Are A4982 drivers supposed to be so noisy?

Just trying out some A4982 drivers (these ones) on my new M90 (RAMPS and Arduino mega). The idea was that I could run these without a fan to cool them, but I hadn't anticipated that they would be so noisy, is this normal?

In particular they have an annoying whine after the motor stops moving (while it is still energised), but they are also louder than my cheap Chinese A4988's during motion, so much so that it completely negates any advantage from not needing a cooling fan.

The sound isn't great in this video (trust me, it's louder in real life) but you can hear the high pitched whine from the A4982 (I can even tune the note by adjusting Vref voltage).
Comparison of A4982 and A4988

I had thought about getting a Smoothiboard someday, but they use this same driver, does anyone know if they have the same problem?

---

Tim

Printers:
Prusa i3 (plywood box frame)
RAMPS 1.4 and Arduino Mega

Mendel 90 (sturdy)
Smoothieware

My RepRap Blog

Are you sure that the extra noise isn't because you are now running the motors at a higher current?

The A4988 and A4982 are very similar chips. Even if the pot settings are the same, the boards may have different value series resistors. I use Duet controllers boards, which use A4982 drivers, and my motors are much quieter than that.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Yes, both boards use 0R05 ohm sense resistors, and in that video they both have Vref set to about 0.58v, which equates to 1.45 amps (the motor is a 17HS19-1684S, rated for 1.68 amps). I've also had the current right down to about 0.5 amps with no appreciable reduction in noise when stationary.

(Good to hear that this isn't a generic A4982 problem though)

Edited 1 time(s). Last edit at 03/08/2015 03:14PM by plankton.

---

Tim

Printers:
Prusa i3 (plywood box frame)
RAMPS 1.4 and Arduino Mega

Mendel 90 (sturdy)
Smoothieware

My RepRap Blog

See a4982 datasheet page7 top right, "low current microstepping" and maybe if you can, try shorting ROSC to ground and see if improves the noise.

@plankton, What supply voltage are you using? If it is more than 12V it will lower the switching frequency.

---

[www.hydraraptor.blogspot.com]

@NoobMan, these drivers already have ROSC shorted to ground, see the schematic here

@nophead, it's a 12v PC supply (a Corsair CX430M, to be specific)

I have 4 of these drivers, all behave the same, however the supplier has sold hundreds of these without problems, which would imply something unique in my setup or environment. I made this video last night of an A4982 on the Y axis of my old printer (A4988's on all other steppers), the whine when stationary is much less pronounced (which I'm putting down to differences in the smaller stepper motor on that printer (Wantai 42BYGHW609)), but you can clearly hear that Y axis moves are much louder than X.

---

Tim

Printers:
Prusa i3 (plywood box frame)
RAMPS 1.4 and Arduino Mega

Mendel 90 (sturdy)
Smoothieware

My RepRap Blog

Sorry my last post was the wrong way round. Your chopping frequency appears to be in the sonic region when it should be ultrasonic. A lower supply voltage or a higher motor voltage make it lower as the on time needs to be longer and off time is fixed. Also higher current will make the frequency lower. I only use 1.25A with the same type of motors.

I use A4988 with ROSC shorted. Shorting ROSC does make the idle noisier but it should make running a bit quieter if anything. I don't think I have tried A4982 but I can't see any reason why it would be noisier than A4988 in low current micro stepping mode. Are you comparing it with A4988 with ROSC not shorted perhaps?

The Y axis is always noisier than X because the base and carriage act as sounding boards.

---

[www.hydraraptor.blogspot.com]

Yes, pretty sure the cheap Chinese A4988's don't have ROSC shorted, and if I look very carefully (through several magnifying glasses) I can just make out a 10K resistor that seems to be labeled R4. I've measured 12.1v from the PSU when I hear the loud whine from motor on idle, and tried vRefs all the way from 0.2 up to 0.65v (0.5amps - 1.625amps), the frequency goes UP as I increase vref (not down), but remains audible (and annoying) whatever voltage I select.

Quote

The Y axis is always noisier than X because the base and carriage act as sounding boards.

True, I think my point was just that these A4982's are louder (than the only other drivers I have to compare them to) whichever printer I put them in, and from the feedback i've got so far I don't think other people are getting this much noise with A4982's either, so I'm just totally puzzled why I am.

