Prusa i3 occasional weird problem during homing

Hi all,
I purchased a Sunhokey Prusa i3 kit, duly assembled it and it has been working well for some months now (although the MKS Gen board fried itself a couple of months ago, replaced it).

I have a weird problem, when I press the "Home" button in Repetier.

Normal operation
1. X axis moves towards home position
2. X axis triggers the endstop for home position and stops moving

Abnormal operation
1. X axis moves towards home position
2. X axis triggers the endstop for home position and stops moving, the printer freezes, comms is lost and I have to reconnect in repetier.

It only does this sometimes (and only on the X axis), I have tried replacing the endstop and checked the wiring. The X axis moves easily when not powered, has no binding of grabbing during movement.

Does anyone have any ideas that might point me in the right direction to identify the fault?

Thanks in advance

I´d look in the y-driver/stepper direction. It seems the 5V regulator gets a "brown out" the very moment the y-stepper would normally start.
Reduce y-acceleration and/or Vref of the driver if possible.
-Olaf

Hi Olaf,
Very good points Olaf, both of them tend to point towards the solution I ended up implementing.

In the past I would normally have access to some very expensive test equipment (I'm an electronics R&D engineer) so I could troubleshoot the fault. I suspected the fault was related to the 5V supply but couldn't prove it without an oscilloscope.

It looks like I have fixed the problem though. I ended up putting a 3300uF capacitor across the 5V supply rail. I actually had one laying around from my R/C hobby (used for my flight controller to stabilise the supply) and just plugged it into one of the 5V headers on the MKS GEN board. I can't believe how well it worked, I have not had any weird issues since with resets occurring. I am starting to trust my MKS Gen board now, I always thought it was rubbish but have not had one failure since adding the capacitor.
I purchased mine from HobbyKing.

In the past, I could click the "Home" button in Repetier and maybe 6 or 7 times out of 10 the board would reset. Also when starting a print, it often took me 3 or 4 goes before it would start properly. Once it started it always worked ok.
Now I can open Repetier and click on "Home" 20 times and the board does exactly that, goes home and works. All my prints start without incident. I have not had a single failure since adding the capacitor.

Hope this helps anyone else that may have a similar problem with their board resetting.

The MKS board seems to have the same crappy 5V regulator as the Mega boards?
Have you considered the inrush current, while power on? Especially low-ESR capacitors are merciless...
Might be healthier for the regulator to use a precharge-circuit in long terms?

-Olaf
PS I love it, when I can talk "business" without the need to explain every term

Hi Olaf,

Well it looks like my excitement has been short lived. It has started doing it again about 30 minutes after writing the post claiming it was working! But I am aware this is how things work, do something, looks like it's fixed, tell the world, then it falls in your face!
I tried adding several uF of capacitance directly at various locations on the 5V rail across the board near the processor and also near the stepper drivers without success.

I had a spare, new MKS Gen board still sealed in its antistatic bag, so I pulled it out, flashed it, installed it, and have printed about 7 jobs with a total print time now of about 12 hours without a failure. I have not incorporated any of the mods that I claimed on the previous board, it is stock standard.
When I have a little more time, I will go over the previous board with my microscope, maybe there is a faulty or dry solder joint somewhere.

Both boards use an ST L5970 switching reg (which I have been using myself in some commercial designs for over 10 years) with a 150uH inductor + 100uF cap. The board that is misbehaving is a slightly different hardware revision. Both baords have the same regulator, but the working one has two inductors, one for the switching regulator and one for additional main supply rail filtering (220uH + 100uF cap) prior to the switching regulator (as far as I can tell).

Almost every day I think about redesigning this board, some weeks back I even started an Altium project doing just that (after I wrote a short product specification document). It is something I will probably end up doing but as always there is never enough time as always too busy doing the jobs that pay the bills.

Being able to talk "business" is such a wonderful thing. I always feel I have to try to explain every detail (even though I often don't).

One feature I would like, completely not related to the hardware is some way of tracking the quantity of filament used. The slicer always calculated the use, but it would be nice if the software (such as Repetier Host) could log the length or volumetric qty plus maybe the running time. Yes I know of the logistical issues when changing the filament, or wasting pieces or if I cut off some manually, etc. However the software could also allow for this to be entered.

Anyway, I will continue to chase this bug and provide an update when I get the change to do just that.

Regards

Even the biggest cap doesn´t work, when the regulator is too weak to feed it.
If you happen to have one of the odd graphic LCD-boards, the background LED draws more than is healthy for the regulator.
Try the other board with an external 5V UBEC, if possible.
I´m not a fan of : "doesn´t work => change the board " solutions.
-Olaf

Hi again Olaf,

The setup is using a 4 line by 20 character LCD with LED backlighting.
The entire setup powered is using 170mA from the 12V supply in idle mode (no steppers or heaters switched on). It drops to about 160mA with USB plugged in.
I doubt the regulator is having trouble with that kind of current draw, it should be good for about 1 amp.

The board that is giving me troubles is made by "Keyes", but the board that is working well is made my an unknown company.

Thanks for the idea, when I get a chance, I will try to bypass the on-board regulator with a good quality 5V UBEC I have laying around (I have a few different ones). I will just have to lift the output inductor and patch it in there.
I did have a quick scan over the board this morning with my jewelers loupe and didn't see anything obvious with the soldering.

Regards
gjt211