6 pin SOT-23 is actually not hard to solder or remove. IC4 on a Rumba is an LT1933ES6 switchmode regulator, you can order it from Digikey or Mouser. Bear in mind, it may not be the only “fried” part, but it’s a start.
1 Like
Lame indeed. The part that blew (if it’s a “real” Analog Devices part) is only rated for about 600mA, and it’s supposed to have an active clamp for over current protection according to the datasheet… which obviously didn’t work.
1 Like
After posting late last nite, I wired up an external 5v regulator and attached it to the onboard 5v jumpers. I was able to power the board with this but none of the steppers would work. Serial COMs to Repetier host worked and the LCD interface worked. I checked 12V supply across the board and everything was as expected. I de-soldered the IC in question and damaged the pad in the process so replacing it kind for kind is out of the question.
I was tired and frusterated so i did a dumb thing; what I think happened was I put a 8825 driver in with the power on and it somehow tripped something. I also could have put it in backwards/upside down, but looking at the pins it would have sent 12V through the stepper before tripping which shouldn’t be a problem?
I tried every axis and every driver and every stepper and nothing would work through the LCD.
I ordered a new set of DRV8825s which will be here tomorrow and a friend is letting me borrow his RAMPS 1.4 with A4988s for testing the steppers.
Whoops. Turn the power off before you mess with stuff!
Did the fried part blow open or did it short the +5V net to Vin? If the latter - toss the board, even if it’s partially working the stresses could result in future failures as you put more operational time into the board… not worth the headaches, especially when it’s used to move sharp pointy things
1 Like
The damage occurred at pin 5, or Vin. It put bulge on the top of the IC. The top layer of the IC was delaminated and scraped off with a very slight touch.
This is an import board with modified components compared to the original RUMBA+ and fysetc (FYSTECH) has no schematics available that I’ve seen.
I assume you have a DMM, but check for continuity between pins 6 (SW) and 5 (Vin). If it’s shorted, this means Vin connected to the +5V net… anything connected to +5V is suspect (pretty much most of the board).
1 Like
I’ll check the 12V short later tonight
I was able to boot up the board on USB Power and then later using a standalone 5V mosfet regulator fed from Vin with the output applied to the 5V jumpers. Repetier Host was able to connect. The LCD interface was working correctly. 12V from the power supply appears in the right places.
I was having issues with Arduino on my Windows 10 machine so I wasn’t able to try reuploading Marlin to the 2560 but I’ll try that again tonight.
Other than that, I’m a little confused as to why the stepper drivers and steppers won’t work. I need to check the DRV8825s tonight.
This means the part died open (good thing), otherwise you’d see 12V on the 5V net - perhaps the dead regulator killed the drivers… who knows. Probably don’t need to check for the short with that info. Did you get your replacement driver modules?
1 Like
My DMM shows continuity between Vin and SW pin 6, but checking the voltage, i show a differential of 7 volts.
I tried to reprogram through the Arduino IDE and just kept getting timeouts. Repetier Host would ‘connect’ to the board on COM4 but all commands would be timeouts. The LCD interface works as expected. I have no idea where else to look. Don’t see any other components that are damaged.
I’m assuming the 2560 is fine but the 16U2 is not. I’m going to try the 16U2 reset procedure. This isn’t really worth my time to diagnose but I hate to landfill this thing.
I’ve got a RAMPS 1.4 board that I’m borrowing from a friend I’ll try tonight.
Ramps board installed, and the steppers are moving great. Back in business.
Now moving on to cleaning up wiring and finalize mounting.
I think the problem with the RUMBA board is that I blew the 16U2 and its junk or the 16U2 needs to be wiped and reprogrammed which I’ll do at some point in the future when I have time.
Also,
When the board blew, I purchased a new set of DRV8825 stepper drivers from Amazon just in case.
I was having some issues pre-board blew with the system ‘missing’ steps; you would call for a +10mm move on the LCD and it would only move the axis 5mm, randomly. So sometimes you’d get a +10mm move or a +5mm move. The new DRV8825 stepper drivers work exactly as they should and are much quieter. The old ones are Geeetech from eBay but I suspect they are knockoffs or just general garbage.
Hi, I’m still here.
I have very very slowly been working on the wiring.
I have a 3D printed enclosure for the Ramps 1.4 board now and am wrapping up cable lengths and XLR connectors for the servos/endstops.
More to come!
2 Likes
Almost Almost!
I need to wire the Z stepper and figure out routing for that and the Makita router power cord.
I’ll be buying some cable chain to clean things up a bit and probably wire up a relay controlled widget for supplying power to the router.
1 Like
Nice printed enclosure! Build is looking good!
Sorry to keep spamming but this is getting exciting.
It appears to be off about 0.25 mm every 100 mm, which i think is pretty good but would love to dial it in even more.
It shouldn’t be. At that level of accuracy it is really hard for us to help though. Pulleys tight, belt tension, loose pen in the barrel, loose pen on the mount, too much pen tension.
Have you tested the actual axis in a long move? Each axis to it’s extremes?
As bad as this is, Are you using import drivers? I had a batch off more than that somehow, I use them on my oldest two MP3DP’s. My Z axis is 788 steps instead of 800 on them, super craptastic 8825’s.
1 Like
I like you Ramps enclosure. I have built one for the Archim that also includes the LCD into the top. I rarely use the LCD and control, so it’s ok to be flat. Still thinking of a good way to print a top with a slant to angle the display without having to do a ton of support material.
I’ve been trying to figure out an efficient way to get the top to stay connected to the bottom. Right now it just slips in a rebate in the bottom section. I now have figured out a way from looking at yours to be able to put screws in at the side to keep them together. That might work.
It’s always a tradeoff between supports, print layout, design angles, intersecting parts. Trying to be efficient, but it’s not always evident to me.
And like that you have the aircraft connectors. I didn’t go for these but perhaps some day I will. It does sure help for being able to work on the control box without having to drag an octopus around.
Weird about your loss of accuracy. I was just floored once I got everything connected correctly and it just worked perfectly over a 700mm x 700mm bed without any loss of accuracy. Ryan’s design and using his firmware and board and motors sure made that easy.
1 Like
My build area is ~ 600 mm x 250 mm
The machine zeros itself to the same point every time. When I move the machine all the way out to 600, I’m getting 2 mm loss of accuracy on the long end. I pulled the belts a wee bit tighter to see what would happen and its now out to 602.5, so it looks like my belt tension is a bit too tight. I’m out of zip ties and need to move to a better tensioning method anyways. I’m going to give this a shot and will report back: https://www.thingiverse.com/thing:1833934
2mm is nuts, Have you tried adjusting your steps? Like I said the cheap import drivers are apparently capably of having the wrong steps somehow.
1 Like
You are seeing 0.33 percent extra length. I had measured stretch of some belts at approximately 0.04 percent per kg of tension (https://www.v1engineering.com/forum/topic/double-belt-for-lr2/#post-114997), so 8kg of tension could create that much stretch.
The tension only needs to be as high as the maximum load your tool will experience, so a few kg at most should be sufficient. I have heard that with that belt tensioner you linked people have a tendency to over-tighten the belts. Tighter does not increase stiffness (once you have covered the maximum tool load) so take it easy with that tightening mechanism.
If you want to keep your tight belts you can also adjust steps per mm in firmware to account for your belt stretch.
1 Like