I don't understand how the frequency goes up with more current. It's a fixed off time of 30us, so the only thing you can do to get more current is increase the on time to get a higher mark space ratio. That means the total period must be longer, so the frequency lower.

One possibility is the switching is ultrasonic but the two haves of the chip beat with each other and produce a sonic frequency equal to the difference.

You could try shorting ROSC on the A4988 to see if it then as loud as the A4982.

---

[www.hydraraptor.blogspot.com]

I think the resting noise is certainly dependant on some interaction between the 2 motor coils, in that first video you can see me moving the Y carriage in 1mm increments from Pronterface, you can hear the tone alternate between 2 notes with each step (in fact, having just checked that video again I think you can also just make out that I'm turning the pot clockwise as the tone increases)

I wonder if PCB layout affects the amount of interaction between the two halves and hence the noise. I seem to get far less whine with Melzi and A4988 but the switching waveforms for A4982 should be exactly the same. I think the only difference is that it has better short circuit protection.

They both have the same fixed off time so the on time is defined by the current and the motor voltage / supply voltage ratio. I made a Huxley version of the Mendel90 with NEMA14 motors and took it to a local 3D printer group. The younger members complained about a high pitch whine but it was above my hearing. The smaller motors must have lowered the switching frequency.

---

[www.hydraraptor.blogspot.com]

I know this is quite old, but I think I am having a similar problem with a Printrboard that has A4982 and a corexy printer .
It is same as you said upper in the posts, the carriage is acting like a sound box and is somehow amplifying the noise especially from one driver I think, as I set the infill to go alternatively between drivers, one of them is generating much more noise and resonance in the carriage than the other .
The question is how to fix it, I dialled up the VREF to 1.45 V but the drivers were very hot, but the vibrations seemed to be less but not completely gone .
Tried different bushings and bearings for the carriage trying to fix a mechanical problem but no fix .
I think in corexy is even worse as the belts work together so they can create also some kind of resonance between the motors which makes things only worse ...
I've order a chinese RAMPS as that is modular with A4988 s with radiators, I am willing also to try the DRV8825 or whichever driver go on the RAMPS and are quieter, if you have some ideas let me know .

Steps per mm is around 81, this is with GT2 and 20 teeth pulley, can somebody tell what microstepping are hardwired in the Printrboard ?

Quote
bbogdanmircea
I know this is quite old, but I think I am having a similar problem with a Printrboard that has A4982 and a corexy printer .
It is same as you said upper in the posts, the carriage is acting like a sound box and is somehow amplifying the noise especially from one driver I think, as I set the infill to go alternatively between drivers, one of them is generating much more noise and resonance in the carriage than the other .
The question is how to fix it, I dialled up the VREF to 1.45 V but the drivers were very hot, but the vibrations seemed to be less but not completely gone .
Tried different bushings and bearings for the carriage trying to fix a mechanical problem but no fix .
I think in corexy is even worse as the belts work together so they can create also some kind of resonance between the motors which makes things only worse ...
I've order a chinese RAMPS as that is modular with A4988 s with radiators, I am willing also to try the DRV8825 or whichever driver go on the RAMPS and are quieter, if you have some ideas let me know .

Steps per mm is around 81, this is with GT2 and 20 teeth pulley, can somebody tell what microstepping are hardwired in the Printrboard ?

I presume you mean that the printer is noisy when the motors are moving, not when they are stationary. Motion noise is caused by the motors being energised in discrete microsteps. Typically there are resonances in belts and sometimes other components that amplify the noise at certain speeds. So you can do two things to reduce the noise:

1. Use drivers with higher microstepping. The Duet WiFi that I co-designed supports up to 256x using TMC2660 drivers. But you need only go up to 64x to get rid of most of the noise. It's also possible to get higher microstepping drivers that plug into RAMPS, such as TMC2100.

2. Reduce the resonances, or shift them to frequencies that are not as easily triggered. Unless the belts are already very tight, tightening them may help.

Edited 2 time(s). Last edit at 03/13/2017 11:59AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

https://www.youtube.com/watch?v=S9Hk6voyhI4

Here is my channel with the movies of my printer, it is definitely a resonance especially when one of the motors is doing the infill, and it is amplified by the carriage, if I put a screwdriver I can feel it vibrating .
The problem is how to fix or counteract this ? The belts are tightened quite ok, friction is very low so something is happening .