My Meltzi kept locking up and I couldn't figure out why.

My Meltzi kept locking up and I couldn’t figure out why. Turned out when the heaters were running on the printer there was so much load on the power supply 12v rail it was sagging and noisy.

Rather than swap out the 500W PC PSU which powers the printer (which would be a bit of a nightmare as its connected all over the place and physically difficult to remove without stripping the printer) I did a bit of a hack. There’s a jumper in the middle Melzi board that switches all its logic components (including its Arduino) from being powered off the Melzi USB connection or the boards built in 5v regulator which is running off the 12v line. My issue is of course the 12v line is garbage and the USB is hooked to the printers Raspberry Pi running OctoPrint/OctoPi and the Pi can’t output enough current down the USB to power the Meltzi when the camera is running and its also encoding video.

So what I did was remove the jumper completely and hook the 5v rail on the PC power supply directly into the Melzi.

Voila, all rock solid stable :slight_smile:
OK the 12v rail is still noisy but this not so much of an issue as the heaters and stepper motors don’t care quite so much.

Check out the 12v rail sag…
The red wire you can see going directly into the middle of the Melzi is the 5v rail hack.

Nicely done! I’m surprised the Melzi isn’t set up by default to take the 5V from the 5V supply of the PC supply, on may units in order to reach the rated capacity of the PSU You have to draw it from both the 12V and 5V supply lines.

I guess the idea was to make it so you only need a single voltage rail so you could use a standard 12v supply. I think I read the Melzi may even go up to 24v which would make sense as you wouldn’t need so much current and if you have 24v its more likely you will only have a single rail because you wouldn’t be using a PC supply.

Nice post. I may start doing that for my own printers: make use of the conditioned 5v lines coming from the psu.

Maybe I’m missing something, but a 0.4V sag doesn’t seem so terrible. That is comfortably within 5%. What is the no-load voltage? At any rate, it should cleanly regulate to 5V.

@Ashley_Webster yes a more powerful PSU would have probably solved it also. Mine was 500W

@Jeff_DeMaagd yes that was my thoughts as well. I think the issue was noise on the voltage rail as the PSU was unable to smooth it.

I’d still swap out the PS for a 1500 watt overkill with major cooling. Lesson here is always build in a swappable component fashion.