Board Crashing / GLCD gobblydigook?!

Machine: RAMPS/Arduino from Moshi

Board: New C3D

Firmware: Smoothie - as per C3D Firmware

Crashing / Gobbledigook on LCD

Hey all,

First, let me say that the C3D upgrade from Ramps (I’ll do a build dairy later) was lovely - scary, but so nice when I got my head around it!

Now, at the moment I’m having an issue with the LCD - it’s come up ‘right’ twice, but the rest of the time it shows the mumbo jumbo as per the picture. I’ve taken out and replace the ribbon and the GLC adapter multiple times - I’ve even ‘removed the cartridge and blown on it’ ;-). Nope, still gubbins.

The other issue, and I don’t know if its a LB or C3D issue, is that I get ‘crashing’. Par example:

I’ll be setting a line up for a ramp test. It homes when I ask it to run - notice it doesn’t run. Then it says that it’s busy, and does nothing. Sometimes even when I power on off the C3D card and close/reopen LB. I’ve moved it from X/Y end stops it case that was it, but no. Restart Mac and hope. Leave card unplugged and hope.

The other thing is that is will frequently get to the origin on the job, then just stop dead. It’ll fire the laser at where it wants to start, then just shut down. If I run it again there’s a 50/50 chance of the job being done, but in the meantime it’s forgotten it’s position, and will crash in to the gantry end. If I remove a load of cut layers and go down to a bare minimum then it will sometimes start.

In short, I’m a td perplexed and hope that someone out there can assist!

FYI - running a Mac! , most up to date version of LB, have made DOUBLE sure that the most up to date firmware is installed by reinstalling ‘just in case’!!

Minor Update
If I unplug the board (power), then plug back in enough times I can get the LCD to load, but then in a minute or so it reboots itself automatically.

I have taken out x & y, and laser modules, leaving a ‘blank’ board other than USB and LCD. Still happens. This may sound stupid, but when I ‘pull’ the board from its rightful place and put it on a flat surface it works better for longer…

Minor Update 2

There seems to be something specifically wrong around the comms. I run a Mac with Parallels for Windows use. I decided to flash the firmware with Windows ‘just in case’. That was fine. However when I connected the board to the computer I got multiple flashes of ‘do you want to connect to you Windows 10 or Mac?’ windows. I’d select, it would come up again, which indicates that the board is dropping the connection to the computer and reestablishing it, dropping, reestablish etc. I’m using a brand new C3D USB cable, and I’ve also used a cable from my printer, which worked on the RAMPS setup previously, and the same thing happens.

Minor Update 3 - last one for tonight (UK!)
I diidn’t know it, but the Laserboard can power the steppers wiithout the K40 being turned on. And as it happens, the crashes take longer to occur. It still can’t ‘do’ anything, and it crashes as soon as you point it in the direction of ‘lasering’ something, even if the laser ‘maschine’ is off, and comms go down, but perhaps some breadcrumbs? I imagine that when the machine is on it causes some conflict between what it thinks it wants to do, and what I’m asking it to do, hence the longer time between crashes as when the ‘machine’ is on I’m clicking buttons, sending to home etc.

Hi Jonathan… Wow, there’s a lot going on. Glad to hear you’ve got things cleaned up since we last saw the RAMPS board and some questionable soldering to the ribbon socket. :slight_smile:

Thanks for posting the photo of the LaserBoard. The first thing I want to check on is if there is any space between the board and the white piece of metal underneath it. Are there spacers or is the back of the board directly touching the metal?

Some of the behavior you mentioned may be related to grounding issues. Can you confirm your machine is grounded properly? It should be plugged into a grounded outlet, and there is typically a grounding lug on the back of the laser, though you may need to scrape the paint off the case for it to make good contact.

You might also try to unplug the GLCD completely and test things again. If for some reason the GLCD itself is causing problems, disconnecting it to take it out of the mix may help diagnose things.

