stepper motor bulk purchase

Hey all,

Adrian and I have been talking about the RRRF doing a bulk purchase of the 'standard' stepper motors to sell to non-european researchers. I'm posting to find out what type of interest there is from you guys in ordering them. I'm not going to lie, they are pricey. A set of 3 is going to cost $180, minimum. Probably close to $200 once we factor in shipping, paypal, etc.

Here is a link to the specs: [www.nanotec.com]

Please reply here either way... i'd like to get a feeling for what you're thinking.. yes *or* no.

Also, if anyone could find a comparable motor that is more readily available in the USA, we could also go that route. Thanks for your help!

~Zach Smith

Deleted as off-topic

Edited 3 time(s). Last edit at 08/19/2007 01:49PM by Forrest Higgs.

A bulk buy is a good way to get the price down a bit but starting from something more reasonable would be better.

As these are both expensive and obscure motors wouldn't it be a good idea to ask the core team to test and approve a cheaper 1.8 degree motor and flip the firmware to half step mode. Or is there some special reason these were selected?

---

[www.hydraraptor.blogspot.com]

Deleted as off-topic

Edited 3 time(s). Last edit at 08/19/2007 01:50PM by Forrest Higgs.

Deleted as off-topic

Edited 4 time(s). Last edit at 08/19/2007 01:50PM by Forrest Higgs.

Forrest, the drive chip on the UCB is L298 which is bipolar.

Note that that the bipolar series resistance is a misprint on the datasheet. It should be 2.46 and the bipolar series voltage should be 3.72V.

The problem with modern stepper motors like this, which are built for speed, is that they are all low voltage. You would either have to add series resistors like Eric did or rely on PWM to keep the current under control.

Perhaps UCB 1.3 should support constant current chopper drive. That gives better performance and allows almost any motor that is physically compatible to be driven.

Although it would be better to use a dedicated chip, I think a fairly simple mod to the UCB would be to add sense resistors and a dual comparator chip with open collector outputs, feed the enable signals through a resistor and allow the comparator to drag them low when the target current is reached. That would be the addition of one cheap chip and quite a few resistors and probably a couple of caps. The firmware would not need to change.

It might also be possible to do the chopping using the comparators built into the PIC but that would require some pin juggling and firmware changes.

Edited 1 time(s). Last edit at 08/19/2007 05:56AM by nophead.

---

[www.hydraraptor.blogspot.com]

Forrest Higgs Wrote:
-------------------------------------------------------
> If you look at the torque/pps curve at the bottom
> of the page you're only getting about 6 oz-in of
> available torque out of it in the 1-2K pps range
> using 12v nominal

I think the Darwin design won't care what the torque is in the kHz range. I don't remember the numbers off the top of my head, but I think I read that a 0.9 degree step in Darwin translates to just under 0.1mm of movement on the X or Y axes. If we extrude at 1cm/s, that's only 100 pps.

OTOH, for the repstraps using threaded rod, high speed performance is a big deal.

> 11: NEMA 23 BIPOLAR STEPPER MOTOR 185 oz-in
> ,

Forrest Higgs Wrote:
-------------------------------------------------------
> Holy crap! You had to size up to a NEMA 23 to get
> enough torque?
>

umm... we've been using NEMA 23 since the beginning.

Forrest Higgs Wrote:
-------------------------------------------------------
> Zach, let's take a deep breath before you sign any
> checks for a moment. I don't keep track of where
> Darwin is most times, but looking at the only pic
> that I can find of the stepper Adrian used...

dont worry... the only one getting excited seems to be you

> [reprap.org]
> nics/65.JPG
>
> I see that it is an ST5709S12(08, I think) with a
> coil amperage rating of 1.2. It's also a unipolar
> stepper with a holding torque of 120 oz-in (85
> Ncm).
>
> If you look at the torque/pps curve at the bottom
> of the page you're only getting about 6 oz-in of
> available torque out of it in the 1-2K pps range
> using 12v nominal
>
> [www.nanotec.com]
> ml?attribute_id=12&option_id=13
>
> I found another supplier with this spec...
>
> 11: NEMA 23 BIPOLAR STEPPER MOTOR 185 oz-in
> ,

Deleted by request of thread originator

Edited 6 time(s). Last edit at 08/19/2007 01:51PM by Forrest Higgs.

Personally, I'm going to try and find some cheaper steppers and take my chances. $180 seems ridiculous to me, given the alternatives.

