[CS900]

looks like I'm going to have too, although they have yet to get back to the email I sent them a while back

[CS900]

sooooo....May be onto the root cause here. After multiple un-returned emails and phone calls from automation technologies, I decided to play around with it a bit more. So since the steppers work great without a load on them I decided to try loosening the belt drive a little at a time until there was next to no tension on the belts with the though that at some point the steppers would work correctly even if it slipped on the timing belt. So wouldn't you know it, the more I loosen the belt, the better they work. I'm now thinking the belt is side loading the encoder a bit, and as i decrease tension the encoder comes back into alignment. This may not be such an issue with the larger 34 steppers, which is why i didn't notice it.

So I guess next step is to make some brackets to direct drive the leadscrews....Thoughts?

[UA_Iron]

sooooo....May be onto the root cause here. After multiple un-returned emails and phone calls from automation technologies, I decided to play around with it a bit more. So since the steppers work great without a load on them I decided to try loosening the belt drive a little at a time until there was next to no tension on the belts with the though that at some point the steppers would work correctly even if it slipped on the timing belt. So wouldn't you know it, the more I loosen the belt, the better they work. I'm now thinking the belt is side loading the encoder a bit, and as i decrease tension the encoder comes back into alignment. This may not be such an issue with the larger 34 steppers, which is why i didn't notice it.

So I guess next step is to make some brackets to direct drive the leadscrews....Thoughts?

So your setup had a pulley driving a belt; the pulley and belt tension would put the encoder out of alignment in a sense and ultimately causing the unstable stalling out behaviour?

Direct drive with a flex coupling is definitely superior to a belt drive (in most cases) if you can get away with it.

[CS900]

That's my theory.Here's the belt drive
Attachment 259138

Same type of set-up on my mill however. But again, much larger steppers.
Attachment 259140

[CS900]

Well that didn't fix it. I made new brackets, ordered zero backlash couplings, and installed everything tonight. Its better than before, I can jog the motors now, but i still get an intermittent positioning error. I'm also getting a LOT of dithering, which is specifically why i bought these so I didn't have to tune them. I'm very disappointed in these steppers, and even more disappointed with automation technologies support. I may sell these and just buy some geckos. At least I know they will work.

[CS900]

for anyone who can think of anything else, took a video for John @ automation technology. Feel free to comment on any more ideas:

trim 5DCCC373 6601 4F2C 8506 871DA311F8F0 - YouTube

[CS900]

so got a snow day today (seriously) and decided to work on this a bit more. Got it to the point where it doesn't fault out nearly as often but it still hunts a bit. I've got it running mostly reliable at 100IPM at 10in/sec^2. I ended up loosening my ballnuts a bit in order to get there, so I got a bit more backlash, but it's much more reliable.

[CS900]

spoke too soon....these things suck...

[Larken]

Whats surprising is how terrible chinese support is. Almost every problem people post of Leadshine or other chinese drives never gets an answer.


I have a new product that will Close the loop on any stepper system. It will be ready in january.. It will work with Gecko drives (or any drive), and an encoder on the stepper.

Step Lock, ViperServo

It monitors the Step/Dir inputs and the encoder.
This way you have the performance of a gecko and the security of a closed loop system. You can set the error limit and it will stop Mach3 if it exceeds that.

I will post some pictures on putting encoders on Keling size 23 motors. I've been using the KL23H2100-35-4B motors and mounting a 500 line western digital encoder on them.

Larry

[ger21]

Larry, you do know that this already exists and has been available for several years..

Page Title

[Larken]

Yes, i've seen it. One of my customers talked about running that on their viper 200 mill conversion so they could have a dro.

Do you have it ?

Larry

[ger21]

Nope. Never really understood the point.

[Larken]

Ya, i've seen posts where people say a closed loop stepper is a solution looking for a problem . But so are seat belts, and airbags. Why do trucks have bumpers ?

