shorted stepper driver by finger?

This is embarrassing and frustrating but i just got my brand new ramps board and wired up one stepper and one end stop and jogged it with Pronterface. I was able to jog it one way, but not the other. I wanted to make sure the driver wasn't too hot so i touched the ic on the top of the x axis and at the same time the smallccontacts under it and it made a funny noise andthen stopped moving altogether. the readout in pronterface then reverted to a strange group of characters and: "Can't read from printer (disconnected?)" After this i tried just with the usb and it still read the same.
Cheers

Perhaps you were carrying a static charge. Electronics can be destroyed by ESD as even a tiny charge can be thousands of volts.

---

[www.hydraraptor.blogspot.com]

but could that have destroyed the whole board or just the driver? is there a way to test?

You didn't say what board it was. If it has plug in drivers you can remove the one you touched and see if the rest works.

Touching the driver may have damaged it but if you were unlucky it may have caused it to feed 12V to the CPU and damaged that as well.

---

[www.hydraraptor.blogspot.com]

well not sure where else to post this ...

after destroying a few A4988 drivers (white color ones from reprapdiscount)

i managed to order a few more cheap A4983 variant from ebay. incidently, these boards came with a larger sense resistor value than those from reprapdiscount (theirs is rated 000, the new ones are marked R200).

for some reason, these A4983s ... are so quiet ...

is there anything that quietness has to do with the sense resistance?

or isit that the A4988 board is badly designed?

on low speed tests, you can feel the A4988 board motor bumping thru the steps, not sure if this is call a near miss step, on the A4983s ... IT IS SUPER SMOOTH.

anybody have any takes on the above? :p

Edited 3 time(s). Last edit at 11/07/2012 10:31PM by redreprap.

attached files for ref

2786 = A4983
2787 = A4988

both clips audio highly amplified.

stepper supply 10VDC
steppers sanyo denki. per coil 10.6ohm, 0.85A/9VDC rated (VR adjusted for a drive approx 0.7A)
arduino setting, test pulse dump cycles at 768microseconds

all driver boards not modded for zero ohm.

*hi nophead --> would appreciate ur feedback, is the 2nd clip showing bad u-stepping? is the 1st clip a good example of smooth u-stepping?*

@ nophead "I don't think constant off-time choppers are the best idea. The current range is too limited and the switching frequency varies wildly. As the two halves of the chip run at different frequencies they can generate beat frequencies in the audio band."

when u say beat frequencies, u are refering to the "beats" like in clip2? it has a distinct high frequency sizzling beat. or does it include the medium audible constant whining rhythm like clip1, doesnt really sound like a beat there but there is a motor rumbly med freq hum. in both case, a zero ohm mod will make these whine go away i assume?

Edited 6 time(s). Last edit at 11/08/2012 12:23AM by redreprap.

redreprap Wrote:
-------------------------------------------------------
>
> stepper supply 10VDC
> steppers sanyo denki. per coil 10.6ohm, 0.85A/9VDC
> rated (VR adjusted for a drive approx 0.7A)
> arduino setting, test pulse dump cycles at
> 768microseconds

These are much higher voltage motors than we typically use.
>
> all driver boards not modded for zero ohm.
>
> *hi nophead --> would appreciate ur feedback, is
> the 2nd clip showing bad u-stepping? is the 1st
> clip a good example of smooth u-stepping?*

It is hard to tell at that speed, you need to step it very slowly so that each microstep can be seen and heard.

>
> @ nophead "I don't think constant off-time
> choppers are the best idea. The current range is
> too limited and the switching frequency varies
> wildly. As the two halves of the chip run at
> different frequencies they can generate beat
> frequencies in the audio band."
>
> when u say beat frequencies, u are refering to the
> "beats" like in clip2? it has a distinct high
> frequency sizzling beat. or does it include the
> medium audible constant whining rhythm like clip1,
> doesnt really sound like a beat there but there is
> a motor rumbly med freq hum. in both case, a zero
> ohm mod will make these whine go away i assume?

