anon4321

If you are an old car guy and have a dial indicator, there is a dial indictor holder that allows you to place the indicator into the head. I'm thinking of going that route because even with a feeler gauge, I just can't get the feeling right. Sometimes the feeler is tight or lose or feels the same. And any pressure on the bed can allow the feeler to go under the head and add in compressible blue tape and I find it very inconsistent.

BTW, if you want to heatsink the pololu drivers, I believe this is the correct tiny heatsink: http://www.newark.com/fischer-elektronik/ick-smd-a-8-sa/heat-sink-dip-ssop-tssop/dp/34M6428?CMP=TREML008-003 I have also been successful at twisting off the heatsinks from the dead drivers and cleaning them off my carefully scrapping off the old compound with an xacto knife and then cleaning with 99% alcohol. HOWEVER, before cleaning, note how little adhesive was used. Then I used artic silver adhesive to attach. Note that you only need a very, very small amount of the adhesive component probably like 1/2 cubic mm or less. You don't want so much that it squeezes out onto the pins. However, I have had success with cleaning the tiny pins of uncured epoxy and a fine bristle brush. At least it hasn't gone up in smoke yet... Start with a tiny amount and see how much it spreads. Adding more is easy, removing excess especially if it gets on the pins, is more difficult. I used this: http://www.arcticsilver.com/arctic_silver_thermal_adhesive.htm However, I wanted to use the cheaper alumina product but the store I went to didn't have it: http://www.arcticsilver.com/arctic_alumina_thermal_adhesive.htm Probably overkill for the "black" drivers but shouldn't hurt anything.

I received and replaced my drivers with "blacks" from Pololu and things are working OK now. I think I need to adjust the current. I really think it was my misunderstanding that made it seem to me that the feeder driver was burned out by me turning the extruder. When I went to diagnose, I didn't flip the driver correctly and that killed it. And of course like an idiot, I burn out another before I realized that the orientation is flipped. Since I'm the only one that appeared to kill a driver by turning the feeder, I suspect that it was just me not understanding the safety feature of the feeder control and the need for the hotend to be up to temp before the feeder driver is enabled.

Duesentrieb, I believe you are correct in that the traces won't handle the current. Gr5, Sounds like the relay solution is the way to go for HBP. So my idea is overly complex.

Hmmm, you do have a point. Neither the switch nor power connector is carrying any of the HBP current. That's the reason why you need to solder the HBP negative to the ground plane. Current for the HBP is only flowing through pin two of the connector, the MOSFET (which I think is rated like 60 amps or something so no problem there) and the ground plane. The only thing I'm worried about would be the current carrying ability of the MOSFET to ground plane connection. The connection from pin 2 to the mosfet is pretty beefy so that shouldn't be a problem. However, as you can see in the attached image, the connection to the pads of the MOSFET look pretty skinny (see image). I'll need to do some research on how wide of a trace is required to carry that kind of current. The simple solution is instead of soldering to the power jack ground connection, solder directly to the MOSFET leg that is connected through the skinny traces to the ground plane. I sense an expensive mistake coming....

After even more thought, you could make the ground connection using the left pins of the limit switches. This would avoid soldering to the PCB. However, it's getting more and more Rube Goldberg.

**** NOTE: based on the discussion below, this is not a good solution. **** I should note that I think this will work but haven't tried it. Do so at your own risk. Damage to your UM or a fire or other horrible thing(s) may result! Also, after looking at the PCB for the UM1, there is an alternate way to connect a HBP so that you use the MOSFET directly. I don't know what "bang-bang" mode is but if it is to slow the number of times the relay switches on and off, you can use this alternate method so that the switching is done by the MOSFET as quickly as required. It requires one warrantee voiding modification to the PCB in the form of soldering a wire to the pins of the power connector. Sorry if it isn't as clear as Jason's diagram. Technically, the relay isn't required but it disconnects the HBP second supply from the PCB when you turn off the UM but the switch. I'm not sure leaving that power across the MOSFET would be a good idea.