I'm going to try to wait this out a little longer. If I can save US$90 out of US$500 ish by waiting a bit for someone to verify that a 200 step/rev stepper motor costing US$33 works fine, then I'll wait. I don't have a robot to put steppers in yet anyway :-) So in terms of the proposed stepper motor bulk purchase for researchers, for now, my response is clearly: "No, thanks".

A few related questions come to mind:

(Q1) Zach: How close are you to being able to verify that the NEMA-23 1.8 deg/step motors you have are sufficiently capable (torque and precision) for a Darwin robot? Days? Weeks? Months??

(Q2) Further: Is there some way someone with a working Darwin robot (with expensive steppers or otherwise) can measure the needed torque, perhaps? I have no idea how one would do that... put a torque wrench in place of the stepper drive shaft somehow, and see how low you can set it and still move the extruder around?? There's probably a better/saner way of testing this that mechanically-minded people know?

(Q3) For the number of steps question, could we write some test-only firmware or hack the host software to double the number of steps requested, so deliberately downgrading the expensive Nanotec steppers into 200 steps/rev ones from a host software/firmware point of view?? Then see how much precision is lost by doing that, and whether it is likely to matter for the Darwin RP-able parts (my theory being that replicating those parts is goal #1, anything else Darwin v1.0 can make is gravy!) ?

Thanks,

Jonathan

>
> (Q1) Zach: How close are you to being able to
> verify that the NEMA-23 1.8 deg/step motors you
> have are sufficiently capable (torque and
> precision) for a Darwin robot? Days? Weeks?
> Months??

i can only really answer this question: it depends on how hot is in the next month. if its fairly cool, i'll be able to get some work done in the lab, otherwise i'll spend most of my time on the wiki and other online stuff.

i'm hoping that this winter is really productive, and i should be able to get quite a bit done. i also plan on getting an air conditioner for next summer. there are also a couple other machines out there with Vik and Simon. i would consider them much more capable than me, so i'm hoping that they beat me to it.

jmarsden Wrote:
-------------------------------------------------------
> (Q2) Further: Is there some way someone with a
> working Darwin robot (with expensive steppers or
> otherwise) can measure the needed torque, perhaps?

The torque wrench is what came to mind for me too, but that seems like it would be really hard to do right. We may be able to get a rough guess by finding out what torque setting people use. If someone could just keep dropping the torque setting until the motor started skipping steps, we may be able to work backwards and estimate how much torque the motor was producing at that setting. I think we have all of the necessary info.

Hmm... the only setting listed on the "tweaking" page has it pegged at 100%.

> (Q3) For the number of steps question, could we
> write some test-only firmware or hack the host
> software to double the number of steps requested

I don't think that would tell you anything useful. The 400-step motor is designed to have each one of its 400 full steps accurate to some spec, usually around 0.05 degrees. Same thing with the 200-step motor. If you try to make the 400-step motor into a 200-step motor by double-stepping, you'll still have the same 0.05 degree accuracy since you're still using the full-step positions it's specified for. When you try to make a 200-step motor into a 400-step one, every other step is a half-step where the motor is stopped between the full steps where accuracy is guaranteed.

ObOnTopic: Newark is currently listing the official steppers with a lead time of <=7 days. Looks like the three motors would run you $176.49 + whatever they charge for shipping. IIRC, they don't tell you what the shipping charge is before you checkout, you have to check your credit card to see what they charged. For me it has always been reasonable, <$10 for ground I think.

I second the 'I don't have a robot built yet, so I'm waiting for a cheaper breakthrough' motion... If my robot build plans work out and the motor situation hasn't changed by say, the end of the year, looks like this might be a christmas present. But order steppers today? Nay!

testing the torque.

Hmm if you need to meassure the torque on the stepper it might be easiest to get it connected to a rod sticking out sideways and then hang a tray with water in it on the side. The rod length X the weight of the water would give you an estimate, just remember that the holding torque is higher than what you can actually move.

Hmmm..... Sparkfun has $15 1.8 degree steppers that appear appropriate from the specs.

These are just about as expensive, but this might be a more accessible option for RRRF. I'm looking into their other motors (brushed DC and brushless DC) for my own projects.

[www.anaheimautomation.com]

One thing is that they're willing to play ball with special modifications over their in-stock designs.

RussNelson Wrote:
-------------------------------------------------------
> Hmmm..... Sparkfun has $15 1.8 degree steppers
> that appear appropriate from the specs.

