UM2 is dead after short circuit.

[SyntaxTerror 344]

like @SyntaxTerror says they don't look right.

 

Well, it could be me not remembering what's what correctly. To be honest I've had enough versions of what's what on the hot end fan connector lately, if you know what I mean

[neotko 1,417]

He could check with a 5V battery if the pos/neg is ok? That's easier than plugging it on a 300€ board

[SyntaxTerror 344]

Yeah that's for sure

I'll have a look at my own printer when I get back home and see what connects to what.

In the mean time, @shurik , have you given up on fixing the board?

[shurik 87]

have you given up on fixing the board?

No, of course not.

I will check it with the multimeter according to your scheme.

The correct connection of the fans is a bit confusing. On the photo of the board the yellow and green wires leading to the fans cannot be seen on the upper right corner because of the thick black cable. How they should be properly connected?

[SyntaxTerror 344]

In the wire loom going to the print head, the model fan wiring is Red --> Yellow and Black --> Green. It should also be impossible to swap, as the connectors are keyed and will only go in one way. So that part looks correct.

[SyntaxTerror 344]

I'll have a look at my own printer when I get back home and see what connects to what.

 

Actually, it IS me who's not remembering things right. On the hot end fan, the blue wire of the extension cable connects with the red fan wire, and the orange wire goes to black. I think I might be damaged by my education. When I see blue, I can't help thinking neutral. And with the gear I work on every day, blue is pretty much always 400Hz AC cold. So my disposition is just wrong

The hot end fan connector is correctly installed. But in a strange place.

[onkelgeorg 36]

@onkelgeorg - the link is invalid, but you have a good idea - will look on the ifixit site later on.

 

Sorry for that mistake. It seems that the link was cut for some reason :(

Here is a working one: https://tinyurl.com/hn5l4wh

[shurik 87]

Alright, as I said I think it's prudent to check if you have supply voltage before drawing any sort of conclusion. So hook up your board again and flip the power switch on.

This is the copper (back) side of the board:

Get your multimeter out, set it to measure DC voltage. Your common measuring pin goes on TP24, which is a ground. Then measure the test points for the fan and look for 24 volts DC. I can't remember which one it is - the silk screen with the test point numbers has an error on this board. One of the two should read 24 volts though, if your power supply is working and the power switch is closed.

If you can find this voltage, move on to measure TP16 next. That's the 5V output from the buck converter.

 

Report:

This is how the board looks like when powered:

Could that be that I just did not plug the power properly when the board was dismounted the last time? I was too depressed and tired then. Maybe...

Voltage between TP24 (GND) and:

 

• TP16 = 4.9V

• TP77 = 24V

• TP78 = 24V

 

In short, the readings are normal. The display is dead, and so the back cooling fan.

What else to check?

P.S. No matter how it turns out, your help, all of you, is absolutely amazing. It is that help and cooperation spirit that won the UM for me over other alternatives back 3+ years ago, and I did not regret the choice ever since. Thank you!

Edited March 1, 2017 by Guest

[SyntaxTerror 344]

Could that be that I just did not plug the power properly when the board was dismounted the last time? I was too depressed and tired then. Maybe...

Voltage between TP24 (GND) and:

 

• TP16 = 4.9V

• TP77 = 24V

• TP78 = 24V

 

In short, the readings are normal. The display is dead, and so the back cooling fan.

What else to check?

P.S. No matter how it turns out, your help, all of you, is absolutely amazing. It is that help and cooperation spirit that won the UM for me over other alternatives back 3+ years ago, and I did not regret the choice ever since. Thank you!

 

Now that is an intriguing and mysterious result.... most unexpected. So, the power supply appears okay, we have 24 volts in. The buck converter appears to be working, delivering 5 volts. Yet the hot end fan is not moving and the display is dead.

The hot end fan being dead is somewhat odd. I can't see any other possibilities than either the fan is dead or the wire loom has a defect somewhere.

I expected absence of 24 or 5 volt supply since the fan was not moving and there's no signs of life on the display. Since 5V presence is now verified, I'd go one step further with the test points and check the 3.3V regulator as well, at TP17:

Next, I think the right thing to do is unscrew the hot end fan from the print head. Unplug it from the wire loom extender cable and plug it directly into the main board fan connector. If it starts running then, we know you have an open circuit in the wire loom to the print head.

No matter what happens, for the fan you should just use J34 instead of the serial header, that's what it's there for. We might as well get that corrected. Hooking the fan up like it is on your printer makes no sense.

