I need a few sets of Pololu-type drivers for some RAMPS boards I'm building. I fancy having a go at building Stepsticks, but I've got a few questions.

1) The Pololu A4988 Stepper Motor Driver Carrier [www.pololu.com] operates from 8 to 35 V and can deliver up to 2 A per coil. The Allegro A4988 driver chip is soldered to the PCB under the chip, and it uses this to dissipate heat into the copper of the PCB. The A4988 datasheet talks about a 4-layer PCB. I emailed Pololu, who replied "the boards for the A4988 stepper motor drivers are 2-layer boards with 2 oz copper". Does anyone know if the Stepstick has the same setup?

2) I read a post from nophead [forums.reprap.org] that said the Stepstick limits current to 1A. Is this true?

3) The BOM for the Stepstick is a little light. So far I have ascertained:
Allegro A4988 (IC1)
Capacitor, 0402, 0.1uf (C1, C2, C5, C6)
Capacitor, 0402, 0.22uf (C4, C7)
Capacitor, 1206, 4.7uf (C3)
Resistor, 0805, 0.2ohm, 0.25W, 1% (S1,S2)
Resistor, 0402, 10k, 10% (R4)
Resistor, 0402, 20k, 10% R1)
Resistor, 0402, 100k, 10% (R2, R3)
Trimpot, 3mm, 10k ohm (T1)

Just need a little help with the capacitors; what voltage, tolerance should they be? If there's any other specifics that would be useful, please add it.

Edited 1 time(s). Last edit at 04/10/2012 08:50PM by droftarts.

The stepstick is electrically almost identical to the pololu stepper driver
[www.pololu.com]
There is a reference implementation that Allegro give in the datasheet which is basically the pololu or stepstick circuit.
[www.allegromicro.com]
Nophead has just done a very interesting post about the subject here:
[hydraraptor.blogspot.co.uk]
I Nophead's blog he mentions the effect of the choice of R4, since you are making them yourself you can short out R4, or pick a different value for it as best suits your motors.

1) The stepstick board can be manufactured with varying copper weight, check with your supplier, the thicker the copper the better the heat transfer away from the A4988. The setpsticks I self assembled on 1Oz boards have been running my Prusa for some months now. I recently stuck some heatsinks on them as I was turning up the speed and they we getting hot.

2) The current that can be drawn though the A4988 chip is limited by the sense resistors, the datasheet explains the effect of the sense resistor(RS):
ITripMAX = VREF / ( 8 x RS)
Where VREF is set with the pot. For the Stepstick, JoeM has chosen RS as 0.2 Ohm and VREF is between 0 and (TrimpotMaxR)/(TrimpotMatR+R1) of VDD. with a 100k trimpot and 20K R1 this is 4/5 of VDD. VDD is 5V on Sanguinololu, RAMPS, etc. Using a Trimpot of 100K, ITripMax is 2.5A, more than the board can supply anyway. Interestingly pololu use 0.05 Ohm, which works out as 10A, well more than the board can supply.
In reality ITrip is a % of ITripmax depending on which coil and what step or microstep its on, this information is in Table2 of the datasheet. The limit is often thermal on the A4988, with the thermal protection kicking in and shutting them down for a short period of time, hence fitting the heatsinks.

3) The trimpot on the original BOM on the wiki did not appear to be specified, it was then posted to be 10k Ohm. I made some with this and they work but given that R1 is 20k and VDD is 5V, VREF maxes out at 1.7V which limits ITripMax to ~1A. this may be what Nophead is referring to when he says it is limited to 1A. This is probably fine for most applications as around this value thermal protection is normally kicking in anyway. I made some more using a 100k trim pot (the same value as pololu use on their boards). This also worked, just has to keep the pot turned down lower! I used 16V, 0.22UF, 10% capacitors because I knew I was going no higher than 12V. if you want to run your stepsticks up to the maximum for the A4988 then you need to go up to 35V+.

These can be challenging to make depending on your experience with re-flow techniques. I did not have much practical experience before hand and I had some failures along the way. It was an enjoyable challenge and I like the fact that the stepsticks on my printer were made by me.

Hope this helps

Tony

The StepSticks I have use 10K pots and do limit the voltage to 1.7V.