Hi Pete,

Yes, little plastic risers installed, machine is grounded through the original laser plug and the same occurs with and without the LCD screen. I’ve even mounted the board on to a wooden plinth to ensure it’s well insulated, and mounted the plinth outside of the case.

After a little more playing this morning I’ve noticed that when it crashes LED L2 will either go out, or stay on permanently - it does both (although not at the same time :laughing:). Again, it will frame etc, home, but crashes as soon as it has to receive complex g-code like cutting a line of test text. When it recovers somewhat the LED returns to normal, albeit typically out of sync, which offends my OCD! Although it still won’t do anything when the LEDs are in the recovered state.

Edit - I’ve installed a new ground, just for the sake of completeness and ‘belt and braces’. Still nada.
The board keeps dropping all comms - even when idle.

Edit 2:
I’ve saved a couple of files to the SD to see if it purely is a comms issue. When cutting a 10mm circle from the SD it completed it OK. However when I ran the ‘test’ text file it got to the start, paused for about 2 mins, competed the ‘t’, paused again and, well you can see the photos. It’s just completed the ‘e’, at 8 mins 48 seconds run time.

!

Anyone from C3D help? I’ve installed LB on Windows, and the same thing happens - crash, restore comms etc…

It looks like you have covered quite a lot of possibilities there.

Both your issues, at first glance, would be explained by electrical interference/ noise inside the cabinet of the K40. For example, the fact that the screen does work proves it is not defective, and that it scrambles sometimes means there is noise/ interference nearby. I suspect it might be the LPSU.

Can you try to run the job with the laser disabled (laser enable switch turned off)?

How is the C3D power brick plugged in - for example, is it on a power strip with other items?

For completeness, you may also want to review this if you haven’t already: https://cohesion3d.com/knowledgebase/usb-dropouts-and-brownouts/

As for the many posts on Fb and the LB forum, your previous post here was literally 23 hours ago from when I am typing this response right now. Quality support and responses require consideration and time, we do not do the instant “shoot from the hip” that you will find elsewhere. All that you are doing is stressing us out, at a time when many of us are barely hanging on as it is. Please relax and be patient.

1 Like

Hi Ray,

Thanks so much for the reply. There was no intent to cause you stress, in fact quite the opposite. I wasn’t trying to circumvent you per se, but I’m aware that things are not always ‘business as usual’ at the moment. I did drop an email, and I had no direct reply, and there wasn’t a ‘please hold’ kinda message on here. I placed an order for LB software with you and while I expect delays, and I have no problem because LB extended my trial, after 4 days or so I haven’t received it. I’m not complaining, just trying to point out that I figured that you guys must be very busy, and while I’m trying to keep my own business afloat (hence the C3D card), you guys are doing the same as well as running a print farm. Rather than keep on ‘nagging’ I thought I’d ask elsewhere. I promise, I’m not an a-hole, I’m just trying to extrapolate from the only data I have.

In reply to your question, yes, I have done that. The card is currently on an isolated wooden board outside the cabinet (just), and I have tried your suggestion, admittedly by mistake initially, and ran the board without the machine being on (so no laser PSU), as well as with the machine on but the laser off. I have run it with and without the x and y stepper motors connected, and with and without the wiring connected to the laser PSU. I have also done the above without the LCD connected. The LCD is the canary in the coal mine (IMHO) – it isn’t the reason for the crash, but when it goes nuts then I know that a hard-reset needs to happen!

I have also tried running the card entirely outside of the machine, on an insulated bench 6ft from the machine, and it does still crash. I just plugged it in now, and with only the board plugged in to the wall, nothing else, the screen came up as per the previous picture. I restarted the board three times before it came up. I ran a g-code file from the SD which was just the word ‘text’ as fill & line, saved from LB, and it came to a halt in the same manner as my previous post. I produced a new ‘test’ text file in LB with layers set to defaults to ensure there was nothing wonky there, and the same thing happened, both with and without the K40 being turned on.

