Page 1 of 4

Integrated stepper motor

Posted: Tue Dec 11, 2018 10:59 pm
by tdarlic
Hi guys
I plan to replace the NEMA17 steppers with these integrated steppers:
http://americanmotiontech.com/products/ ... pper-motor

I think I could get at least 30% faster movements with this. These steppers should directly replace ones from the kit and they would be driven directly from the tinyG.
Has anyone tried this approach before?
Any advice?

Re: Integrated stepper motor

Posted: Wed Dec 12, 2018 11:09 am
by tdarlic
Actually, the motor I plan to use is
http://www.americanmotiontech.com/produ ... ervo-motor

Re: Integrated stepper motor

Posted: Wed Dec 12, 2018 11:43 am
by JuKu
Interesting, I'd like to know about your results! The motor you linked has about double the torque than the supplied motors. Also, being closed loop it doesn't stall so you get to actually use all of it. TinyG controller (or any other stepper motor controller for that matter) doesn't have any means to optimize the torque, so the current ans speed need to be set so, that the motor doesn't stall in the worst case situation. I doubt that this results to the motor operating at max torque.

Re: Integrated stepper motor

Posted: Wed Dec 12, 2018 12:44 pm
by tdarlic
I'll buy one NEMA17 closed loop servostepper and then test it on a Y axis. Only problem is that I need to pull the motor control signals directly off the tinyG board as in this link
https://github.com/synthetos/TinyG/wiki ... al-Drivers
This is signal directly from the MCU so some isolation and/or level shifting would be prudent. Most probably with something like this:
https://www.ebay.com/itm/Optocoupler-Is ... fresh=true

I need to get this signal reliably to the motor and then adjust the motor for 3200 microsteps per revolution (I believe that this is what Litepnp is using now). I'll do this for Y axis and if this works reliably I might get one for X axis....

Re: Integrated stepper motor

Posted: Wed Dec 12, 2018 9:56 pm
by wormball
tdarlic wrote:[$100 motor]
I think I could get at least 30% faster movements with this.
Any advice?
Use 1.8 degree motors, luke! I use $10 42BYGHW609 and got stable operation at 45 (X) / 36 (Y) m/min. Which is about 90% faster than in the last Juha's video. And their 1/8 microstep is 25 micrometers which is way less than the total mechanical inaccuracy. I really can not understand why Juha is using and supplying 0.9 degree motors.

Also i will try 30-toothed pulleys soon, with which i hope i will get at least 60 m/min by X axis.

Re: Integrated stepper motor

Posted: Thu Dec 13, 2018 3:56 am
by Jet
I’ve noticed the bearings on the X/Y axis start whining way before the steppers start maxing out, which generally indicates the bearings not keeping up which will lead to increased wear. Also it’s a lot of mass to move around at higher speed if something goes wrong (like hitting nozzle holder etc.), and there’s a lot of jerk to the machine at higher speeds effecting placement/small parts on the table.

I’ve tried it at a variety of tensions for the idler wheels, but it seems to be the limit.

My speed is set to 10,000mm/min (around 12,000mm/min, bearing skip started to show up) both for X/Y.
Acceleration set to 1000.

Plus there’s a limit with that amount of mass on the head and the axis length for how fast the mass can be accelerated and decelerated over the length before there’s issues.

So “personally” the stock steppers seem spec’d adequately for the machine design. Possibly better v-wheel idler bearings may help can a bit.

Re: Integrated stepper motor

Posted: Thu Dec 13, 2018 10:41 am
by wormball
Jet wrote:I’ve noticed the bearings on the X/Y axis start whining way before the steppers start maxing out
around 12,000mm/min, bearing skip started to show up
It's not bearings, it's stepper motor resonance. If you try speeds more than 14 m/min, you will see that this sound will decrease with speed until the motor stalls. And with 1.8 degree motors this resonance occurs at about 24 - 28 m/min. I think the best way to deal with it is just to ignore it.

Also the larger is your acceleration, the shorter time is spent at the resonance speeds, so the lower amplitude is reached. However, if the carriage moves at some angles, the speed by one axis may be close to the resonant speed.

I placed 1 mm rubber sheet between the motors and the motor bases and placed the Y motor on an old PCB bolted to the table, it seems to slightly decrease the sound, but i had not measured it. I will also try damping coupler like this: https://ru.aliexpress.com/item/D20L25-s ... Title=true .
Jet wrote:My speed is set to 10,000mm/min (around 12,000mm/min, bearing skip started to show up) both for X/Y.
Acceleration set to 1000.
It's very slow. My jerk is 9000 (X) / 7000 (Y). Have you tried adjusting the current? I found that the optimum values are slightly less than half by X and slightly more than half by Y.
Jet wrote:So “personally” the stock steppers seem spec’d adequately for the machine design.
I do not think your liteplacer is very different from mine or Juha's or Mark's. So i "personally" advise you to try higher speeds and 1.8 degree motors.

Re: Integrated stepper motor

Posted: Thu Dec 13, 2018 6:41 pm
by Jet
The sound, sounded like it was coming from the bearings not the stepper motor.
Plus it’s around 2000 steps per minute, axis is loaded and it’s micro stepping, so thought it was unlikely it was stepper
resonance when I was testing pushing the limits.

That said, hearing isn’t great these days so I may not have localized the sound correctly, and / or it maybe another form
of resonance, so if other’s are getting higher speeds, I’ll revisit and check everything again (I’d like it faster, just it’s a lot of mass to fling around :-) )

And I may have jumped to that conclusion based on encountering bearing skip on the Makerbots whilst developing the Sailfish Firmware acceleration code when I was testing the limits of the machine.

I’ll post my findings.

Re: Integrated stepper motor

Posted: Fri Dec 14, 2018 10:21 am
by tdarlic
Cheers. Thanks for this input. All I need is to get reliable 700 0603 pcs/hr. I've ordered the integrated servo/stepper from AMT and plan to install it in few weeks. Also, I have ordered 1.8 deg stepper from AMT with a little bit more torque.
I just started to play with machine and the motions I get are reliable and repeatable, at least when seen trough the camera. It seems to me that the camera/nozzle distance is always off by the same distance regardless of what I try. I am sure that I am calibrating the camera wrong or that's because the parts and the pcb are on different heights or something there is something else I am doing wrong.

Re: Integrated stepper motor

Posted: Tue Dec 18, 2018 7:30 am
by Jet
wormball wrote:If you try speeds more than 14 m/min, you will see that this sound will decrease with speed until the motor stalls.
So I revisited this today.

Any speed >=12000mm/min started whining. At 18000, the steppers gave up. Tried at various acceleration speeds (250, 500, 1000) and tried reducing step count to 4. Also tried increasing the current from just under half where I have it now to 100%.

The only difference was below 100% current it would give up at 16000, and at 100% it would give up at 18000.

Whine didn’t reduce past 14000, it just got louder.

When the stepper was skipping, the noise was coming from the stepper, but when not it sounded like the whining was coming from the bearings.... although I can’t be 100% sure because the whole axis acts as a sound board too, but it seems to be localized where the head is.

I do have some gecko drivers, so I’ll try those at some point (they don’t suffer from stepper resonance, but those also microstep at 10).

Is anybody managing >10000mm/min without whining?