These aren't cheap but "Mean Well" makes several "desktop" supplies. Since they aren't cheap, they have all the safety certifications required so I tend to be more comfortable with them. Sadly, the biggest won't run both the heated platform and the UM based on the platform using 10amps and the UM taking 6. BTW, this is the adapter UM selected for the UM2 (which seems to run both that platform and the UM2 electronics and steppers). This one is marginal based on the platform using 10amps: GS220A24-R7B http://www.meanwell.com/search/gs220/default.htm This one will cover the platform's amperage with 10% to spare: GS280A24-C4P http://www.meanwell.com/search/GS280/default.htm There is also the 20V version the might, just might, run both at the cost of slower heating GS280A20-C4P http://www.meanwell.com/search/GS280/default.htm Around 90 USD.

Just some advice to the OP, let the people here try to help you. I made the mistake of assuming a bunch of stuff and blew up a stepper driver because I didn't ask these people for help first. I haven't had much time with my UM1 before I broke it but the feeder seemed to work well. Im new to this but I think there are a lot reason why you might have feeder problems beyond the design of the feeder. For example, what size filament are you using and where did you get it? Or do you have the tension cranked up to the point it is crushing/distorting the filament causing it to get jammed at the hotend? Could also be that the end of the Bowden tube was crushed or distorted where it ends in the hotend preventing the filament from passing through it easily and so on... People like illuminarti, sander, gr5 and others have a lot of time on these printers and can probably get you straightened out...

Re:Printbl Their site seems reachable now...

Yeah, forget about the flyback diode.... The issue with using a 24v fan is it will run slower at 19V. Ideally, If you could get another fan that is the same make and model of the one on the UM1, you might be able to support two at 12V overvolted to 19V and really get some air blowing. a 24V fan (or two) might still work even running slower since obviously there will be more fans blowing. Not sure what to recommend you do..

19V to a 12V fan might kill a lesser quality fan and if it fails shorted, it will also kill that transistor. I'd would recommend that if you connect a second fan after replacing the transistor, reduce the maximum to 50% -70%... I was wondering how that little fan produced so my air. Its overvolted... I noticed that the electronics cooling fan is a 24v fan...

Ahh yes the datasheet shows an internal flyback. However, are you sure about the 12v? VCC/2 is coming directly from the power switch and is connected to pin 1 of the regulator which I believe is the VI (voltage in), VO is pin 3 and is the 12 V going to powering the auduino. The transistor has a max current of 4A which is quite a lot for these two fans. If the flyback is integrated, why is it so easily killed?

Well, I concluded that I am probably stupid. Sander IM'ed me to see how I've been getting along. I told him the UM1 was working well until I broke it. Then he laid down this tidbit: "Putting in the stepper drivers the wrong way around is indeed deadly (for them), but when you are switching around stepper drivers to test, note that the E motor is indeed positioned in the opposite way as the other, but there is also a firmware safety measure that it doesn't work unless the print head is at least 180ºC. This is done so you won't damage anything trying to extrude PLA through a cold nozzle." Notice that last part about the safety measure?!?!?!?! I did when it hit me that the reason WHY the driver appeared to not work was because it was disabled by the FW! DOH! So when I switched it with another putting the other in the wrong way, that blew the replacement. However, I'm willing to bet that the original was OK but wasn't enabled due to the cold hotend. Furthermore, the audrino is probably fine and doesn't need to be replaced... I just need to preheat the hotend to see if the audrino is OK with a good driver in the extruder socket... This is probably why no one else has since this issue. Probably because it doesn't exist... Sigh... It sucks to be stupid...

In case you are wondering, this is what we are discussing: http://en.wikipedia.org/wiki/Flyback_diode When the flyback diode is used to simply dissipate the inductive energy, as with a solenoid or motor, cheap 1N4001 and 1N5400 series diodes are used instead. I recommended the 1N5400 because its higher current capacity but I'm not sure how to size a flyback diode. I think I read it should handle the current that is seen in the coil. So in this case, that would be something higher than 100ma x 2.

