Closed-loop stepper controller progress

Ah, OK, I kind of understand then why the encoder was used with a DC motor, tks.

Quote
LoboCNC

Most of my print failures are - suddenly curling-> head strike-> lost steps
Does your solution solve that problem?

Yes, closed-loop control of the motor will allow it to move past minor nozzle collisions with curled edges. The motor may get pushed out of position a little, but it will spring back into position after it has moved past the obstruction, and the rest of the print will not be affected.

With a really major obstruction, though, the closed-loop servo will eventually give up. I usually set the maximum allowable position error to about 1/2 rev. of the motor so that the motor really has to be blocked completely before the servo gives up.

Quote
LoboCNC
Yes, closed-loop control of the motor will allow it to move past minor nozzle collisions with curled edges. The motor may get pushed out of position a little, but it will spring back into position after it has moved past the obstruction, and the rest of the print will not be affected.

With a really major obstruction, though, the closed-loop servo will eventually give up. I usually set the maximum allowable position error to about 1/2 rev. of the motor so that the motor really has to be blocked completely before the servo gives up.

1/2 rev is 100full steps(1.8 degree). So you are able to correct up to 100 lost steps failure??? Most of my falures are 1-2 lost full steps.

Edited 1 time(s). Last edit at 12/12/2013 05:06PM by karabas.

Yes, the closed-loop servo controller can correct for pretty much any size of position error. With an open loop stepper, if the motor gets out of position by more than 1 full step, the motor is actually locks onto the adjacent detent position which is 4 full steps off from the desired position.

With the closed-loop servo controller, the motor is constantly pushed back to the correct goal position no matter how far away it gets.

If the ultimate point is to provide closed-loop positioning, and this solution is roughly $390 more expensive*, then would that money be better spent on some kind of machine vision? Then the feedback mechanism could be stuck onto any computer controlled positioning system and could probably be upgraded to provide quality control of the part itself.

*
$125 for motor and encoder
$500 for 4 axes
vs
$8 for the driver [www.pololu.com]
$20 for the motor [www.pololu.com]
$110 for 4 axes

In the original post, I think I mentioned that the incremental cost of the closed-loop control over an open-loop system would be more like $60 per axis, or $240 more for a whole machine. But it might make sense to just use closed-loop control on the extruder, or maybe just on the X & Y axes.

Some kind of machine vision would be cool, though, if you could discern a position errors of a few steps.

Machine vision:
It would be very useful to be able to mount a microscope to a delta printer, and have the ability to track the elusive amoeba.
This is one of those things that is on my bucket list.

LoboCNC make it so!
I'd buy one!

Quote
LoboCNC
Yes, the closed-loop servo controller can correct for pretty much any size of position error. With an open loop stepper, if the motor gets out of position by more than 1 full step, the motor is actually locks onto the adjacent detent position which is 4 full steps off from the desired position.

With the closed-loop servo controller, the motor is constantly pushed back to the correct goal position no matter how far away it gets.

How to buy your board? I may ask my friends about making it localy for low cost as diy kit f.e.
I think Ultimaker style printers are most suitable for such upgrade especially direct drive ones as it's easy to mount encoder onto the XY rods.

Or maybe you can sell programmed chip if you decide not to opensource it?

Edited 1 time(s). Last edit at 12/14/2013 11:17AM by karabas.

Right now, I'm trying to decide if I want to make the boards available as a product or not. There is also the possibility of just selling the pre-programmed PIC-SERVO S3 controller chip. (Unfortunately, I can't open-source the controller chip firmware because it is based on proprietary code not owned by me.) Hopefully, I'll have my plans sorted out early next year and will make an announcement if the board or controller chip become available.

Quote
LoboCNC
Right now, I'm trying to decide if I want to make the boards available as a product or not. There is also the possibility of just selling the pre-programmed PIC-SERVO S3 controller chip. (Unfortunately, I can't open-source the controller chip firmware because it is based on proprietary code not owned by me.) Hopefully, I'll have my plans sorted out early next year and will make an announcement if the board or controller chip become available.

Ok, can you post schematics to estimate soldering?

The data sheet I put together (ps3.pdf) has a schematic.

Is it reasonable to use existent driver boards? DRV8825?
A4988 board is available for less then 5$

The driver boards you mention are really nothing more than a driver chip on a carrier board. For closed loop control, you need a micro reading the encoder and then executing a fair amount of code to perform the closed loop control. You also need to be able to control the REF input for the driver chip with a high resolution PWM output. The data sheet describes all the gory details.

Quote
LoboCNC
The driver boards you mention are really nothing more than a driver chip on a carrier board.

Yes I know. I mean the breadboard approach is reasonable here for lower cost when DIY kit. Honestly saying I am a bit sceptical about selling this as a product for low demand.

Quote
karabas

Quote
LoboCNC
The driver boards you mention are really nothing more than a driver chip on a carrier board.