As for the apparent unresponsiveness of the printer, you should try plugging in the USB cable and firing up Cura. See if it can connect with the printer.

Edited March 1, 2017 by Guest

[Torgeir 242]

Hi,

I must say this is really strange or.. Just checked the schematic and found that pin 1 is + (positive) and pin 2 is ground. If you connect this one to the J34 connector the polarity will be wrong.. There must be a history behind this?

Another thing is about those small fan and the polarity, some of those fans do not tolerate wrong polarity as the electronic controller will be damaged if polarity is swapped. Just so you know.

As far as I'll remember the connector at the main board is bigger than the connector at the 5VDC fan.

Well, you got plenty of good advice in here.

Good luck.

Thanks.

Torgeir.

[SyntaxTerror 344]

Just checked the schematic and found that pin 1 is + (positive) and pin 2 is ground. If you connect this one to the J34 connector the polarity will be wrong.. There must be a history behind this?

 

Pin 1 is 5V and pin 2 is GND on both J22 and J34, so there won't be any difference between the two. I'm thinking maybe the colors of the extension cable wires has you confused like it did me?

 

As far as I'll remember the connector at the main board is bigger than the connector at the 5VDC fan.

 

On both my UM2 and my UM2+, the hot end fan terminates in a JST connector that mates with the onboard J34.

[shurik 87]

TP17: 3.27V.

Cura or the computer do not see the printer being connected with USB.

[SyntaxTerror 344]

This is all very strange. I think we can safely conclude you have somehow suffered two separate problems at the same time. The 24V adapter is working, the 5V buck converter is working, the 3V3 regulator is working - the only supply line we haven't checked is from the 13V boost converter on the controller board. Given that the printer refuses to connect with the computer, it's probably nothing to do with the display anyway. The fan issue must obviously be a problem with either the extension cable or the fan itself. The power is there, so. I'm betting on the fan, it's probably just dead. The hot end fans lead a rough life.

If the 5V supply is up, then both of the microcontrollers should be doing something. I'm thinking maybe the next logical step is to try to feed the printer some new firmware through ICSP? I've never had reason to do so myself, and I know there are a bunch of pitfalls there - maybe someone has a fail-proof method?

[shurik 87]

How can I check if the display receives power? There should be a reason why it is dead.

I tend to agree that the fans are probably a separate issue.

[Torgeir 242]

Hi,

Sure, it was the color on the wires that's not correct according to the picture.

Sorry, just learned that the original printer is built different than my printer.

As I've built my printer, I've got a fan with the smaller JR connector preinstalled so I used this type and "installed" them inside the extruder head. I've later seen most of the printers have the connectors outside the extruder that's I think is to bad -and actually lead to this problem..

I've been reading through all write up in here and noted that you said you had a problem with spark (static electricity?) since day one, I'll thinking you are living in a very dry environments, right?

Working on electronics board without any antistatic protection is very risky, as it can permanently damage such sensitive electronics.

Its hard to see this card is completely dead, at least you should see the “connect light” for the USB as this is controlled by another micro controller (ATMEGA16U2-MU) on the main board. This one can also be “programmed” by using ICSP2 port J2.

The main micro controller is ATMEGA2560-16AU, this one can be programmed by ICSP1 port J1.

You will need a programmer for this update, such programmer could be another Arduino controller card with proper software and connecting wires. As I have never updated my printer this way I cant advice in this matter.

Display power (5VDC) can be measured at TP22 and TP45, but I'll think you cant see anything without the main card working.

Been thinking about a master reset line that’s kept down all the time, but not likely..

Anyway, good luck in fault finding.

Thanks.

Torgeir.

[shurik 87]

Display power (5VDC) can be measured at TP22 and TP45, but I'll think you cant see anything without the main card working.

Indeed, there is 5V between TP24 and each of them - TP22 and TP45. So why the display is dead and the system itself is dead? Weird...

[SyntaxTerror 344]

How can I check if the display receives power? There should be a reason why it is dead.

I tend to agree that the fans are probably a separate issue.

 

Well, I think the display is dead not because of lack of power but because it's not receiving anything to display. I'm working from the observation that the PC won't interface with the microcontrollers, which we've determined should be under power. Like an arduino, the ATMEGA 16U2 should provide a USB interface to the ATMEGA2560 and we should be able to communicate with the printer. They should start up when the power switch is closed, and you should be able to connect. If the problem was with the display itself, the USB connection should be unaffected. I think the right thing to do right now is try to ensure the microcontroller firmware isn't corrupted. As most people with the 16U2 equipped arduinos can attest to, they're just not as stable as the older FDTI equipped ones.