I think if you go any higher the sense resistors dissipate too much for a resistor that size. That is why Pololu use such a small value. It allows a small cheap resistor and I think they also limit VREF with a 10K pot, so the wiki is correct. The downside is more sensitivity to noise in the sense circuit.

Edited 1 time(s). Last edit at 04/11/2012 07:11PM by nophead.

---

[www.hydraraptor.blogspot.com]

The stepstick is more sensitive to noise? or the pololu?
I am trying to decide which drivers to go for. If I got the stepstick board, couldn't I just put on .05 ohm sense resisitors in place of the .2 ohm sense resisitors?

The Pololu will be because the sense voltage is smaller so any noise has a bigger effect. Doesn't seem to be a problem though.

Yes you could build a StepStick with 0.05R resistors. VREF should then be limited to 0.8V which gives 2A and 0.2W dissipation in the resistors.

If I were designing it I would use 0.1R sense resistors and limit VREF to 1.2V to give 1.5A max and 0.225W dissipation in the resistors. I think that is the best compromise.

---

[www.hydraraptor.blogspot.com]

Thanks a bunch, I think I will try to throw that together.

Thanks for the info Tony, nophead. I've add a BOM, links and some extra info to the Stepstick wiki page to help others. If someone could check it over for mistakes, that would be helpful. [reprap.org]

nophead, if the sense resistors change to 0.1R, would the value of the trimpot need to change as well? And what if the value of the sense resistors were 0.05R? Having a range of options might be useful, but I'm not confident in my ability to interpret the formulae in the posts above to work out the proper values. The commonly-used Zapp SY42STH47-1684A has a current rating of 1.68A, which would be good to match. [www.zappautomation.co.uk]

I've just heard from Roland at think3Dprint3D, who is planning on ordering 2oz Stepstick PCBs when his current stock runs out.

The pot and its series resistor define the maximum VREF = 5 * R9 / (R5 + R9). Ideally that should not allow the voltage across the sense resistors to exceed the dissipation of the sense resistors. I think the footprint is for a 0.25W resistor. The Pololu one does at 5V because they values are chosen to work at 3.3V as well. It isn't an issue though because the current would be more than 2A, so the chip would shut down.

If Roland has changed to a 100K pot then you could set the current high enough to burn out the sense resistors before the chip shuts down.

It is easier to limit VREF by changing R5 because fixed resistors are available in more values than pots.

The 1.68A current rating of the motor is not really relevant. That is for two coils on and 80C temperature rise. The Allegro current is stated for peak with one coil on and when both coils are on it is 70% to each. That means to run the maximum motor power you need to set the Allegro current 1.4 times higher, i.e. 2.4A, which it can't do. If it did the motor would hit 100C and the brackets would melt.

We typically run them at about 1A, peak, which is only 0.7A with two coils on. So we get 0.7/1.68 of the torque on the datasheet and (0.7/1.68)^2 * 80 temperature rise, ~14C. I tend to run my extruder motor at 1.3A, which is gives me 23C temperature rise. This is why I suggest a sensible limit of 1.5A and that allows a 0.1R 0.25W sense resistor.

To get 1.2V from 5V and a 10K pot R5 should be 10 *(5 - 1.2) / 1.2 = ~30K.

The thing you have to watch out for with 2oz copper is the minimum track and gap sizes get bigger because the etching is less accurate due to undercut.

Edited 1 time(s). Last edit at 04/17/2012 05:30PM by nophead.

---

[www.hydraraptor.blogspot.com]

Thanks again, nophead, for the extra info. I'm just a little confused by your notation. I've based the BoM on the Stepstick page on the notation used on the Stepstick schematic [reprap.org]
Just to be clear, when you say R5 and R9, do you mean the trimpot for R9 (marked T1) and it's resistor for R5 (marked R1)?

To summarise, to change the current limit from 1A to 1.5A, you change the trimpot resistor (R1) to 30k, and the sense resistors (S1, S2) to 0.1R 0.25W

What's the best way to measure the current going to the motors? I remember on my Gen6 there was a test point you could test the voltage across, to know how much current was going to the motor. Is there something similar on the Stepstick/Pololu?

droftarts Wrote:
-------------------------------------------------------
> Thanks again, nophead, for the extra info. I'm
> just a little confused by your notation. I've
> based the BoM on the Stepstick page on the
> notation used on the Stepstick schematic
> [reprap.org].
> Schematic.png
> Just to be clear, when you say R5 and R9, do you
> mean the trimpot for R9 (marked T1) and it's
> resistor for R5 (marked R1)?

