Stepper motor + encoder resolution

[JoeyS]

Hey everyone,
I'm designing a syringe pump from scratch for a medical device. This is my first project with a stepper motor so I have a question regarding encoders.
I'm currently using the full-step configuration of the stepper motor, so 200 steps per revolution. Which kind of rotary encoder resolution should I be looking for?
Will a 1024 line up properly? It seems like I'd have trouble with mismatching PPR and steps per revolution.
Should I go for an encoder with 200/400/600/1000 PPR?
On the other hand I'd like to have the option of using the same encoder if I decide to go half- or quarter-steps.
Any advice will be appreciated.

 

[BrianPetersen]

Prime control loop needs to work by counting the steps, with the encoder only acting as a home-position switch or back-up of sorts to confirm that the stepper hasn't gone way out of whack (mechanically jammed). If done this way, the encoder doesn't matter much.

 

[FACS]

There is no benefit of an encoder with more than 200 pulses per revolution.
You can't do a "Half step" on the stepper motor. All you want to do is verify 200 steps per revolution.

It's more important to match each step of the stepper motor with a desired quantity of material moved by the syringe.

 

[Compositepro]

Well you can do half-steps, or smaller (micro-stepping) by controlling coil voltages and energizing more than one coil at a time. You are not likely to be using a stepper motor without a gearbox for this application, so the pulses get multiplied by the gear ratio.

 

[Rodmcm]

Closed loop stepper motors usually have 1000 segments per rev, but with quadrature encoders you can get up to 4000 positions per rev.
Micro stepping can be used on 1.8 deg stepper motors. This holds the position between 1.8deg using current control

The max is 51200 pulses per rev but you don't often see this with small drivers.

Have a look at the drivers for 3D printers especially the TMC5160 or newer version range for smaller applications which yours probably is. Programable with Marlin.

See

https://www.youtube.com/watch?v=ek4qZk0aRbE

on Closed loop controller

 

[IRstuff]

Seems odd that you'd be dinking around with the encoder without having specified what the minimum travel of the syringe, or minimum incremental volume dispensed; either of which would dictate the minimum travel of your actuator and lead to the minimum stepping increment required from the motor.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[JoeyS]

Thanks for the replies - they are very helpful!

IRStuff -
I appreciate your input very much. I absolutely agree that the minimum travel distance of the syringe is a determining factor in step size.
However, we're designing a syringe pump for 120ml syringes, so the volume for each step is most likely negligible even for full steps.
Of course we will have to verify this in the near future. In the meantime we're assembling the mechanism as a proof-of-concept for the required forces and velocities.

 

[IRstuff]

Nevertheless, the minimum required dispensable volume dictates the stepper requirements. Is the minimum dispensable volume anywhere in the nL regime?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[JoeyS]

The minimum dispensable volume is in the mL range.
Our syringe dispenses approximately 1 [ml/mm].
This would mean that with a 200 step per revolution and a pitch of 1[mm] we'd be well within our bounds.
The reason I'm researching other step options is that I'd like to consider using that can be reconfigured to suit
more needs, since we still have a lot of research and consequent prototyping to do.

 

[IRstuff]

That's a choice between resolution and speed. Having a sloppy thread limits your accuracy; even "precision" threads, like 80 tpi, have slop. Nevertheless, 80 tpi is 0.3175 mm per revolution, which would get you 3 times the precision with no change to the servo motor part of things, and with 200 steps/rev, you'd be able to dispense 1.6 uL, albeit, the accuracy might be worse, unless careful preloading, etc., is taken into consideration.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[SwinnyGG]

IR raises an important point - you need to be thinking about the accuracy of your dispensing process, not just sizing your controls (i.e. stepper and encoder) to provide the quantity you need. Being a medical device, I'm going to assume you also need a high level of accuracy with regard to dispense volume. In most general engineering disciplines, you would target a measurement accuracy of 10x the requirement (i.e. if you need to dispense 1 ml, you need measurement accuracy of .1 ml or better. Don't forget measurement accuracy when you're sizing things and choosing sensors.