Yes I know. I mean the breadboard approach is reasonable here for lower cost when DIY kit. Honestly saying I am a bit sceptical about selling this as a product for low demand.

Oh, I see. And thanks for your opinion - that's what I'm looking for - to see if this is worth doing as a product.

Quote

Oh, I see. And thanks for your opinion - that's what I'm looking for - to see if this is worth doing as a product.

I'm more interested in this as a method to detect and correct XY skips. My extruder never misses a beat.

it would be useful to make a filament sensor that could see how much filament was actually getting pushed into the extruder.

Seems like there are two camps - those with extruder issues and those with occasional skipped steps. Sounds like additional sensing and feedback control is needed all the way around to make 3D printing completely reliable.

Quote
nophead5 April 2008
This is essentially flat, the slight rise is probably due to the motor windings getting warm, increasing their resistance and thus lowering it's torque. The 160°C reading was taken after the motor had had time to cool down again. This is a good illustration of why a shaft encoder is necessary to control the feed rate.

So if the viscosity is changing, but it has no effect on the motor duty cycle, I have to conclude that most of the torque is required to overcome the friction in the filament guide. That also explains why more torque is required to extrude PCL than is required for HDPE, despite it being less viscous and requiring less spring force. If I rub my fingers over PCL it is obviously a lot less slippery than HDPE.

[2.bp.blogspot.com]

[hydraraptor.blogspot.com]

@A2: That post probably concerns "old style" DC motor extruder with PWM control, so that argument about necessity of shaft encoder is not universally valid.

Quote
LoboCNC
Seems like there are two camps - those with extruder issues and those with occasional skipped steps. Sounds like additional sensing and feedback control is needed all the way around to make 3D printing completely reliable.

So what have you decide about kit/chip selling?

Quote
karabas
So what have you decide about kit/chip selling?

I've been pulled off on another project, but I'm hoping to get back to the closed-loop stepper controller soon. I am leaning towards selling a kit (board + enclosure) and putting together a list of compatible dual shaft motors. Hopefully in the next 2 months.

BTW there are many steppers with already integrated encoder on ebay. Are they suitable?

Quote
karabas
BTW there are many steppers with already integrated encoder on ebay. Are they suitable?

For simplicity, my closed-loop stepper controller needs the number of encoder counts per rev. to be equal to the number of microsteps per rev. (or otherwise some integer multiple). Right now, I'm using 16x microstepping which gives 3200 microsteps per rev. I'm then using an AMT103 encoder set for 800 "line" resolution, which with quadrature counting, gives me 3200 encoder counts per rev.

The other thing is that mechanically, I've got the encoder neatly tucked away underneath the controller on the end of the motor, and the cover I've got also acts as a heat sink. If you were to use a different motor/encoder, you'd have to work out your own mounting & heatsink.

Hopefully I'll get back to this project soon. I've been hip deep in a new printer design - more on that later.

Any news?

Quote
karabas
Any news?

Not yet, although there has been a lot of interest from both 3D printer folks and CNC folks. I've been up to my eye balls in another project, as well as mentoring a US First robotics team. (I helped several kids on my team design and build their own 3D printers over last summer and they printed a bunch of parts for their robot, which was pretty fun. Search Sehome Seamonsters for videos they've posted.)

When the dust settles in a couple of weeks, I need to do a production rev. of the board with some minor mods, get some covers laser cut, update the documentation, and then figure out how I'm going to sell them. So all-in-all, it'll be a couple of months yet before I have something I can let people play with. In the meantime, any feedback on the documentation lobocnc.com/ps3.pdf would be great.

After a bit of a delay, I finally have done a pilot production run of my closed-loop stepper controller that mounts right on the back of the motor. (Actually, it piggy-backs on the AMT103 encoder which mounts on the back shaft of the motor.) It can be used with Step & Direction inputs directly (no stepper driver modules needed), or you can send it commands directly over a serial port, and in fact, you can do both at the same time. That is, you can have it moving under Step & Direction commands while monitoring the motor positions and/or servo position error over the serial port while moving. This is a great way to monitor the extruder forces to check for jams or slippage.

You also, of course, get the advantages of closed-loop control: smoother, quieter, cooler motion with no lost steps.

The controller board includes an aluminum enclosure, and you supply your own dual-shaft stepper and AMT103 encoder (available from Digikey). There is an updated data sheet at: lobocnc.com/ps3.pdf. I've got 20+ of these that I'm selling at my cost of $40 ea. plus about $3 shipping. If you are interested in getting one (please look over the data sheet first), just PM me and I'll send you payment info. If this is well received, I'll be ramping up volume production where I expect the regular price will be about $50.

Can you sell a pair with AMT103? It's a bit expensive to buy internationally separate parts...

Quote
karabas
Can you sell a pair with AMT103? It's a bit expensive to buy internationally separate parts...

I suppose I could do that. Of course, I'd have to include the shipping of the AMT103's to me. PM me with your address, and I can work out all the shipping charges for you.