Making stuff explode is fun when you have a replacement on hand ....

I barely know enough to make recommendations so take this with a grain of salt... You might add two diodes such as a 1N5400 just at the end of the Y cable that for each fan across each connection. Just don't get the polarity wrong. Then shrink wrap them. I believe that the line in the diode would connect to the plus side and the other side to the - lead. Anyway, that is what I would try.... However, I blew up my stepper drivers last night so it's important to consider that I don't know what I'm doing....

It looks like the print head fan that is used in the UM1 is 12VDC 1.2W so that would imply that it draws 100ma at 12 volts. Probably higher at 19V but I doubt that it is more than 200ma. The transistor can handle 4 amps or 4000ma. So two fans shouldn't be a problem. However, if the fan load is inductive, it might cause high enough voltage spikes to kill the transistor. I would check the Y cable for shorts and check the voltage rating of the fan. It's possible the high voltage killed the new fan causing it to fail shorted and that caused the death of the transistor. Of the high spikes from the fan caused the death of the transistor. Definitely should have a flyback diode on it....

So I looked at the specs for the transistor and it seems like it should handle two fans without a problem. On issue might be that the fan is supplied with the full 19V from the power supply. Most fans are 5V or 12V so this higher voltage might cause a problem. Do you know what voltage the fan is designed for?

If the print header fan connector is the one just right of center on the top of the 1st picture IRobertI posted then Q4 is the one that probably needs to be replaced. It is just below the top, more right of center mounting hole. It's a transistor, BD679 (the one previously linked) so it's orientation matters. Before removing the existing one, determine it's orientation. Usually they differ from front to back so find the same front or back indicator and ensure the orientation is correct. For example, front is probably the one with the markings on it or has a dimple on it. They are cheap so buy two in case you get the orientation wrong and the magic smoke escapes. You will need to be able to solder it back in. You will need solder wick to clear the holes after removing the dead one. The easiest way to remove it is (FIRST NOTE THE ORIENTATION) use flush cutters to clip the leads and remove the "head" if possible. Then heat each remaining lead and remove one by one. Then use the solder wick to suck the solder out of the holes. Place the new on in the correct orientation and solder.

Daid, Actually, I think the PCB is designed fairly well. Or at least better than I could do. Although I do believe there should be a screaming warning on the board about the inverted drivers. I can understand that because of the physical placement of the connectors, the layout is better with the side-to-side flip. I actually tried to redesign the board to add external power for a HBP and possibly external power for the stepper drivers. Eagle is not for the novice so I gave up. Would be an interesting project that I might tackle again...

Someone pointed out that I wasn't being polite. The only part I would clarify would be obviously that you shouldn't shoot people over bad board design. I thought the humor was implied. Anyhoo, I think the key is the feeder motor. Given the gearing on feeder, it's possible to get the motor spinning very fast using the big gear and that probably produced enough voltage to kill the stepper driver. The drivers have a max supply voltage of 35V, my guess is if you spinning that big gear fast enough that a voltage higher than that is produced, it's dead Jim.... If anyone has a spare driver and time, remove the board and then the Arduino and disconnect everything and spin away! See if it still works after that.... In case you are wondering what the heck I was doing: http://www.thingiverse.com/thing:40334 See that little knob... yeah that's a BAD idea!

Already ordered Blacks from Polulu along with a new audrino from them. Too late for the weekend but hope to have them early next week.

The thermocouple amp seems OK as does the hotend heater MOSFET. However, I did confirm that you can move the XY axis fast enough to cause the same effect so be careful....

meant to add that this is probably something that is specific to the extruder since the gear ratio allows you to get the motor spinning pretty fast. I'm not sure that you could drive the XY axis fast enough to cause the same effect. I'm also not sure that having it powered would have helped but that would have at least allowed the PS to help clamp the back EMF. So be warned.... don't manually drive any of the steppers by hand.