Help 'Splain Sumfin to Me: Steppers, Drivers and Parallel Ports

R

Ripthorn

Active User
Registered
Joined
Dec 8, 2013
Messages
147
Yesterday I was using my CNC on my first real machining project and found myself wondering how I could help my machine perform better. I have a MaxNC15 (I know, I know, but it was free, so I'm using it) and right now my rapids are limited to something like 14 ipm. I am using a mechatronics 4 axis driver (with socketed driver chips) and the motors have been upgraded to Keling 280 oz/in dual shaft steppers. I have tried to increase the rapids, but whenever I get up any higher than about 12 or 14 ipm, I hear a whining and the motion stops, like it's getting locked up trying to provide the pulses. I am running LinuxCNC on an older Dell machine. I am wondering what I might be able to do to increase the rapids. I hear of people with other small machines like Taigs and Sherlines getting much higher rapids. I think that there are four possible causes:

1. The computer: it's old and may not be able to send the pulses fast enough
2. The driver board: I suppose it's possible it doesn't provide enough power or something to the motor. I have heard that socketing the driver chips is kind of a no-no, and I have burned up several chips (4 in the last year or so) that I have replaced.
3. The steppers: I think their torque rating is sufficient for the machine, but maybe not
4. The machine itself: friction could be an issue. It runs on 1/4-20 threaded rod and it seems I can never get the gibs adjusted tight enough to eliminate racking while still sliding smoothly. Of course, I have only been using 3-in-1 oil on the ways...

So what might be the most likely suspects, and what are the things I could do cheaply to try to get it going better. I would love a new driver board, but I can't plunk down the cash on the gecko g540 for what amounts to only a little more than an experiment with CNC. If all goes well, I will get a bigger machine. Mostly I am just cutting aluminum and plastic right now, but would love to be doing brass and steel (but found that this machine is nowhere near rigid enough for steel).

Sorry for the long post. Any help is greatly appreciated.
 
It sounds like the problem is the motor torque (along with the resistance of the mechanics). 280 oz/in would seem to plenty, but that is holding torque value. What you are describing is the motor decoupling (the whine). About the only fix for this is to increase the voltage to the motor, and this probably means new drivers and power supply. Those motors are rated at 30 Volts, using the Keling driver, the limiting factor is the driver. If you look at the torque specification, it shows that the torque drops rapidly as the speed increases. http://www.kelinginc.net/KL23H276-30-8BT.pdf

If the lead screws are 1/4-20, standard V-thread, then that requires about 10 times the torque of ball screw at the same load.

Let's look at the motor speed vs movement.
The motor is a 1.8*, 200 full steps / revolution
20 TPI = 0.050 / revolution
So running in full step mode:
15 IPM = 15/0.050 = 300 RPM x 200 Steps = (60000 Steps / minute) / 60 seconds = 1000 pulses / second

Assuming the driver is running in half step mode, which would be common, then the to get 15 IPM it would require 2000 PPS. At this speed, the torque drops to about 80% of the rated holding torque. This assumes that the motor is being driven at it's rated voltage. Running it at a lower voltage, the torque would drop off even more dramatically. I suspect that your system is running at 24 Volts.

If it's possible to attach heat sinks to the driver chips, it would probably help the service life.

I don't think it's the computer, it sounds like the pulses are being output to the drivers and I would assume that this is well within the limits of the computer.

If it is possible to re-calibrate the steps / inch on your machine, you might try adjusting the step mode to 1 in the driver if that is possible. Reducing the friction would be the best way to improve the performance, but you are limited by the hardware. Buying new drivers rated for a 48 Volt power supply and a new power supply would give you a dramatic increase in performance.
 
Jim, thanks for the great reply. I have been reading up on some of this stuff and I do believe that I have a 24V power supply. I think my driver board is rated up to 28V or something like that. I will start looking in to other drivers. Any recommendations that won't break the bank, or is that a fool's errand?
 
I would not want to recommend any driver / power supply combination that I haven't had personal experience with. The ones that I would recommend might break the bank. ($89 to $215 each for the drivers and $140 to $178 for the power supply) There are a lot of drivers advertised on Ebay and Amazon, but I know nothing about the quality or functionality.
 
I've been reading some of the reviews on those units, and it sounds like the generic ones aren't worth the pain. I'll probably save up a little more for a good unit, as I intend to move more towards CNC, and a good driver and power supply could be used with more than one machine or another machine if need be.
 
my Sherline runs with an older tb6560 board running 24v and 170oz motors. but at best I get 25ipm
I know nothing about Linux but I doubt the problem is the pc.
over tightening the gibbs or the backlash in the screws leads to problems, once the machine warms up they can lock up. It took me a long time to find that problem because you might be into a project for an hour and suddenly 1 screw stops making all the steps and the part is destroyed. loosen everything and find max ipm then set to a speed below that as max. if your gibb adjustment or backlash adjustment keep you from operating at that speed back off a little at a time. you have to have a small amount of backlash for these small motor to work the heat up and expansion will lock up the screw.
heat sinks help and a good cooling fan are a must, most of the boards don't have enough cooling if placed in an enclosure so you need an enclosure fan.
steve
 
You must log in or register to reply here.
Back
Top