DOH! NEMA-17, not NEMA-23. Sorry. But still, NEMA-17 motors might work, and they're a lot cheaper. It's just that they're not what is currently specified, so my suggestion was junk.

Oh, and they're out of stock of the motors until November, too.

So Do 1.8 deg Motors work?

are are of these any good? [www.oceancontrols.com.au]

... testing the needed torque of the motors can be easy made with a string wound around the axis (or a bigger roll on it for easier calculating), straightened horizontally and feed over a fixed roll and then append a weight on the string (maybe a small bag or cup for stepwise increase mass with heavy parts).

So the motor can turn freely around and you can calculate the torque from radius of the motoraxis (or roll) and the tangential applied weight-force ...

Viktor

Any progress reports on this? anyone find a cheaper motor?

well,

i've had success with both of these motors:

[www.jameco.com]
and
[web4.automationdirect.com]

the latter seems to be higher quality / stronger / etc. unfortunately it doesnt really work too well with the current v1.2 UCB because it doesnt have chopping. works like a charm with the new stepper board though

I think between us we may have run Farnell out of stepper motors

I ordered mine about a month ago and they haven't turned up yet. At the time I noted that they had a stock of 2 and I was wanting 3. A quick look on their site today reveals they are still out of stock.

I also read on various threads that there are a few others in my situation.

Maybe it would be better to do a bulk purchase for the store of one of the motors above in anticipation of the flood sure to come with the new boards.

I'm looking at the pages now and I got quite a shock at the price difference when I saw the ones Zach was pointing at. They seemed to be 1/3 the price of the ones I bought. Then I noticed the currency was different and maybe don't include shipping.

Still, I estimate the prices are around half. Is that right?

Edited 1 time(s). Last edit at 11/16/2007 08:41PM by reece.arnott.

With the new Arduino boards or the existing boards with the Half step enabled you can use a standard 200 step motor Nema 23 these are much cheaper in UK about

ian adkins,

i agree that is a major benefit of the new stepper boards.

currently i'm using some motors from jameco very successfully, and i do know they offer a bulk puchase.

i'll probably only be able to order 10-25 motors to start out, can anyone find a better price/performance than this one:

[www.jameco.com]

Zach,

I assume that your 10-25 motor order is a money issue? Perhaps those of us wanting a set can pre-order through the RRRF store so that you have more on-hand capital with which to get a larger lot of motors and thus a better price. Obviously you can calculate the cost per lot of 3 and add in an appropriate shipping cost so we can pre-order them. If this isn't the issue, just ignore me. :-)

Demented

Hi Zach

That motor is cheap but not same spec as currently used.

I'm sure it would work but it will not fit the RP parts as standard and is considerably less powerful (although i think the current motors are larger than needed).

I've not given the required time to really investigate the Arduino yet as I've been full on doing moulded parts (nearly there) but certainly Micro-stepping is the way to go no question.

UK side we can do 200 step versions of the motor Adrian & eD are using for

Ian Adkins Wrote:
-------------------------------------------------------
> Hi Zach
>
> That motor is cheap but not same spec as currently
> used.

true, there are 2 major differences:

1. only one output shaft
2. its 200 step rather than 400 step (solved with half-stepping)

however, its half as expensive which is huge.

>
> I'm sure it would work but it will not fit the RP
> parts as standard and is considerably less
> powerful (although i think the current motors are
> larger than needed).

well, its the exact same size (NEMA 23) so it fits perfectly in the current mounts. i've wired them up and tested them with both the old and new boards and it works just fine.

> I've not given the required time to really
> investigate the Arduino yet as I've been full on
> doing moulded parts (nearly there) but certainly
> Micro-stepping is the way to go no question.

agreed (obv.)

> UK side we can do 200 step versions of the motor
> Adrian & eD are using for

Demented Chihuahua Wrote:
-------------------------------------------------------
> Zach,
>
> I assume that your 10-25 motor order is a money
> issue? Perhaps those of us wanting a set can
> pre-order through the RRRF store so that you have
> more on-hand capital with which to get a larger
> lot of motors and thus a better price. Obviously
> you can calculate the cost per lot of 3 and add in
> an appropriate shipping cost so we can pre-order
> them. If this isn't the issue, just ignore me.
> :-)
>
> Demented

yup, its a money issue (anyone have a home for the RRRF in NYC that is rent-free?). making a pre-sale item is a good idea. how about i put the pre-sale motor up on the store then in 2 weeks i'll do as big of an order as possible.