Yes, I used a soft UTP cable with stranded wire. by changing the attachment of the cable and bowden, it is under minimal stress. It works absolutely perfectly. Twisted wires are an advantage in any communication. And yes my printer now uses CAN-BUS
So my printer started doing the random disconnect thing again and could not reset the toolhead board via software. I finally took the time to swap the cable out for a proper unit of a cable (3DO CAN/USB) great quality imo just very thick. I got the one with USB because Iām going to throw a cartographer probe in there at some point. They have a CAN only 4 conductor version.
Iām getting pretty fed up with this, so I removed the 4 pin connector and soldered the wires straight to the board like Exxon. I only need it to last long enough to print the EBB42 adapter he made and the new screen mount because with my findings on the toolhead board I have zero faith in its reliability, so I will be putting a Manta M5P in it.
I believe the fab that made these either has a bad stencil or just straight up did not use enough solder paste. Iāve seen so many posts of connectors just coming off and failing at the solder rather than the pads. This happened to 3 connectors on mine, and they all had the same failure at the solder. If thatās how the connectors are then one could only imagine the poor connection the RP2040 and other BGA chips have.
TLDR:
Ill post again if this lasts but the tldr is that I donāt think there are any klipper SW or FW issues going on here. I think the PCBs were poorly made and are causing intermittent connections. I also checked continuity of the original cable and moved it around and saw no issues.
Very strange, upgraded mine to latest klipper this weekend, and everything is working flawless⦠the only disconnects I got have been caused by me increasing the extruder run_current (works ok for few minutes before random disconnects) ā¦
To me that would be further indication that the boards are made poorly. Increasing the current is probably putting too much heat/stress on a poor solder joint, causing the issue to show up.
Since soldering the new cable directly to the board, I have another 10 hours or so of printing without issue.
I donāt think so. The solder joints shouldnāt be a problem.
Looking at the posted pictures without a schematics, I make a harsh guess.
Generating 5V from 24V:
Some switching DC converter.
Generating 3.3V from 5V:
LDO derivate of LM1117.
Incoming 24V and buffer capacitors (donāt know, if ābuffer capacitorsā is the right naming):
You may increase the buffer capacity for the main supply of 24V. Easy to do.
The fact the decoupling capacitors are oddly offset from the chips too. If you look at where the stepper driver is on the front of the board (near the power/serial connector) and then look at the back of the board everything is funky.
Especially when there is still real estate on the front of the board.
The decoupling caps should be close to the pins.
I canāt even begin to visually trace out the power paths through a picture. But itās plain to see this board wasnāt designed very well.
The RP2040 has the same issue, the capacitors are oddly offset when they had plenty of room?
Zero clue what they were thinking.
The MCU decopling capacitors are on the back side of the board, there are plenty of them. I did not trace them to the MCUpins, though
PCB design 101 is place your capacitors as close as physically possible to the power pins of the chip to minimize the effect of trace impedance. Too much trace resistance/impedance and youāve created low pass filter which is the opposite of what you want.
You also want a low impedance path to ground directly beside the power pins so any noise is shunted at the lost possible moment, likewise the first thing a power pin should āseeā is a decoupling/bypass/however you want to look at it capacitor which acts as a quick reservoir of power to buffer the chips power needs.
āAnother aspect of the power supply design are the decoupling capacitors required for RP2040. These provide two basic
functions. Firstly, they filter out power supply noise, and secondly, provide a local supply of charge that the circuits
inside RP2040 can use at short notice. This prevents the voltage level in the immediate vicinity from dropping too much
when the current demand suddenly increases. Because, of this, it is important to place decoupling close to the power
pinsā - Page 8 - Hardware Design with the RP2040, even though thatās applicable to pretty much any IC.
They even provide a handy dandy photo to show what they mean.
Thatās what kinda of makes me scratch my head at that board, thatās SUCH a common basic thing they seemingly ignored for whatever reason.
I can have the nicest, lowest ESR capacitors but if theyāre on the otherside of the board from the power pins it kind of defeats the purpose.
I am pretty sure you a saying the right things. Just would like to clarify that under āother side of the boardā I meant not the farther corner of the board, but rather a via right near the RP2040 power pins leading to back side of the board, where the capacitors are. The lead to the capacitor is extremely short.
Dear Gents , from the user perspective view . Is it possible even if only for failure checks to connect USB cable from Klipper MCU to the toolhead. Or as quick solution use USB cable . Does the USB port of the toolhead behave different in terms of signal stability ??
Adittionaly any short description how to flash potentially the toolhead ⦠sk1 mainboard uses Armbian and can be used via putty or something else. Looking forward to interviewing reply.
please knock if there is a solution for westernmen305@gmail.com
@westernmen :
And what about your contribution?
You just come in and request a signal when the topicās issue is solved.
A forum like this lives from helping each other. Not from waiting for solutions.
You may explain the issues you have, what is different to the OP, what you tried yourself to solve it, etc
Guys, I have good news, I managed to return to the initial state, I just formatted the EMMC completely, then I recorded the image and everything worked, although before that I flashed the head board many times, if you show me how I can make a backup copy
Sorry, Iām not familiar with Clipper and such sites, and I also live in Ukraine, itās become very difficult to live here, Iāll try
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.