I think clip two is simply chopping at audio frequencies. With high voltage motors the on / off ratio needs to be large so, with fixed off time, the on time has to be much larger and hence the frequency drops. You make need to reduce the value of R4. Shorting it will force the off time to 30us, but that is probably too long for your motors.

By beating I mean that when stationary the two coils chop at two different frequencies. Even if they are both ultrasonic the superposition creates the sum and difference frequencies and the difference may well be in the audio band.

---

[www.hydraraptor.blogspot.com]

this is a A4983 on 1/16 step ... approx 600ms++ per step
i tried varying the VR to see what are the effects of current control on the "linearity" of the steps. seems like nearly max rated current is a good bet.

there are some audible hiss.

i assume that this is approximately a good paced 1/16 stepped? if ignoring the hiss?
i do note that the 5 steps per set square mark are not entirely perfectly spaced, but in this case, has anyone attempted to test this to the extent of perfect spacing? and to go that end, does the R4 mod do well in this?

Edited 1 time(s). Last edit at 11/08/2012 12:58PM by redreprap.

this is a A4988 on 1/8 step ... approx 600ms++ per step (on 1/16 i noticed some steps turning more then usual, is this miss stepping?)

even on 1/8 step, 2 of the 8 steps make bigger "leaps" then the usual. i assume this is what we term "miss-steps" ?

so in this case the shorting of R4 can be used to attempt to correct this?

this is how it looks like from the rear

stepper taped to a set square, and a wire taped to the pulley pointing to some angles on the set square

come to think of it... im quite happy i burnt those "white" stepsticks ... otherwise, i think i would not have dived into details of stepper driving.

and tnx for all your help nophead :p

Edited 1 time(s). Last edit at 11/08/2012 01:15PM by redreprap.

just to confirm, the SMD circled in blue is the R4 on a A4983? before i gut the wrong SMD to test mod lol

Edited 1 time(s). Last edit at 11/08/2012 01:43PM by redreprap.

May sound silly, but have you confirmed that all the screws in the motor are tight?

I found a loose grub screw on my extruder motor gear (it has 3) and it was rattling like crazy.

good idea, jus checked. its all tight. the audio sounds loud cos every clip is edited +40dB

Cefiar Wrote:
-------------------------------------------------------
> May sound silly, but have you confirmed that all
> the screws in the motor are tight?
>
> I found a loose grub screw on my extruder motor
> gear (it has 3) and it was rattling like crazy.

redreprap Wrote:
-------------------------------------------------------
> just to confirm, the SMD circled in blue is the R4
> on a A4983? before i gut the wrong SMD to test mod
> lol

Yes that is the correct resistor.

---

[www.hydraraptor.blogspot.com]

redreprap Wrote:
-------------------------------------------------------
> this is a A4988 on 1/8 step ... approx 600ms++ per
> step (on 1/16 i noticed some steps turning more
> then usual, is this miss stepping?)
>
> even on 1/8 step, 2 of the 8 steps make bigger
> "leaps" then the usual. i assume this is what we
> term "miss-steps" ?

The missed ones are where it doesn't move (or doesn't move enough) but they are then followed by steps that move twice as much since it has to catch up. They are not "missed steps" as when there isn't enough torque, they are only missed microstep positions.

>
> so in this case the shorting of R4 can be used to
> attempt to correct this?

There are several causes, if it is due to the off time being too long to get the current low enough then yes shorting R4 will correct it because it switches to fast decay, so doesn't need as much off time.

You can also get opposite with a high voltage motor. I.e. if you set Vref higher than the supply voltage (minus the drops in the chip) allows with the coil resistance then it fails to reach the peak current, so the top of the sine wave is flat. You can see both forms of distortion here: [hydraraptor.blogspot.co.uk]

People seem to assume you can drive any motor with a Pololu driver and indeed they are sold as general purpose, but R4 needs to be different for every motor.

