An Electronic Lead Screw controller using a Teensy 4.1

[WobblyHand]

Got a little tired of the 0.001" resolution of the magnetic DRO's and ordered some 1um glass scales from AliExpress in mid January, pretty much just in time for Chinese New Years, when the whole country stops working. Got word today that the scales have left China and are passing through Narita, Japan. So they are on the way! My ELS can accommodate scales of 0.001"/div, 5um/div, and 1um/div. Doesn't care if you mix them up either. Making this ELS has been a great project involving electronics, software and machining, it was quite enjoyable.

In other, related news, I have created a private gitlab repository, and am working on getting it ready for testing. Did a successful burn of the code using the hosted code, so I am happy about that.

 

[WobblyHand]

Woohoo, the scales are at US Customs in Louisville, KY! Won't be too long now.

 

[WobblyHand]

Yippee! The scales came in. After one minor misread of the data sheet, I connected my GND to 0V on the diagram. Bingo, both scales worked. That's a relief. Now, I have to decide how to truncate the display. I can display to 0.001mm, but need to handle dealing with inches and truncation or rounding. 1um is 0.3937 tenths. Found some bugs in the serial output, it still thinks it is in inches, but the display is showing the units correctly as mm.

 

[WobblyHand]

Spent some quality time with the short scale and my lathe. Been scratching my head how to install the glass scale on the X axis. I don't think it is possible with this normal sized scale. The overall height is the primary issue. I measure 47mm from the bottom of the scale to the top of the read head. However the top of the cross slide to the prism on the ways is about the same thing. I could space it up 1mm, but then the scale goes above the cross slide, which I do not want. If I turn it sideways, then I lose tail stock quill travel (in effect) by requiring a longer stick out of the quill. The "slim" version is about 43.5mm high, which might work out. Also with the slim version the read head is 70mm long, rather than 80mm, which helps.

To ship this scale back to China is probably not worth it. I will eat the cost and just order a slim scale. Maybe I can use the short scale for some project or another. Perhaps in a fixture of some sort. I have code that will do a DRO display.

The larger scale is fine. I just need to install it, which may require me moving the lathe.

The measured travel of the cross slide is 175mm, might be slightly longer, but I don't want the lead screw to fall out of the nut. So the body extrusion should be 175mm + the read head length, or 175 + 70 or 245mm. I think the standard is to order by travel length? 175mm? It seems the place I bought from has a 170mm or a 220mm. Probably go with the 170mm, according to the datasheet it would be 282mm long. That's slightly shorter than the 150mm normal scale I have now. (289mm measured, 292mm on data sheet.) Or I could cut down a 220mm scale on my belt grinder.

 

[WobblyHand]

Ordered a slim scale 220mm long, will cut it down to size. I want those extra 5mm of travel. As the clock goes ticking... Hopefully it won't take so long this time. There's practically no point in making any brackets, since the actual object is doesn't match the drawings. So I will wait on it. If there was only some way to machine the head a little smaller, I'd be able to use the one I have...

 

[ivel03]

The Sino KA-200 slim scales are the smallest glass scales i've seen - 16mm x 16mm. The only thing to becareful about is that they come in 1um and 5um and of course i ordered the 5um by mistake!

Subscribed and watching to see how the new scales fit your machine.

 

[WobblyHand]

The Sino KA-200 slim scales are the smallest glass scales i've seen - 16mm x 16mm. The only thing to becareful about is that they come in 1um and 5um and of course i ordered the 5um by mistake!

Subscribed and watching to see how the new scales fit your machine.

I wasn't clever enough to get that one. I got the one that is one size up, but should still fit. I was primarily limited by the read head length on my lathe. I found a 1um version that has a 70mm long head, rather than 80mm. This should prevent it from scraping my lathe ways prism for the tail stock. It has a few mm less height, which also helps. It is the KA-500.

 

[WobblyHand]

The Sino KA-500 scale arrived yesterday. Less than two weeks! Rather speedy for AliExpress, usually takes a bit longer. Using my spare ELS board, it seems to indicate a little bit, even when the head is locked in place. A tiny bit of pressure on the head moves the reading around. Need to unlock the head to do a do an actual test. But it is a good sign. The read head is 70mm long - as the drawing indicated and a critical dimension for me.

The scale was strangely packed. The outer plastic bag was captured by the housing. It wouldn't come out of the bag. I had to slice the bag in a couple of places and had to disassemble everything, to get the bag released. It did have the advantage that there were no loose parts jangling about in the bag, so I guess it ended up ok. There was no damage anywhere, so I suppose I can't complain.

It came with some pieces that might be useful for mounting, at least as a starting idea. Came with a few shims for alignment. Much nicer than using random washers.

Next steps are to do a sanity check on the cross slide. After that, measuring the true length that I need and marking the scale. Then figuring out how to cut the scale down without mangling it. This scale is different from other ones that I have, in how the end cap is mounted. It will make the cutting more difficult, as the housing has to maintain a flat section for the end cap to mount to.

 

[WobblyHand]

Boy, that end cap was tough to get off. It was glued on using silicone rubber. Used a razor blade and eventually cut enough of it and was able to remove the end cap. The glass is mounted at an angle. Hope I can shorten this unit without the glass coming free of its glue. I will have to mill off an end, leaving an 11mm flat section for the cap. Put the end cap back, as well as the head locks, so I don't mess anything up while I'm taking measurements to figure out how long it really needs to be. I might mill the aluminum away and then score the glass and snap it to the score. Got to think about that a little. It doesn't matter if there's a ragged edge, only that any cracks not propagate more than 6mm.

The scale seems to indicate - at least I moved it 10mm and it read 10mm, and moved it one inch and it showed 1" movement. Hard to tell over a remote slow connection if there are any missed counts. I don't think so, but I will have to test with a direct connection. 10mm/sec with 0.001mm scales should be 10K x 4 counts per second, which should be fine. Scales are rated to 60mm/sec or 240 KHz count rate. I'm not feeding that fast!

 

[WobblyHand]

After a lot of thinking and just plain procrastinating, I was able to draw up the first piece that I need to mount the x scale. Modeled it in FreeCAD, along with the top cross slide piece. I am 3d printing the model to see if I got it right. Printed in draft mode, it will take about 1 hour and 10 minutes, which is fine. There was a lot of construction from reversed or mirrored views, and I'm not sure if I did it right. None of this is hard to mechanical designers, but to me, it's a mind scramble!

The large pockets (on the back side) are to go over the gib set screws and nuts. The two counterbored holes on the top mount the bracket to the cross slide, and the two holes on the ends mount the scale to the bracket. To fit this on my 3d printer, I had to print it diagonally. I'm trying out some black PETG for this. Hope to use this piece to help me figure out the back splash spacing.