Sure if you have a new machine with steppers, or a hobbiest who lightly uses a machine then you may not think you need closed loop, but for anyone running a machine 8 hours a day and cutting expensive material, 1 jam up can waste more than the cost of the closed loop system.

After a stepper cnc router has been running in dust for years, they start to jam and mess up parts. ( i know, ive sold a lot of routers in the last 24 years) All you need is dirt in a linear bearing to make a machine unreliable and jam with you don't expect it.

A closed loop stepper is piece of mind that you won't destroy that $600 4x8 sheet of aluminum or expensive signboard.

Larry

[UA_Iron]

Here's one of the best implementations of a closed loop stepper driver:
Copley Controls - Accelnet - Servo Drive - Servo Amplifier for Brushless & Brush Motors - CANopen - CAN bus - EtherCAT

A few neat features:
1. Close a stepper velocity loop with a rotary encoder and close a position loop with a linear encoder at the same time.
2. Fully adjustable gains - will do auto tuning of your stepper as well.
3. Current shaping can allow you to squeeze more performance out of any stepper.
4. Considerably lower power consumption.
5. You can set one input count to any number of position counts - meaning your bandwidth limited db25 breakout board can run at a relatively lower bandwidth with the benefits of running at a higher bandwidth.
6. You can set number of encoder counts as limits - getting rid of limit switches.
7. High speed inputs operate at something ridiculously high - like 5mhz.
5000 line encoders work well with these things - post quadrature you'll get 20,000 ppr. It's a true servo controller, it doesn't care that it's pulsing to drive a stepper motor.

Pushing off I/O capability to the individual motor controllers is more robust way to handle logic components of a machine - in my opinion.

Of course, learning curve is high with these controllers. Programming a lot of the logic is a big deterrent.

[CS900]

so little update....Finallt got John at Automation technologies to help me out. We got leadshine (the OEM) to do some investigation as well, and no one could figure out what's going on. I put my 282oz/in steppers with a G540 from my minimill on the lathe and they work like a charm. John is gracoulsy letting me return them so I can buy a standard stepper set-up. I as really hoping to get these working becasue I love my Nema34s that I have on my large mill, but at the same time I'd like to get this thing making chips at some point. So for now I'll just get some Gecko drives and some openloop steppers and call it a day.

[Larken]

Try to get steppers with a Dual shaft, so you can add an encoder. I make a little board called a Step-Lok that can close the loop with any stepper driver.
Try to get a size 23 motor since they can use a cheaper encoder than a size 34.
Here are a few pictures of one of my machines with USdigital encoders on KL23H2100 motors and my controller with my CobraSV closed loop stepper drives. The blue cable is cat 5 (flexible) for the encoder

The Step-Lok will work with Gecko, Leadshine or any step/dir drive.

Larry

[CS900]

As luck would have it, I already ordered new steppers in a nema23 with dual shaft.. So how exaclty does your drive work? will it actually generate steps if it sences missing encoder counts? Or will it just give a position fault that you can used to stop the machine motion?

[Larken]

As luck would have it, I already ordered new steppers in a nema23 with dual shaft.. So how exaclty does your drive work? will it actually generate steps if it sences missing encoder counts? Or will it just give a position fault that you can used to stop the machine motion?

The Step-Lok doesn't control the drive, it just is a watchdog that keeps track of both encoder and step positions and outputs a fault if the error exceeds a programmed value. But since the gecko is a high performance stepper drive that gets the most out of the motor anyway, adding foolproof security of never losing position or messing up a part , is all you can ask, isn't it ?

My cobra-sv drive does correct to a small degree, which helps get rid of resonance . I've sold a few machines with it and its reliable, but it has a bit of noise when holding, so i need to redo the pcb layout a bit.

Just like inevitable as death and taxes, an open loop stepper motor system WILL jam up sometime and on a router or mill, it can plunge the spindle into the table, ruin the workpiece or on a lathe, ram the cutter into the chuck.

Larry