The other thing to note is that you won't get perfect micro stepping with a cheap motor. The ones on HydraRaptor cost well over $100 and are supposedly optimised for microstepping. The obvious difference is they are cylindrical rather than square.

I think the best way to measure it (suggested to me by Adrian Bowyer) is to put a small mirror on the shaft and shine a laser pointer at it and project a spot on a sheet of paper the other side of the room. It is on my list of things to try.

Edited 1 time(s). Last edit at 11/09/2012 04:05AM by nophead.

---

[www.hydraraptor.blogspot.com]

hmmm

walls are flat, if mirrors rotate on the shaft, they need to strike a round target so that results would be linear for comparison i think. that will be tricky

getting 1/16 right seems really a handful if driver and stepper is not a "resonant" pair. did you managed to get a driver/stepper combi to work well @ 1/16 step?

i think for my case, i can rule out that A4988 will drive the sanyo stepper properly. the A4983 seems an ok match, at least it doesnt contain steps that exhibit large jumps.

i think i might next add more condensers around the drivers to help stabilize it.

Edited 1 time(s). Last edit at 11/09/2012 05:14PM by redreprap.

redreprap Wrote:
-------------------------------------------------------
> hmmm
>
> walls are flat, if mirrors rotate on the shaft,
> they need to strike a round target so that results
> would be linear for comparison i think. that will
> be tricky

The error due to wall being flat is minute over the sort of distance in involved. One micro step is only about 0.1 degrees which is about 5mm I think across a 3m room. So 16 would be only 80mm. When evaluating the driver we only need to look a 16. When evaluating a motor more would be required, so yes you might need a cylindrical scale for that.

>
> getting 1/16 right seems really a handful if
> driver and stepper is not a "resonant" pair. did
> you managed to get a driver/stepper combi to work
> well @ 1/16 step?

The 2.8V motors is use (SY42STH47-1684A) sound OK with R4 shorted. I.e. each microstep can be heard and at least moves some distance. I haven't done the mirror test to see how equal they are.

>
> i think for my case, i can rule out that A4988
> will drive the sanyo stepper properly. the A4983
> seems an ok match, at least it doesnt contain
> steps that exhibit large jumps.

I don't think there is any difference between A3983 and A4988 other than "no smoke no fire" pinout is there?

>
> i think i might next add more condensers around
> the drivers to help stabilize it.

Yes poor decoupling can increase the beating between the two channels.

---

[www.hydraraptor.blogspot.com]

yes it is strange isnt it, i think the only diff i have here between the A4988 /4983 ... is really not much ... but the way A4988 pushes the stepper, it is quite a long way off even at 1/8 step. i have tried R4 shorting on this chap, my guess is ... 4988 doesnt like high resistance / high voltage steppers? or isit a case of badly designed PCB, tons of spurious capacitance, etc? or theres something wrong with china A4988s lol?

with all things being equal, so either the cheap A4983s i got is really well made, or i should try one of those expensive A4988 with dual VR ...

hmmm yes i think you are right, 16 steps over 1.8degrees on a wall, curvature doesnt really make alot of diff :p

what stepper supply do you use for your 2.8v motors? 12v ? or somewhere in the region of 5v ?