The display itself is powered by 3V3 (VDD) and 13V (VCC). The former we've already checked, the latter is only available on the controller board. It's created from a boost converter on the 5V line and only used by the actual display. It's available on TP5 (on the controller board, not the main board) should you want to measure it (be careful removing the cover, it holds the display in place). I'd probably leave it alone though, my suspicions lead me elsewhere.

[shurik 87]

So it looks that the board itself has all the power it needs and the prime suspect is the microcontroller whose firmware probably got corrupted. Is that correct assumption at this point?

If so, how that firmware can be checked/restored? I have absolutely no tools to interface it to anything except the USB and PC.

[SyntaxTerror 344]

So it looks that the board itself has all the power it needs and the prime suspect is the microcontroller whose firmware probably got corrupted. Is that correct assumption at this point?

If so, how that firmware can be checked/restored? I have absolutely no tools to interface it to anything except the USB and PC.

 

Well, I can't be sure that's the case. But at this point in the troubleshooting, it would certainly serve as my main suspect. You can spend untold hours looking for a hardware problem when in reality the microcontroller just insists on mismanaging the peripherals. I would most definitely attempt to get fresh firmware on both microcontrollers before doing anything else.

Since you can't connect and upload firmware via Cura, it looks like you'll have to use the ISCP headers on the main board to upload it. That's where my experience is pretty much zero - never did it on the Ultimaker. I'm hoping someone with experience could help out here? As for programming adapters, an arduino usually works very well for this sort of thing if you have access to one.

[shurik 87]

@SyntaxTerror - your help is invaluable!

I will try to find around someone with the needed equipment and experience, although no idea at the moment where to start.

Meanwhile, how can I check that the microcontroller gets all the due power supply it needs?

Just to be sure it should work, at least.

[SyntaxTerror 344]

@SyntaxTerror - your help is invaluable!

 

Well, we haven't fixed anything yet

 

I will try to find around someone with the needed equipment and experience, although no idea at the moment where to start.

Meanwhile, how can I check that the microcontroller gets all the due power supply it needs?

Just to be sure it should work, at least.

 

The ISCP headers are what you use to upload firmware directly to the micros directly. Most Arduino people have been in contact with those at one point or another, I'll bet. Arduinos of course are cheap and readily available, and make for excellent ICSP programming interfaces. Computer --[uSB]--> Arduino --[iCSP]--> device to program.

I'm hoping someone who's had an unresponsive USB interface can lend us a bit of experience.

The microcontrollers run off the 5V rail, so barring any unexpected supply noise or burnt tracks on the board, they should be good to go.

[shurik 87]

Update.

In incredible 11 days the replacement board from here arrived and was installed today.

The printer is dead yet!!! :angry:

There are some behavior changes, however:

 

1. When powered, there are periodic sounds and kinda weak clicks, like something tries to get properly up but just can't. I cannot determine for sure where the sounds come from but it looks like they originate from Z-motor or its area. The display stays dead.
    
   
2. While the sounds are happening, there is yellow blinking light just next to the normal green one on the board.
   
3. When connected via USB, the computer (Win10) initially started the "New device" installation and duly rings when the printer is being connected/disconnected or powered on/off.
   
4. Cura 2.4 cannot see the printer but performed firmware update that changed nothing in printer's behavior.
   Cura 15.4 agrees to start the firmware installation but almost immediately fails with error message:
   

   Failed to install firmware:
   This firmware is not compatible with this machine.
   Trying to install UM2 or UMO+ firmware on an UMO?

 

 

Here is a short video:

 

Besides some noise from me holding the phone and the board, you can hear the weak background zummering and the yellow blinking begins.

So, there is some progress, so to call it.

Any ideas, please?

Edited March 15, 2017 by Guest

[neotko 1,417]

Umm. Maybe you really need to go to step 1, unplug anything except the oled and see what happens. Also, maybe is the psu itself??

Edited March 15, 2017 by Guest

[shurik 87]

Well, I checked the voltage of PSU and it is correct - 24V.

BTW, the clicks are coming from Z-motor. I can feel pings when holding the printing plate. It looks like the Z-motor tries to run but fails. The printer lays on its side but it never was a problem for it in the good old days.

P.S. Just measured the voltage on the new board. Just as good as on the old one.

Edited March 15, 2017 by Guest

[neotko 1,417]

Disconnect the motor to see if at least goes on?