mixed measurments). For the other face I added two side slots to pass a screwdriver, and the vertical slot to take the screw that will go into the mill's fixed stop bracket. I'll be fastening that with an 8-32 screw.

[vtsteam]

And here's the X axis reader mounted to the mill's fixed stop:

[vtsteam]

DTI measured scale resolution appears to be ~ 2640 lines per inch, or about .0004" (.01mm) and repeatability is excellent. No display glitches. I haven't fully calibrated it yet, but quite pleased with that.

[efrench]

That's certainly better than mine which were probably from the bottom of the barrel on AliExpress    I'll probably try to use mine on another project.

[vtsteam]

One difference in using the original displays vs using this shop built display is that I can set the calibration in software. ie. how many lines per unit length as a setting. This is likely to vary for many reasons, including even the physical setup.

If you have a consistent error per length rate then I would suspect the need for calibration, and the scales themselves may be better than they appear. Since the original scale heads do not allow calibration, you might consider a project like this one.

[vtsteam]

On to the Y scale mount:

I just didn't want to drill into my carriage horizontally in my cramped space and tap blind holes into cast iron in situ. I also didn't want to remove the table. Yet I wanted to mount the scale on the carriage and mount the reading head stationary on the base. It took a lot of thought to come up with something, but it occurred to me that maybe some kind of clamping fixture would work -- maybe fabricated out of steel angle and strip.

Below in the photos is eventually what I worked out. Not the most elegant piece of work, but it does what it is supposed to do -- it's a close fit on the extended bearers, and a couple of setscrews lock it in place. I did drill dimples into the castings for the setscrews to register in.

The main body is 1-1/2" angle iron. The downward facing leg will be where the scale mounts, inside the angle. The horizontal leg above will shield it from swarf.

The angle iron isn't quite square to the mill surface. It is rotated 8 degrees from vertical, because the cast iron bearer surfaces were at that orientation, and in fact so is the mill base edge -- that's the draft angle it was given when cast. This is quite convenient since with both the scale mount and base surface at the same angle, reader head will be straightforward to mount, and I won't need an angled bracket.

Well, I'm definitely glad to get this puzzle worked out after a lot of head scratching, and trial fits. The scale will go on tomorrow.

[vtsteam]

I had to shorten the Y scale by 4" to fit my mill. That was easy to accomplish with just a hacksaw. Then I reversed the end brakets so that the back of the reader head faced out, I drilled and tapped the new new Y axis mount, which I'd painted last night. Here's the assembled scale and mount:

[vtsteam]

I made a fixed bracket for the read head and mounted that to the mill base. With that the Y scale installation was complete.

[vtsteam]

I did a quick test of the Y axis using the same resolution setting I had for the X. I got different results which surprised me until I remembered that the X calibration was against a DTI set on the table and bearing against the quill. This time with the Y, I was just looking at the dial readings on the crank. I then checked the X against dial, and got a similar result. So that begs the question, which is corret, the cheap DTI, or the dials? Or neither.

I'm in need of a standard......

[PekkaNF]

Yesh...standards are good. Do you have two micrometers that are larger than 1"? Their calibration rods are remarkably accurate. I also have found that bearings are made to very exacting standards and their tolerances are according to published tables and nominal diameters are comfortable numbers.

other thing: Those scales are capacitive, right? I have used the very little, but I think that they are very sensitive to scale-head distance variation....therefore the head should "float" and no load to twist or influence perpenticular to scale should be present. In normal caliper that is no problem, because the body is pretty effective guide rail.

Pekka

[vtsteam]

Thanks Pekka. I do have a big micrometer, I was thinking I'd have to make a bar up with it, but if I have a cal rod with it, I'll use that. In mounting, I've been very careful not to bind the scales in any way, but we'll see.

I think friction or binding error would show up as scale reader backlash, in other words not returning to the same point after a traverse and then reversal, vs calibration error which would show up as a one directional measurement error.

The main thing I'll be looking for, after calibration adjustment would be inconsistencies for traverses of varying lengths and their returns. That should show what the limits are.

I think I remember reading something somewhere that the leadscrews on these mills is actually a metric approximation of inches, and the inch dials will show some error, but I'll figure out -- will also check my DTI, out of curiosity.