PSU troubleshooting

Printable View

08-05-2014, 12:35 AM
wingnut1000

PSU troubleshooting

So I've been searching for help on various sources like reddit/r/3dprinting, reprap forums (which is down atm) and haven't gotten much help.
Printer: Prusa i3V 8"

First post
Purchased a Prusa i3v 8" build area from makerfarm. Have been upgrading and tinkering with it for 3 weeks now. Printed and installed the improved hexagon cooling system and the Z probe servo mount for auto bed leveling. The heated bed takes AGES to heat up. Average is about 30 minutes from room temp to 110 for printing ABS. Print bed is 3/16" borosilicate with aluminum foil/cardboard underneath for insulation. Power supply is 12V DC 30A PSU found here...

Anything I can do to speed up the bed heat time? PSU strong enough? Can I test with a volt meter to ensure PSU is outputting correctly? I'm currently looking into building an enclosure for it but it's not possible ATM. I also plan to rig up some LED's and possibly some additional cooling fan for printing PLA, should be fine right?

Also Arduino board fried on me yesterday, looks like the USB controller burnt up. Can control using LCD screen but PC will not recognize board now.
http://i.imgur.com/RfdrQdA.jpg
http://i.imgur.com/4OXeZaj.jpg
http://i.imgur.com/jvcZTkB.jpg

Smelt burning PCB after powering on, pulled power and USB connection to PC. USB cable was hot.
~~Waiting word from makerfarm to see if they will warranty~~ Makerfarm sending replacement.
Anything I can do to prevent this in the future? Possible fault of PSU?
Any direction is much appreciated.

Second Post
I received my replacement board today, and hooked everything up. I octuple checked everything. Began a print and about 5 min in the printer froze. I got that faint scent of burning PCB. Pulled power and usb immediately. The board is still recognized by the PC, and looks to be responsive.
At this point the only thing I have left to suspect its the PSU.
So, am I right?
What should I replace it with?
PSU or PC power supply?
I'll spend more if it would heat the bed quicker (average is about 15-20min for 110C for ABS)

Any help is much appreciated.
08-05-2014, 02:19 AM
Mjolinor

First guess would be a ground problem somewhere allowing current to flow up the USB cable. Where in the world are you. That board could be mended quite easily and I suspect you need help to find your fault if you have followed all the instructions correctly. Check your new board in the same location as your pictures, does it show any sign of the same problem?
08-05-2014, 02:23 AM
wingnut1000

Old board appears to have damaged component, not just the run on the back. Don't think it would be easy to repair with my limited knowledge. I'm in the San Francisco bay area. No sign of damage on new board, think I got to it in time. Followed instructions to the letter, checked everything numerous times.
08-05-2014, 04:00 AM
Mjolinor

Which damaged component? The 220 ohm resistor pack, if knackered will be easy enough to replace though it is unlikely to actually be damaged. What has probably happened is that the Mega 16U2 is fried due to current flowing where it shouldn't have. It is hard to tell from your pictures whether the burnt track is a 5 volt line or ground. I would guess a 5 volt line so something has pulled a huge current down there. This could be due to voltage differences between your PC and the printer, you will be unlikely to fix that by replacing the board as you have done. You can't get a fix for this in a forum, you need to methodically go through the board and its connections with a meter and maybe a scope. There are loads of things that can cause this, I would start by checking that grounds from teh meter to all of your equipment are sound.
08-05-2014, 02:19 PM
printbus

Has the long time to bring the heat bed to temp been a problem all along, or a problem that just started? EDIT: Please clarify the warm up time. One part of the first post says 30 minutes average. Another part of the post says it takes 15-20 minutes to heat up. That's a big difference. Do you have a heat bed relay installed? That would speed up the heating at least a bit since a relay has less voltage loss than the on-board FET normally controlling the heater. Heftier gauge wire in the path from the power supply to the heater makes a difference, too.

Sure, you can measure the output of the power supply with a DC voltmeter. Put it across the + and - outputs of the power supply, or the power input pins at the RAMPS board. You should read about 12V DC. If the power supply is providing an output but it's crappy, the DC reading may be misleading since your meter will provide an interpretation of what it is seeing. You could also set the meter for AC volts and measure that. Any substantial AC component (maybe more than 0.2V?) suggests the quality of the power supply output is suspect. If the AC component is low and the DC voltage is low, try adjusting the trimpot on the power supply by the LED while you are monitoring the DC output voltage for the desired 12V. If you didn't do it, this should have been done as part of your build.

On my i3v, I've noticed that the RAMPS and LCD will power up if the printer is turned off but connected to the computer via USB. What I don't know is how much current the printer will try to draw from the USB in this case, or possibly if the printer power supply is on but set too low or has issues. There's a limit to how much 5V current should be drawn from the USB.

