More problems bringing up my ShaperCube 2.1

[Bryce Schroeder]

I got the firmware to compile, but the stepper motor drivers get really hot when I use buttons on the printer interface. Too hot to touch. They have heat sinks. Any suggestions? (I am using all three jumpers under the drivers and made the appropriate configuration adjustment to the firmware. Does that effect this?)

Also, the temperature sensors always read "357." At this point I am not powering up the extruder or bed because it does not seem to have an accurate temperature reading. Suggestions?

Thank you for any help you can offer.

[RedoX Ws]

Hi,

I also got the same kind of problem when I first started with my Sc2.1. You can try to lower the current flow through the driver by turning counter-clockwise the small pots on the drivers. The drivers should cool down. It also reduces the torque of the motors, so make sure your guides (smooth rods) are correctly aligned (and parallels) and that the carriages moves well (without too much efffort) maybe add some grease/oil to make the movement smoother and easier.

I also added a small fan (40mm) which provides a small airflow on the drivers, just to make sure they stay cool. I currently have two drivers without heatsinks (until I get some sticky thermal grease to keep them in place) and it works well.

If the motors don't move and the drivers gets hot, it might be fried. Try to swap it with an other one or put it on another axis to test it (be careful not to fry an other driver if you have a short somewhere though)

Do both the temperature sensors read "357" ?
They might be incorrectly wired or plugged in.
Did you select the good thermistances number in the firmware ? They are (at least on my cube) both 100k thermistor, simple model, which I think is number 1.

Don't power up the extruder/bed until they read something like 27-30°C (depends of the ambiant temperature) when idle.

I ran into a few problems too when I started so don't hesitate to describe well what you're facing ;)

Good luck =)

[Bryce Schroeder]

Thank you for your comments.

I have measured the resistances at the thermistor connector, where it plugs into the pin header, and found it to measure ~100k, like the ones you have. This was the setting in the configuration file of the firmware.

Both the temperature sensors read 357. Occasionally, when the sensor is unplugged, I have seen readings of 347, 345, 350 and 327, erratically. But with the sensor (or a 100k resistor) plugged in, they read 357.

The voltage across the sensor inputs T0 and T1 is ~0.04V. In contrast, the voltage across the T2 sensor input is about 4.95V. (Which I assume is within spec, or maybe this voltmeter, which was very cheap, is not well-calibrated.) So it does seem like something has burned out THERM0 and THERM1 somehow. But I am yet to figure out how it happened, so I am reluctant to just plug the thermistor into THERM2 and recompile...

[Bryce Schroeder]

Sorry for double post, but to clarify: I also physically inspected the heatbed underside and found no evidence of a burned-through short, which would explain how it seems electrically normal now.

[Bryce Schroeder]

Okay, after further confirming no electrical problems with the extruder thermistor, connecting it to THERM2 and altering the firmware accordingly, it is now reading 21C. I will attempt to bring up the hot end now. Wish me luck.

It looks like there are four unused analog pins on the RAMPS 1.3. Anything preventing me from using them for the heatbed thermistor? (I have the needed resistor, capacitor etc shown in the schematic.)

[Bryce Schroeder]

Thanks again to everyone who has helped so far.

Update: I transferred the hot end thermistor to T2, and consequently the hot end seems to be basically working. I have even made some basically successful prints, although due to warping because of not having a heated bed, they are not what they could be in terms of quality.

It appears that the thermistor inputs T1 and T0 are burnt out, as are the heater/fan outputs D8 and D9. I will take the logic analyzer to it and see if the fault with the heater outputs is due to fried MOSFETs or if it is digital, and replace the FETs / change their control inputs as appropriate. I also plan to use some of the apparently unused analog channels on the arduino mega to replace the damaged T1 and T0. If there are any pitfalls with this course of action, let me know.

I've also ordered a new arduino just in case, but I am as of yet holding off on ordering a new RAMPS, partly because the fried FETs should be easy to replace if that's the problem, but also partly because the RAMPS is expensive and of limited use compared to an extra arduino :)

Thanks again.

[Bryce Schroeder]

Update: I have bypassed the damaged ADC on the arduino by using A12 for the heated bed thermistor - I attached a little bit of perfboard to a pin header installed in the RAMPS AUX-2 connector, and put the capacitor, resistor, and another 2-pin header for the sensor to connect to on it.

Via the logic analyzer, I have determined that the D8 MOSFET is burnt out, as is, oddly, its associated output line on the arduino, apparently. The D10 MOSFET is working fine and controlling the extruder heater at this point. The D9 MOSFET seems to be working, and I suppose it remains now to plug it to the bed.

What is the resistance of the shapercube's heated bed supposed to be? Mine is only about 1.8 ohms. This would burn ~80 watts at 12V. Can the MOSFETs really control this much power without a heat sink or anything? It seems like it's a lot to ask of them.

Thanks

[RedoX Ws]

Hi,

I didn't have much time these last three days, but I've been following your problems... Seems like your mega took a lot, I'm surprised it still partially works. FIY, You can find mega2560 clones on dx.com for 20$, might be worth the try if it happens again (since you already ordered one).

RAMPS is mostly passive components and  headers, except for the three big mosfets, so until the pcb is damaged, I think you can keep yours.
(If you don't have any spare, I recommend sourcing some extra Pololu drivers just_in_case. It's really frustrating to wait only for these little pcbs...)

Be careful when switching mosfets, I'm not sure D9 can handle the load of the Heatbed... Might be a good idea to check before trying. The heatbed of the Sc is a standard Mk1 heatbed (should be written on the top of the bed), resistance should be around 1 ohm (http://www.thingiverse.com/thing:3919) so I think the MOSFET can handle it. On my RAMPS (1.3) I don't have any heatsink and the MOSFET works fine, even for (quite) a long time.

I don't see any reason you couldn't find a replacement for T0 and T1 to keep your changes just be careful when you update your firmware. I'd say, leave the RAMPS board untouched, so you can return to normal when your arduino arrives.

If the first layer doesn't stick enough to the cold heatbed, try with kapton taped, and clean it with a little bit of alcohol just before the print, it might help a little bit...

Good job on your debug, it doesn't look really easy.

[Bryce Schroeder]

I was informed by Joachim that only the D8 channel had large enough traces on the RAMPS to support the heater, so I replaced the MOSFET from stock onhand. (It's equivalent from a different manufacturer.) Rather than connecting the MOSFET's gate to the pad, I bent it up carefully and attached it with a small wire to D42 on the AUX-2 connector where the small PCB for the thermistor header and ancillary components are located. If I decide to replace the old arduino with the new one, I can cut the small wire and bend down the leg to touch the original pad and move the thermistor connector back to its original place, so the modifications would not be difficult to reverse.

The D8 channel - properly the D42 channel now - is working, in the sense of turning on and off in response to heater bed commands. I will try it cautiously tonight. The new MOSFET has also been outfitted with a heat sink.

Thanks for that tip about the first layer. That has been a minor problem on some of the prints I've done, or rather, tried to do and had to start over because they didn't stick.

Take care,

- Bryce