I have run the board on a separate wall socket, as well as on on a strip, with no perceivable difference.

IMHO the comms dropping is not the cause, but a symptom – something happens on the board, it resets itself and comms drop/come back up.

I am happy to take videos etc, more detailed photographs for you if you’d like.

I apologise for the delay – I’m not being a hypocrite, unfortunately my wife just lost a family member to the virus.

All the best, and thanks for ‘keeping on’!

If your LightBurn license is part of a physical parts order, then that will be automatically sent once the order is shipped.

I have not yet seen a board itself have comms issues. I have seen plenty of comms issues due to other factors, however.

Do you have any appliances like a compressor or fridge nearby/ on the same mains circuit? Is the mains side of the power brick cable firmly plugged in? Can you put a multimeter on the output (DC Barrel Jack end) of the psu and see that you are getting 24v?

Is there another computer you can try?

Ultimately, if you’ve got a board outside the laser and with nothing plugged into it dropping comms, that doesn’t leave a lot of things to look at. Stumped would be a good word.

My condolences on your family’s loss.

Hi Ray,

Thanks for the thoughts.

I thought I’d go about it a different way, and decided to change the SD - and it made a BIG difference. At least until I unplugged the card. I know this will sound weird, but if I format the card, replace the firmware and the config.txt, let it boot up etc then it will run OK (ish) for a bit. There are still issues, so it keeps forgetting where it is - if you home then go-to-origin then it’s a bit better, but if you home then ‘print’ then it just seems to cut the y axis in half (not actually half, but you get the idea),for some reason and hit the gantry. These are all using absolute coordinates. If I run the same job in the ‘right’ way, then it’s off by a little bit every time.

When it crashes, it still crashes, but if I reformat the SD card and reinstall the firmware it gets better again.

I’m in an industrial building, with industrial grade power etc, so I’m pretty happy there, but I have turned off everything else in the area to run the tests, but it doesn’t make a difference.

I’ve tried running a multimeter over the power brick, but as soon as it detect the mm it must have a detection circuit for short circuiting, and it powers itself down.

Edit, playing with the stepper motor power and default seek rate sorted the gantry issues - lets see if it’s done more than that…

Well I’ll be a monkey’s uncle. It seems like it was the SD card. The ruddy SD card. It’s now been fine for about an hour… fingers crossed!

Ah! That was my next suggestion. I was testing trying to pull out and push in an SD card while a job was running via USB to see what would happen. It seemed like if I was sending via USB I could pull the card after the job started, and the job would run fine as long as the power to the board remained constant. The SD card really needs to be inserted at boot.

Obviously pulling the SD card when running a job from the SD card is a terrible idea. (I tried it anyway, you can guess what happens.)

Did you end up replacing the SD card with a different one? Did that solve the problem?

It seems to have, yes. I replaced it with a 64Gb (der, bad move), but formatt!ed it with Mac to a lower capacity on FAT.

That, plus changing the power to the Y and max acceleration has made a massive difference.

I’m having massive banding issues though, really consistently spaced (about 5mm apart) the power surges it seems. I’ll see if I can post a pic!

My gut, and some very casual observation says that it’s going over the same lines twice at these intervals.

Hmmm. Looks like more than one thing, but I think I get it…

I replaced the board inside the machine, and turned it on and it started getting…screwy again. Replaced the firmware and config.txt and it was fine, then it crashed on a piece of work I’ve already done 5 off today.

Took the board out, reflashed etc… all is well.

Placed it inside with the laser off… it was OK. Turn the laser on, and it crashed.

So, and tell me what you think here… there’s a problem with the LPSU that is somehow corrupting the data on the SD card…

Replace LPSU?

To confirm - are you saying that you can put the board inside the laser, and run the same job the same way… if the LPSU is off (I guess you can disable it by the panel test switch, yeah?) then the job runs successfully, but if the LPSU is enabled then you have the cutout/ board crash issue?

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.