We’ve just had our second SKR Pro board fail today. First one failed maybe a year ago. Had a replacement, so we’re back on line, but now have 2 failed boards. Can you repair them? We’re ordering another to have on hand. Since they’re only $50 or so, hardly seems worth it to attempt repair. Shipping back and forth would probably cost about that much.
3 Likes
Symptom of both failures was TFT screen went blank.
Fix for both failures: Replaced main board, reused same drivers, flashed SKR firmware.
I want to re-emphasize that the SD micro card to flash the SKR Pro V1.2 board can’t be more than 32GB in capacity, and must be in FAT32 format. This is not noted in the SKR setup instructions on the V1 website. I found it somewhere in this forum.
After talking to the guys in the shop, found out they tend to move the router gantry rapidly with the power off to the point that the back EMF from the steppers lights up the display. I have told them not to do that any more. Previously, I didn’t know that could cause a failure.
2 Likes
This is 100% why you were killing SKR boards, and would probably have been killing most all of the controllers that cost less than $500.
There are “brake boards” that are used on 3D printers, but what that does is short the coils when power is off. The effect is it gets really difficult to move the axis by hand. Better practice is what you’ve done in changing the workflow to not let back EMF risk damage to the board.
Is it possible in your workflow to jog the machine to a parking spot using a gcode macro?
1 Like
Our jobs always return to 0,0, but rather than leaving the router where it stops, then removing the work from behind the router, they would shove the router gantry down the table to pull the work out from the front.
You might want to add a script at the end of your gcode file to move the gantry out of the way when the job finishes.
Or teach your guys to use the equipment properly.
The script would probably be easier and more effective.
1 Like
Someone will always find fault with any or all of those suggestions 
but if a end of file parking to where they want the gantry hard code it to be entered by the software automatically on file completion.
If you have to move it manually, don’t go faster than the speed where the screen first lights up. That should be safe.
I agree with Bartman that making the gcode move it at the end of each job will make them happier and make it easier for anyone to follow protocol.
1 Like
Ask them where they want the gantry parked at the end of the job. I’d expect that to be a location where they can get the finished piece out and load the next “blank” in. The start of the new job can re-home the machine.
1 Like
That is exactly what i was trying to say but just spwed 
1 Like
I blew up 2 arduinos this way!!!
1 Like
So now we have had 2 more main board failures in less than 2 weeks. The last replacement PCB ran for one day before failure today, and one of the main board failures has also taken out one of the TFT display boards such that it kills the 5V and 12V LEDs don’t light on the main board. The other TFT still appears ok. Of the 4 main board failures, 2 give us a blank TFT, one has a blown U2 DC/DC buck converter, and one constantly reports no printer attached. Our first SKR Pro ran for over a year, and now we’ve had 3 failures in 2 weeks.
The way electrical overstress works is that the damage accumulates until a failure occurs. The currently failing boards were likely mortally wounded earlier, but soldiered on until they ultimately failed.
So boards that weren’t outright blown by a super rapid move were still taking damage the whole time. If it lights up the display while you’re moving it, you’re likely stressing it slowly as you do.
Some boards can last a really long time with some latent damage.
1 Like
Are the boards that failed on the same machine, or is this multiple machines that have failed?
Are you re-using the drivers each time you replace the board(s)? I’m not 100% sure, but a faulty driver MIGHT be causing a short that kills the DC power converters.
You may want to check the replacement boards for debris. I had a replacement board from BTT that had a piece of loose solder bridging some pins on an IC, causing all sorts of weird issues. I used a sharp pick to remove it (with power off), and was able to recover the board for use. For that matter, debris from your shop environment may be collecting on the boards as well.
1 Like
These failures are on the same machine, the last 2 after only a few hours of use. The steppers are not being moved with power off. Is the Jackpot more reliable? I’ve lost all confidence in Big Tree Tech.
1 Like
Yeah, I was unimpressed with their customer service, and their quality control seems to have slipped (IMO)
1 Like
I’m not sure, but in general, the Jackpot does seem reliable according to all reports. (I’m still in the SKR camp, so can’t comment directly).
The firmware and workflow commands are somewhat different (FluidNC vs Marlin), although there are lots of similarities. There will be a bit of a learning curve (reportedly a fairly small curve). The interface is also different (wifi connected phone or tablet vs.onboard LCD screen).
Most of the issues with Jackpot from what I’ve seen in the Forum seem to be around poor connection to wifi, often related to the placement of internal wiring or distance from the wireless access point (WAP or Router).
If you’ve lost confidence in SKR, then Jackpot may be the way to go.
1 Like
Can you post some high-resolution pictures of your boards, front and back?
This is very concerning that you don’t seem to be able to get usable boards new stock.
1 Like
I’m wondering if there is an ESD issue during handling during installation or due to static buildup in your dust collection.
1 Like