Yes I was using the designators on the Pololu schematic.

>
> To summarise, to change the current limit from 1A
> to 1.5A, you change the trimpot resistor (R1) to
> 30k, and the sense resistors (S1, S2) to 0.1R
> 0.25W

Yes.

>
> What's the best way to measure the current going
> to the motors? I remember on my Gen6 there was a
> test point you could test the voltage across, to
> know how much current was going to the motor. Is
> there something similar on the Stepstick/Pololu?

There is a test point for VREF on the Pololu but it is missing on the Stepstick. Since it is just the wiper of the pot you can measure it there and it is easier as it is a bigger target. I hold the meter probe on the shaft of a metal screwdriver so I can see the value while I am turning the pot.

The current will be VREF / (8 * RS).

---

[www.hydraraptor.blogspot.com]

Many thanks, again, for the clarification, nophead.

One last question (I hope): on your blog you suggest shorting R4 to enable Low Current Microstepping. What are the highest currents, at both x8 and x16 microstepping, that R4 should still be shorted? Just trying to work out if, for all cases that are relevant to us, this should be shorted when using the A4988.

That depends on the supply voltage, the motor resistance, the sense resistors and the limits of your hearing. As the current becomes lower the off-time needed to achieve that current with the minimum on-time gets longer. You have to set the fixed off time a little bigger than that. Then at maximum current the maximum on-time plus the off time give you the minimum frequency. The limiting factor that prevents you making it lower is when it becomes audible.

The formulae are on my blog.

Edited 1 time(s). Last edit at 04/19/2012 10:28AM by nophead.

---

[www.hydraraptor.blogspot.com]

@nophead


I shorted R4 on all of my drivers because I had very low resistance motors and it helped a lot.
Is there a time when shorting R4 will have a negative impact? What would the negative affects be?

The only downside I have noticed is the motors are more noisy when stationary. I think that is because the ripple current is higher because it is using fast decay more often. That also might give a bit more heating in the motors and a slightly higher power consumption but I haven't noticed such.

---

[www.hydraraptor.blogspot.com]

@bethebeer: did you make up some stepsticks with 0.1R sense resistors? If so, what value trimpot did you use?

I'm just coming back to this after a very busy summer, and want to make up another set of stepsticks that provide a bit more current, as well as swapping the R4 resistor for a 0R to enable "Low Current Microstepping" mode - with the standard Zapp NEMA17 motors moving at slow speed, I notice the pulsing noise like on nophead's video. I successfully built 6 stepsticks, using 0.2R sense resistors, but found they were right on the limit of the power the motors needed, sometimes leading to skipped steps. Having just a little more 'oomph' would be handy.

If you (or anyone else) could provide a link to the trimpot you used, that would be very useful - especially if you got it from Farnell in the UK! The trimpots I bought [uk.farnell.com] have an incredibly tiny adjustment pot, and no stop at the limit of their movement, so it is tricky to set them. Also, I didn't seem to be able to measure the VREF through the pot wiper. Is there a better option? Perhaps a 4mm pot would fit?

@droftarts: i have never used stepsticks. I have been using pololus with r4 shorted. I just replaced the resistor with a strand of wire from 7 strand 24AWG. They have been working great and supply enough power. I have the 0.1ohm resistors, but never bothered to put them on because everything has been working. I'm also scared to do it because I just have a soldering iron and no good way to work with smt parts yet; shorting r4 was scary enough. After I get my next bot together I would be willing to test it out, if you haven't yet done it yourself.

@bethebeer - Okay, thanks. I'm actually using genuine Pololu A4988 boards as well, and haven't replaced/shorted R4 on them either. There's no point changing the sense resistors, though it's probably doable with a soldering iron as they are quite large, as they would be badly mismatched with the 100k trimpot, I think.

Reading the thread back, I realise that I want, per stepstick, two 0.1R 0.25W 0805 sense resistors (S1, S2), and change the trimpot resistor from 20k to 30k (R1) for a current limit of 1.5A (as opposed to 1A) in conjunction with the existing 10k trimpot. Also, change R4 for a 0R 0402 (or wire, or solder bridge), to enable Low Current Microstepping mode. S1, S2, R1 and R4 are the stepstick position notation from [reprap.org] - the Pololu notation is different.

