Y axis issue?  

Re: Y axis issue?

Hi guys

Just finished a 14+ hour overnight print with firmware version 2.1.0. All OK.

Now I know the pin configuration of the motor current limiting reference voltage, I am not surprised that the print was good.

I have just downloaded the latest firmware source (2.2.1) and increased the reference voltage for the XY motors and am uploading to the RAMBo.

I will start another 15 hour print, but I have to go out. Hopefully when I return, all will still be OK, but I will be checking regularly via my OctoPrint camera and can cancel the print if necessary.

Peter

Re: Y axis issue? [Solved]

Latest Update...

The print ran for about 20 hours and completed successfully.

A little bit more theory:

There are settings within the firmware used to set a reference voltage to control the maximum current of the stepper drivers. The settings start off as numbers which don't really mean very much.

Firmware V2.1 had these settings as {400, 550, 650} and V2.2.x as {450, 450, 270}. Now normally within the firmware, the order is XY (the X and Y motors use the same value), Z, E. However in the I3 firmware, the pins have been defined incorrectly with the XY and E pins being swapped over, so the order is E, Z, XY.

In order to make the pins quieter, the last value (650) was reduced to 270, lowering the current available to the motors quite significantly.

However, if the motors do not have sufficient current, then shipping can occur.

I have the XY value set now at 400.

As I mentioned, these numbers are quire meaningless, until you know what happens within the firmware.

A PWM signal is output from these pins. The numbers indirectly determine the mark/space ratio of the PWM output. The formula is n/2000 * 255.

A divisor of 2000 has been chosen (no idea why) and the resulting fraction is multiplied by 255 - the number of mark/space ratios available.

As the AT Megs runs at 5 Volts and an analogue reference voltage is required, the PWM output is fed through a DAC circuit, generating a reference voltage between 0 and 5 Volts. However, the DAC also contains attenuation circuitry, and the output from it is actually in the range 0 to 1.67 Volts.

The stepper motors on this printer can handle a maximum of 1 Amp and a reference voltage of 0.8 Volts is required for the stepper drivers to limit the current to 1 Amp (the motors will allow over 2 Amps to be passed at 12 Volts, so current limiting is necessary).

Back to the firmware. 960 generates a reference voltage of 0.8V. The original Vref was 0.54V (0.67 A motor current) which was reduced to 0.225V (0.28A) and I now have it set to 0.33V (0.42A).

Apparently, noise from the motors depends on how hard they are being driven, therefore less current = less noise = more potential for skipping.

One final thought: for most printers, Josef's settings will be fine. The manufacturing tolerances of the components will occasionally combine to produce what I have experienced.

Peter

Re: Y axis issue?

It never happened to me so far (but I dont do such long prints). With original current settings my printer was very noisy. Even last change from 260 (beta firmware) to 270 is quite noticeable on my printer in terms of noise.

Re: Y axis issue?

Glad you got it Peter.
So updating the firmware a good thing to do or wait? I'm on 2.2 and everything is going just had that on oops but since then did other prints and they came out fine.

Re: Y axis issue?

Hi Roman

Yes, on the initial (beta) 2.2.0 there were some people reporting skipping, hence the increase to 270. As mentioned, I think that 270 is OK for most, but there will be some where skipping still occurs.

And yes, my Y axis was noisier than the X axis at 270; they are both noisier at 400, but not as loud as the V2.1.0 value of 650.

If 260 (or less) works for you, then why not make the change yourself.

I am currently printing Hogwarts school at approximately 1:80 scale. I have some very big and long prints to do. So far I am about three quarters the way through "Bell Towers", which is almost 70cm tall

My clone which uses the Melzi V2 board is set to slightly under 0.8V, but the stepper motors (all 5 are the same) are capable of handling just over 2A and at 1A are extremely quiet, so I think that the noise generated is actually a proportionate to the power handling capabilities of the motors and the current fed to them.

Peter

Re: Y axis issue?

Hi Scott

2.2.0 and 2.2.1 both have the same Vref for the motor current, so there should be no problem for you to update.

However, it you do experience the problem again, it is worth baring in mind that you can amend the firmware to slightly higher values.

I now have 3 minor firmware changes that I have to put into place with every release (not that there should be many more releases...).

Peter

Re: Y axis issue?

Is that Melzi V2 board OK? I am going to build second printer (based on i3 but different construction) and still dont know what board to use (I dont want typical ramps+arduino). So far I tend to buy Megatronics 2.0 + DRV8825.

Sry for OT.

Re: Y axis issue?

Hi Roman

It's not even a true Melzi - a Chinese clone. It has the same stepper drivers as the RAMBo. It does have trimmer pots which I find better than having to change the firmware to get the motors working correctly.

I's been working for over 9 months now without any issues (with over 6 months constant use). The only thing I did do was to mount a 12cm fan over the stepper drivers.

I would say that it's solid. And I do like the all-in-one construction.

The only issue I had was on delivery there was a dry joint on one solder pad of the E motor which was easily rectified. I was probably caused by me tightening the screw terminals too tightly.

Unfortunately, I have not dared fiddle with the firmware as I don't know what settings the Chinese use and it is quite impossible to find out, so it is running on the original firmware; the motors have quire low speed limits even though they are big motors.

Peter

Re: Y axis issue?

thx for info 😉