If it were me, I'd disconnect the power supply from the printer and work with it by itself until I felt I could trust it. To test the power supply under load, you could do something like power some 12V automotive light bulbs. They tend to draw a lot of current. I'd also leave the USB unconnected until you understand what is going on or you risk damage to the computer you're connected to.
08-05-2014, 09:34 PM
wingnut1000

Quote:

Originally Posted by printbus

Has the long time to bring the heat bed to temp been a problem all along, or a problem that just started? EDIT: Please clarify the warm up time. One part of the first post says 30 minutes average. Another part of the post says it takes 15-20 minutes to heat up. That's a big difference. Do you have a heat bed relay installed? That would speed up the heating at least a bit since a relay has less voltage loss than the on-board FET normally controlling the heater. Heftier gauge wire in the path from the power supply to the heater makes a difference, too.

Sure, you can measure the output of the power supply with a DC voltmeter. Put it across the + and - outputs of the power supply, or the power input pins at the RAMPS board. You should read about 12V DC. If the power supply is providing an output but it's crappy, the DC reading may be misleading since your meter will provide an interpretation of what it is seeing. You could also set the meter for AC volts and measure that. Any substantial AC component (maybe more than 0.2V?) suggests the quality of the power supply output is suspect. If the AC component is low and the DC voltage is low, try adjusting the trimpot on the power supply by the LED while you are monitoring the DC output voltage for the desired 12V. If you didn't do it, this should have been done as part of your build.

On my i3v, I've noticed that the RAMPS and LCD will power up if the printer is turned off but connected to the computer via USB. What I don't know is how much current the printer will try to draw from the USB in this case, or possibly if the printer power supply is on but set too low or has issues. There's a limit to how much 5V current should be drawn from the USB.

If it were me, I'd disconnect the power supply from the printer and work with it by itself until I felt I could trust it. To test the power supply under load, you could do something like power some 12V automotive light bulbs. They tend to draw a lot of current. I'd also leave the USB unconnected until you understand what is going on or you risk damage to the computer you're connected to.

Thanks for this. Heat up time has always been an issue but it has gotten worse, I'd say 20 min sounds about right, however I did not time it. Wires to RAMPS is 14 gauge stranded copper, to heated bed was pre-wired.
" You should read about 12V DC." Reading 12.19 and stays there, doesn't seem to fluctuate.
Sorry but I didn't understand what you are talking about regarding the AC voltage.
"RAMPS and LCD will power up if the printer is turned off but connected to the computer via USB." Mine does this as well.
08-05-2014, 09:55 PM
printbus

Quote:

Originally Posted by wingnut1000

...Sorry but I didn't understand what you are talking about regarding the AC voltage.
"RAMPS and LCD will power up if the printer is turned off but connected to the computer via USB." Mine does this as well.

All I was getting at with the AC measurement is to put the meter in AC volts mode and look at the same power outputs from the power supply. It's not a good test, but it might reveal something. Ideally, the 12V DC output should read 0V AC. If the meter shows a considerable AC (ie, not fixed like good DC would be) component, the power supply might have issues.

My point on powering up RAMPS and LCD via the USB port is I wouldn't do that. I'm not saying you are - it's just one way that people can have unexpected current flowing through the USB.
08-05-2014, 10:00 PM
wingnut1000

Quote:

Originally Posted by printbus

All I was getting at with the AC measurement is to put the meter in AC volts mode and look at the same power outputs from the power supply. It's not a good test, but it might reveal something. Ideally, the 12V DC output should read 0V AC. If the meter shows a considerable AC (ie, not fixed like good DC would be) component, the power supply might have issues.

My point on powering up RAMPS and LCD via the USB port is I wouldn't do that. I'm not saying you are - it's just one way that people can have unexpected current flowing through the USB.

Thanks for the clarification. I tested using AC volt mode and it does indeed show 0V. Thought I was doing it wrong.
How would I avoid powering over USB. Don't think I have control over it. I just ensure that I power on the power supply before connecting the printer to my PC
08-06-2014, 12:26 AM
printbus

Quote:

Originally Posted by wingnut1000

Thanks for the clarification. I tested using AC volt mode and it does indeed show 0V. Thought I was doing it wrong.
How would I avoid powering over USB. Don't think I have control over it. I just ensure that I power on the power supply before connecting the printer to my PC

So from what we can tell, it's sounding like your power supply is OK. One additional clarifier would be to check the power supply output when you're heating up the bed and the hot end. That'll put a pretty good load on the power supply. For comparison, I ran a timed test on my i3v. It hit 110 degrees in about 8 minutes, although I have only 1/8-inch glass and do have the bed heater powered through a relay. Is the thermistor under the heat bed still snug to the bed? If that starts coming loose, it'll be slow to measure temperature changes. Your reddit post mentions 3/8-inch glass. I assume the 3/16-inch mentioned here is correct? 3/8-inch would be really thick stuff.