nophead Wrote:
-------------------------------------------------------
> redreprap Wrote:
> --------------------------------------------------
> -----
> > hmmm
> >
> > walls are flat, if mirrors rotate on the shaft,
> > they need to strike a round target so that
> results
> > would be linear for comparison i think. that
> will
> > be tricky
>
> The error due to wall being flat is minute over
> the sort of distance in involved. One micro step
> is only about 0.1 degrees which is about 5mm I
> think across a 3m room. So 16 would be only 80mm.
> When evaluating the driver we only need to look a
> 16. When evaluating a motor more would be
> required, so yes you might need a cylindrical
> scale for that.
>
> >
> > getting 1/16 right seems really a handful if
> > driver and stepper is not a "resonant" pair.
> did
> > you managed to get a driver/stepper combi to
> work
> > well @ 1/16 step?
>
> The 2.8V motors is use (SY42STH47-1684A) sound OK
> with R4 shorted. I.e. each microstep can be heard
> and at least moves some distance. I haven't done
> the mirror test to see how equal they are.
>
> >
> > i think for my case, i can rule out that A4988
> > will drive the sanyo stepper properly. the
> A4983
> > seems an ok match, at least it doesnt contain
> > steps that exhibit large jumps.
>
> I don't think there is any difference between
> A3983 and A4988 other than "no smoke no fire"
> pinout is there?
>
> >
> > i think i might next add more condensers around
> > the drivers to help stabilize it.
>
> Yes poor decoupling can increase the beating
> between the two channels.

I use 12V supply with 2.8V motors. As the current I use is only ~ 1.2A and they are 2.8V motors at 1.7A then they are only 2V motors at 1.2A.

---

[www.hydraraptor.blogspot.com]

hmmm interesting, 2v ... but still the driver does dump the full 12v into them dont they? only thing is the driver will clamp the current to 1.2A. somehow the power has to go somewhere, either the motor fully sinks this power, or the driver takes up the excess (which i doubt is the case, unless its some kind of linear driver).

i dont have a proper scope here to measure the voltage drop of drivers. have you tried?

when i put my DMM to 1 of the terminals of the stepper, there was an instant i caught what i think is a fly back spike that is higher than that of Vin. which made me to think i need to put in some kind of arrestor condensers of sorts. i am thinking 224J caps in various places, or maybe something smaller maybe some 63v 473J right at the motor terminals.

what do you think?


nophead Wrote:
-------------------------------------------------------
> I use 12V supply with 2.8V motors. As the current
> I use is only ~ 1.2A and they are 2.8V motors at
> 1.7A then they are only 2V motors at 1.2A.

redreprap Wrote:
-------------------------------------------------------
> hmmm interesting, 2v ... but still the driver does
> dump the full 12v into them dont they? only thing
> is the driver will clamp the current to 1.2A.
> somehow the power has to go somewhere, either the
> motor fully sinks this power, or the driver takes
> up the excess (which i doubt is the case, unless
> its some kind of linear driver).

They are chopper drives, so there is always +12, -12V or a short across the coils (minus a little for the on resistance of the MOSFETs, which is a fraction of an Ohm).

Because the motor coils are inductors though, the current doesn't obey Ohms law. When +12V is applied it starts to increase but as soon as it gets to the target value it switches off and the current starts to decay again. It never gets as high as it would be if 12V was applied permanently, which would be the Ohms law value.

So to say "the power has to go somewhere" is untrue. This is the difference between a linear constant current source and a switch mode chopper.


>
> i dont have a proper scope here to measure the
> voltage drop of drivers. have you tried?

Yes, the waveform is a very messy square wave.

>
> when i put my DMM to 1 of the terminals of the
> stepper, there was an instant i caught what i
> think is a fly back spike that is higher than that
> of Vin. which made me to think i need to put in
> some kind of arrestor condensers of sorts. i am
> thinking 224J caps in various places, or maybe
> something smaller maybe some 63v 473J right at the
> motor terminals.
>
> what do you think?

You shouldn't need to. The chip has the four body diodes of the MOSFETs that clamp the outputs to the rails. In fact it also has synchronous rectification, which means the MOSFETs are turned on to short the diodes at the right moments to avoid the power loss in the diodes.

If there is no other load on the 12V rail you can it get being increased by the back EMF from the motors, but with several motors I don't think it is a problem on RepRap machines.

Edited 1 time(s). Last edit at 11/11/2012 07:40PM by nophead.

---

[www.hydraraptor.blogspot.com]

hmmm ... ima need to go try out the free lesson from EDx on electronics lol