This will likely be a long post but I'll try to keep the information to the relevant. I have a 100-watt Chinese laser with a 600x1000 bed. Motion was handled by two tiny little NEMA17 motors, with the Y axis being driven by a single motor turning a shaft that ran the length of the long axis to turn two timing belts. I was having problems keeping my mirrors aligned, and suspecting the motion system on the machine to be the culpriit (it was a strange mix of good profiled linear rail and plastic-pulley-in-a-groove-with-slop), I removed it all and built myself an Acro system inside the cabinet. I lost a little but of room with the motor width but I had room to spare. I fired it up and tested it out and it ran MUCH more smoothly and quietly, speed was better also. The problem was distance. I expected to have to reset the steps on it since I had changed a lot of stuff, but that wasn't the problem. Even if your step settings are off so bad that your machine moves 200mm when you tell it to move 100mm, when you tell it to immediately go BACK 100mm it should return to the same spot. The error in step values is constant. My machine did not return to the same spot. Sometimes it was more, sometimes less, within a range of 2-8mm. I doubt very much that the belts were slipping, as I would be able to hear the teeth jumping. It ran VERY quietly, for all the world like it was working just fine. It just didn't seem to know how to far it wanted to go when I told it to move. SO, since I had replaced two steppers that were set at .46 amps on the drivers with three steppers that could handle 2.1 amps each, I suspected a weakness in the factory power delivery system. It wasn't designed to handle these motors or this many of them. I removed the factory 24-volt power supply and all three drivers (the Z axis has a NEMA23 motor with its own larger driver) and replaced them with a Meanwell 48-volt 12.5-amp power supply and I replaced the anemic drivers with KL-5056 drivers. After getting all of this wired up I tested the motion as before, moving 100mm right (+X) and then 100mm left (-X). It STILL didn't come to the same place. After adjusting some of the settings like steps, acceleration, etc I tried it again. Now, even though the machine homes perfectly when it's turned on, any time I move the X axis with the keypad on the machine it moved a couple of mm and stops, and the red light on the driver comes on solid. From what I can find, this is an over voltage alarm. It happens when it sees 52 volts or more. How is it possible for the driver to see more voltage than the power supply gives?? I tried turning the voltage output down on the power supply to its minimum of 40 volts. No change, still instant Over Voltage alarm. Sooooo, now I have two issues - X axis will not move without triggering an alarm, and even if it DID, the machine STILL doesn't travel correctly. Y axis seems to behave normally, though I haven't gotten around to measuring movement. I tried swapping the X and Y drivers also, trying to isolate a faulty driver. No change - Y axis works fine on the driver that was alarming on the X axis, and the X axis still locks up with alarms with the second driver. I REALLY hope someone can help me out here or point me in the right direction. It's frustrating as hell having a piece of equipment in your shop that cost as much as this one did and have it capable of nothing but being in the way. HUGE thanks to any help anyone can give me.
There's no reason a laser shouldn't be able to run at very high speeds and accelerations with a four-motor 24V NEMA 17 belt-drive system. I don't suspect that was the culprit on the mirrors from the start. Replacing profile rail with extrusion and delrin wheels seems questionable too. However, that aside... Not sure why the driver would give an overvoltage alarm at 40V. What happens if you feed it with the 24V PSU? It may also error similarly- hard to find much documentation- if it detects a short in the motor lines, did you check those too? Variation in distance travelled typically tends to be caused by overly tight V-wheels, slippage in the toothed pulleys/motor shafts/grub screws, etc. Check the entire mechanical chain. Personally I'd put the profile rail back on the V-Slot and dump the wheels, since you already have it in roughly the right sizes, but that's just me.
Thanks for this. The belts aren't slipping, I assume I'd be able to hear the teeth jump as quiet as it is. The screws are all tight but I don't honestly know how to tell how tight the belts should be.
I don't usually assume the belts do slip- they have to be very loose, or under very high dynamic forces, for that to happen, and when it does it's quite loud. Slippage of the pulley itself is what I'm talking about; the grub screws frequently back out undetected without some blue Loctite on there. Belts should be tight enough not to pull up more than an inch per couple feet of axis length. That's an approximation, I'm not near a machine to check, but it's pretty **** tight.
I tighten my belts on my router with a spring balance, the type used by fishermen. I pull to 12kg. I first tried 6 but that proved to be too little.
This is excellent information, the first I've found with actual number associated with belt tension - thank you!
my goal was not so much 'a number' but rather 'the same number on all of them' since I am a little OCD like that. I had the thought that particularly the 2 Y belts must be the same tension so that they stretch the same amount. but knowing the number also means you know what to do when it appears they are too loose (-: I don't recommend doing it alone though,, you need someone to hold the machine down while you pull and tighten the belt fastener, doing it alone is possible but quite difficult and will involve a lot of retries. I suspect that longer belt runs will need more tension.