I've also looked at the board damage shown in the pictures. I have no answers on what is going on, but here's more information. Maybe it'll help someone else jump in with an answer for you.

For the ATMEGA2560 board, I used this as a representative schematic - http://arduino.cc/en/uploads/Main/ar...560_R3-sch.pdf. For the USB connector pinout, I used this as a typical USB type-B connector - http://www.mouser.com/ds/2/154/61729-335969.pdf. First, the schematic shows that the 22-ohm resistors in the two USB data lines are what your pictures show burning up. The part is marked as 220 since that's how the industry does it. The first two digits are the significant digits in the value, and the third digit states how many zeros get added. So, a marking of 220 means 22 ohms. These parts burning up is not good. Either there's a problem with one side of the interface driving the lines way wrong in voltage, or something has happened that has both sides of the interface trying to drive the lines at the same time.

The heftier trace in your third picture is almost more troublesome. If I follow the pinout right, that's on pin 4 of the USB connector, which is ground. What is the spot that looks like a spec on the first rectangle pad the trace goes to? Almost looks like something may have shorted there - especially since there's no discoloring after that pad. Ignoring that, I don't know - a bad cable shorting +5 to ground? Some really funky ground loop-problem between the USB shield and the ground pin, caused by some grounding issue going on between your computer and the printer?

Sorry I can't be of more help.
08-06-2014, 12:50 AM
wingnut1000

Quote:

Originally Posted by printbus

So from what we can tell, it's sounding like your power supply is OK. One additional clarifier would be to check the power supply output when you're heating up the bed and the hot end. That'll put a pretty good load on the power supply. For comparison, I ran a timed test on my i3v. It hit 110 degrees in about 8 minutes, although I have only 1/8-inch glass and do have the bed heater powered through a relay. Is the thermistor under the heat bed still snug to the bed? If that starts coming loose, it'll be slow to measure temperature changes. Your reddit post mentions 3/8-inch glass. I assume the 3/16-inch mentioned here is correct? 3/8-inch would be really thick stuff.

I've also looked at the board damage shown in the pictures. I have no answers on what is going on, but here's more information. Maybe it'll help someone else jump in with an answer for you.

For the ATMEGA2560 board, I used this as a representative schematic - http://arduino.cc/en/uploads/Main/ar...560_R3-sch.pdf. For the USB connector pinout, I used this as a typical USB type-B connector - http://www.mouser.com/ds/2/154/61729-335969.pdf. First, the schematic shows that the 22-ohm resistors in the two USB data lines are what your pictures show burning up. The part is marked as 220 since that's how the industry does it. The first two digits are the significant digits in the value, and the third digit states how many zeros get added. So, a marking of 220 means 22 ohms. These parts burning up is not good. Either there's a problem with one side of the interface driving the lines way wrong in voltage, or something has happened that has both sides of the interface trying to drive the lines at the same time.

The heftier trace in your third picture is almost more troublesome. If I follow the pinout right, that's on pin 4 of the USB connector, which is ground. What is the spot that looks like a spec on the first rectangle pad the trace goes to? Almost looks like something may have shorted there - especially since there's no discoloring after that pad. Ignoring that, I don't know - a bad cable shorting +5 to ground? Some really funky ground loop-problem between the USB shield and the ground pin, caused by some grounding issue going on between your computer and the printer?

Sorry I can't be of more help.

Your help is MUCH appreciated.
As I'm pulling a pretty heavy load working from home atm I can't do much troubleshooting. I can say that today I did a successful small test print (about 6 minutes) by printing from cache and detaching printer from PC. However I attempted another print directly through Repitier host and after noticing some strangeness on the LCD after a few minutes of the bead heating, I felt the USB cable and sure enough it was getting hot. Pulled everything before any damage again.
So, I mentioned that I installed the auto level bed probe, this required a large amount of modification to the firmware. I've successfully re-flashed back to stock. As soon as I get a bit more time, I'm going to read your previous more thoroughly, and do some more testing on stock firmware. If all goes well there I will methodically calibrate the firmware up to spec for the auto level bed probe. I always ignored the "modifying the firmware could cause irreparable damage bla bla bla" warnings, but wouldn't that be something if it was the case here? Can't think of any lines I modified that would cause this but who knows. I'll report back with my findings as soon as I get the time....