TSL change gears, or the lack thereof

I admire your industriousness, kb58. I think you will learn a lot about control systems in this project.
 
Thanks. I used to be in the field but it's been a long time. I think what shocks me the most is how cheap the development hw and sw has gotten. Slightly disappointing though, that even though it's a TI board, it's still made in China. Oh well.
 
The display board, stepper, driver, and power supplies are on hand. The question is: wait and see if the developer sells his interface boards, or assemble it the way he did with the prototype, using Ebay bits and bobs (as the Brits say). The parts are certainly affordable; for example, the logic level converter boards and power supply are about $2 each. Because of the very low price, the boards were ordered through Ebay, and if/when the developer starts selling his boards, it'll plug right in. In either case, a kit of connector parts for the 0.025" pins was ordered, so it's covered either way.

The developer used cogged belts and pulleys to drive the QCGB in his G0602 lathe. In Part 11 of his video series, he shows how another hobbyist instead used the existing gears instead of belts. I haven't yet checked to see whether that can work on the TSL-800, as gear diameter (and ratios) will dictate whether that's workable.

On a related note; I have no idea how much torque will be required to drive this whole affair. It's not a straightforward issue because of all the variables, such as: feed rate, cutter speed, cutter sharpness, gearbox ratio, pulley ratio, lead screw ratio, machine rigidity, and probably half a dozen other variables. The developer's lathe is pretty small, and he used a ~230 in-oz NEMA 23 stepper. Suspicious that this might be too small (based upon nothing other than me having a more stout lathe) I went with the next model up with ~440 in-oz. Will that be enough? Who knows!
 
Last edited:
Brief update. Because the prices of the various boards were so cheap, I went ahead and bought them anyway, figuring if and when he produces his own board, I have that option. Anyway, connected everything up today, turning it on, and son of a bi..., it came up. At this point it's just the controller board and display, but it's nice to see it doing something. I think while I wait for him to start producing boards is time best spent figuring out and mounting the servo and encoder. I'd "like" to not have anything visible, but unless I place the servo motor down near the lathe motor, its' too large to hide. We'll see.
 
Decided against placing the servo down near the motor due to all the stuff that could drop into the gear and belt. The placement below is safer but requires a hole through the triangular filler panel, oh well. The pulleys and belts are from SDP-SI, and the pulleys were chosen with hubs large enough to bore out to fit the 22mm shafts on both the QCGB input and headstock output above it. Still to be installed is the encoder; it'll live above the servo. Haven't decided whether to mount the servo and encoder off the left end of the QCGB, or to the triangular filler panel.

IMG_20191008_182834.jpg

IMG_20191008_182842.jpg
 
Last edited:
After making a whole lot of chips, both the servo and encoder are mounted. In hindsight I probably should have just welded or even bolted some aluminum plate together, but I was looking for an excuse to use the mill.

For TSL owners, the intermediate shaft just below the left end of the spindle spins at the same speed as the spindle. I marked both, ran the lathe for a while, and when stopped, the marks were still aligned. That's good because it means that 1:1 gears on that shaft and encoder will accurately reflect the spindle speed, without having to fuss with different encoder ratios.

Next will be wiring it up temporarily to see if the encoder encodes and the servo... servos.
 

Attachments

  • IMG_20191013_180114.jpg
    IMG_20191013_180114.jpg
    727.3 KB · Views: 44
  • IMG_20191013_180135.jpg
    IMG_20191013_180135.jpg
    590.7 KB · Views: 43
  • IMG_20191013_180127.jpg
    IMG_20191013_180127.jpg
    718.2 KB · Views: 45
  • IMG_20191013_184119.jpg
    IMG_20191013_184119.jpg
    606.6 KB · Views: 52
Last edited:
Got the whole thing running. Cobbled together the power supplies (36V, 5V, and 5V-to-3.3V), the logic level converter, controller board, display board, and servo driver. One thing that was unclear was whether the controller feeds active high or active low signals to the servo, turns out it's active high. Once connected properly, the servo started making a faint sound. Started the lathe and sure enough, the rpm was reading values and the servo was turning!

The whole setup was thrown together just to see if it works. It does, but it's such a dicey setup I don't dare cut anything with the lathe for fear of chips flying into the mix, or a wire popping off and causing it to do strange things, or making something blow up. Next step is making enclosures for everything and cleaning up the wiring.
 

Attachments

  • Untitled.jpg
    Untitled.jpg
    147.4 KB · Views: 44
Last edited:
Who's afraid of chips or tools crashing or any of that negative stuff? When I am at this stage, I am eager on trying this setup out on something real. Note that real doesn't mean junk steel. It means a junk piece of PVC pipe found in the scrap bin or street with a hastily cut thread run out area (generously sized, for maximum peacefulness). PVC chips should be safe even if they fall in circuitry or gearing.

Good job!
 
Very interesting! I'm anxious to see how this works for you. I have all the change gears for my TSL-800, and have had it for about 6 months I suppose, and have not yet had to change gears. Really it looks like the only real need would be if I want to cut 13TPI for some 1/2-13 threads, as the standard QCGB has plenty of good choices.
Also the minimum standard feeds are a bit fast in advance per revolution, but the lathe seems to have plenty of power and the surface finish is good.

So, I'm standing by.
 
Honestly, it's mostly an excuse to build something cool. I like building embedded controllers, and having one that moves a servo is even better. (Back when I was in college, there was no FSAE or robot battles, so I'm making up for lost activities... already have the car!).
 
Back
Top