Any help sourcing a better 10k trimpot would be greatly appreciated. To paraphrase Princess Laya, "help me, Nophead, you're my only hope!"

[www.futurlec.com] from new york
or
[www.taydaelectronics.com] from thailand


"would be badly mismatched with the 100k trimpot":
yea, didn't think that one through.

I have made up a set as per nophead's suggestions, and I have tested them briefly to check all is okay. I've only been able to check the VREF voltage so far, and have set them at around 0.8V for 1A ( A = VREF/(8 x Sense resistor ohms) = 0.8 / ( 8 x 0.1 ) = 1 ). I'll double check the Amperage, hopefully tomorrow. I'll install them on my reprap and see how they go.
I also set the ROSC resistor at 0 ohm for Low Current Microstepping mode, as I am using the same low resistance motors as nophead, from Zapp Automation. Moving a motor with feed rate of 1, it seems to be smooth, but I need to compare it with a Pololu to check for an improvement.
I made them up with 0603 SMT packages; seems to work fine, and much easier to cope with than 0402! The main chip is still the biggest headache to solder, as it's a QFN package. I had a friend cut a solder stencil for me on a craft vinyl cutter (Silhouette), which seems to work well, and I built myself a solder over like Adrian's: [reprap.org]
I've built 10 stepsticks so far, only had one failure, which should be recoverable anyway. I don't plan to go into production or anything - though I have bought enough components for 40 stepsticks, it is rather labour intensive - just wanted to see if I could do it. All 40 have been spoken for already, I'd better get on with making them! Thanks for the help.

I have read through this article and can't fully get my head around all the different resistor selections. Let me summarise what I think it all means and would really like to hear your thoughts.

I also number all questions so you can send a quick response to any questions you know the answer

Assuming the stepstick layout,

Release says use S1, S2 = 0.2 Ohm, 0.25 W
R4 = 10 k
T1 = 10 k

Most vendors sell stepstick with S1, S2 = 0.05 Ohm
1) presumably they still use T1 = 10k - would that work?
2) Also, is 0.1 W ok for these pots?

3) Nophead reccomends S1, S2 = 0.1 Ohm, 0.25 W and a 30 k pot?
4) Is there really an advantage over the 0.05 / 10k combo?
5) Does it matter?

6) I want to buy some new stepsticks, and vendors sell 0.05 Ohm, should I just buy these or wait for someone with Nophead's selection?

7) Also, shorting R4, are we talking R4 on the stepstick (i.e. the 10K?
8) Is this just to change what you hear or does it affect the performance?
9) If performance, is there a real penalty for not shorting R4?
10) Why do vendors sell with the resistor, is there some good reason for not shorting it for other applications?

Sorry, I set out to summarise what others said, and ended up asking a load of questions instead

BUMP

Hey guys, hope you all had a nice Christmas break!

Now would someone like to revisit this post and give some feedback?

Much appreciated!

Happy new year to all!

Is anyone able to comment on the above?

I think all the answers and the pros and cons are in the thread above.

---

[www.hydraraptor.blogspot.com]

@JNOD

When it comes down to the fine detail there is not a lot in it, there aren't really any hard and fast rules like X is always better than Y. It depends on your precise requirements. It's only worth worrying about if you are going to design your own boards, or encounter problems.

Most of the boards are "good enough" for most purposes, given the limitations of the tiny Pololu form-factor which is not at all ideal.

Hi,

Does anyone know where to get the "latest G3D" drivers as they are being sold here [gadgets3d.com] ??

They are not being updated anywhere online according to the GPL...

Edited 1 time(s). Last edit at 02/28/2013 12:28PM by Simba.

That link is where to get them, or on eBay?


Also, the new Rambo board puts a 100k ohm resistor between the ROSC pin and the ground; this is what the wiki schematic looked like to me. I just want to make sure that it would be worth shorting that out. Why would there be such a large resistor there?

Its a 0 ohm resistor used to jump another trace, the 100k resistor is on the sleep and reset pins.

awesome. Thanks a bunch, glad I misread that.