OH NO! Where did that smoke come from????

[RTRyder]

It was bound to happen sooner or later and I'm glad it happened where and when it did. I was going to do a test flight over the house to check out the most recent changes to the camera mount settings on the Hexa and when I plugged the battery in I saw a little puff of magic smoke inside the dome! (Here's a good reason to NOT paint the inside of the dome if there ever was one!)

I think I set a record for pulling the power plug apart and fortunately there was no further smoke nor any sign of flames, so that's a good thing, I'm happy whatever it was that decided to blow up did it while still on the ground and not 50 or 100 feet in the air.

I popped the dome off and sure enough there was the distinct aroma of burnt electronic component(s) underneath. I removed all the boards from the stack and gave each a close look with a magnifying glass, can't find anything anywhere that looks like it burnt, so that leaves the BL's attached to the PDB. If that's the case, the I2C isolation board may well have just paid for itself, won't know until I pull the ring and have a close look at all the BL's.

Stay tuned...

Ken

 

[Bartman]

it's the MK?

 

[RTRyder]

Yes, the MK, and now I have the worst possible scenario, visible smoke on power up, distinct aroma of burnt electronics, and on close inspection no visible sign of anything burnt on any of the boards, nothing, nada, zilch...

There's no way I would dare put this thing in the air until I know what happened, so I guess there's only one thing I can do, put it back together, plug the power in, and see what goes up in smoke. This time I'll have to let it get hot enough to see what's burning, or be close enough to see where its coming from before I pull the plug. If I can at least narrow it down to which board I can fix it, right now it could be anything. I was actually hoping to see a burnt FET or some component on a BL, I have two spares on the shelf and while it would be a bit of a PITA to change, it wouldn't involve spending any $ to fix.

I'd much rather a plainly visible hole burnt in a component or right through a board to knowing something blew but not be able to find it.

Ken

 

[Bartman]

ken,
i'd take it apart and power the flight control board with the jumper power from the usb port. if that's ok you can then power just the power dist board to see if the bl's flash correctly without anything attached. i wouldn't hook up full battery power after the dreaded plume of smoke without disconnecting things and going through it board by board.
then again if you've been wanting to buy a new MK you could just hook it up and see how it does with 220 v single phase from the house.

 

[RTRyder]

Yeah, but unfortunately it may not show up under low voltage, low current conditions, and if it's a noncritical circuit like the servo ports it may not show up on diagnostics at all. I will try connecting the F/C to the laptop and see what I get and go from there, but I'm very much afraid this may take a full power up to get to the root of the problem. If the F/C or a BL is going to go up in smoke, I'd much rather it do it on the bench and be done with it, then I know what needs to be fixed.

If I can't find a definite reason for the smoke I have no alternative but to send all of the boards on a German vacation to be checked out, otherwise I have a very expensive shelf ornament.

Ken

Update, there is a definite short somewhere on the F/C board, the recoms got hot just being plugged into the USB connection to the laptop, more later...

 

[RTRyder]

So the verdict is in, DO NOT change the servo recom to a 6 volt unit to supply more power to camera mount servos. It appears that there is potential for the 6 volts to get back to the control circuitry side of things and there are components there that do not like more than 5 volts apparently, I now have a short between the 5 volt line and ground across both circuits, so it looks like my F/C board is going to get a trip to the MK hospital in Germany...

Ken

 

[Bartman]

ken,
did you happen to verify that the servos were good for 6 volts before putting the juice to them?
sorry to hear your FC is toast. got your PM too, don't hold your breathe, if you know what I mean...
good luck getting it sorted out.
bart

 

[RTRyder]

ken,
did you happen to verify that the servos were good for 6 volts before putting the juice to them?
sorry to hear your FC is toast. got your PM too, don't hold your breathe, if you know what I mean...
good luck getting it sorted out.
bart

Sure did, I wouldn't have bothered to change the recom if the servos weren't 6 volt digitals. Anyway, to get back in the air in a reasonable amount of time I just ordered a replacement F/C from Adam, I'll deal with the bad one at some point down the road. If I could find a full schematic for it I could probably fix it myself, but good luck getting one of those and even if I did the notations are all in German anyway

As I mentioned I don't expect much to come of what I asked about, if it happens, it happens, if not, oh well, it's not the only game in town! My bad for not thinking it all the way through first I guess. I knew the potential for this to happen was there, just didn't think it was likely to happen if others had done it, now I wonder how many have and are still working.

Ken

 

[Crash]

If I could find a full schematic for it I could probably fix it myself, but good luck getting one of those and even if I did the notations are all in German anyway

The schematic for MK FC 2.x? All of the schematics are available and mostly readable.

If any over voltage came back out the servo, then hopefully IC5 (servo driver) contained the damage.

 

[1001 Copters]

Hi RTRyder,
we just have a discussion about this yesterday in our weekly MK meeting near Paris.
One of has changed his recomS (the S is important..) to 6.5V recoms. He plug the battery and fortunatly the FC board didn' t want to start. He checked the electronic scheme and he saw that there is one RECOM for the tower stack ( IC4 place : FC/navi ctrl/mkmag/mkgps) and one RECOM for the servos(IC3) . The µchip on the FC shouldn't receive more than 6V (according to this datasheet) !

It's a basic question but did you changed both RECOM(IC3 and IC4) or just the servo RECOM (IC3) ?

 

[Bartman]

i'm trying to follow this conversation but, is a RECOM another word for voltage regulator? do the RECOM's reside on the FC board? is that what is putting reduced voltage from our 4s packs to the FC?
thanks,
bart

 

[RTRyder]

Hi RTRyder,
we just have a discussion about this yesterday in our weekly MK meeting near Paris.
One of has changed his recomS (the S is important..) to 6.5V recoms. He plug the battery and fortunatly the FC board didn' t want to start. He checked the electronic scheme and he saw that there is one RECOM for the tower stack ( IC4 place : FC/navi ctrl/mkmag/mkgps) and one RECOM for the servos(IC3) . The µchip on the FC shouldn't receive more than 6V (according to this datasheet) !

It's a basic question but did you changed both RECOM(IC3 and IC4) or just the servo RECOM (IC3) ?

I replaced only the recom for the servo power, I knew the recom for the controller circuits has to remain at 5 volts. I did find the F/C board schematic that is publically available and it's not as detailed as any of the ones I used long ago when I did board repair for a living (we also didn't have SMD devices then either!). From what I can see there should not be any way for the two circuits to connect, but as I said the schematic isn't as informative as it could be.

It did work for a while, I flew it and did some setup work on the bench before it smoked, so I don't understand what happened or how it happened, or how it could have happened. The only thing I can figure out is that there was some feedback of the 6 volts on the signal line from one of the servos and it got out on the 5 volt line for the control circuitry.

The really strange thing is that I can plug the F/C into a laptop, power it via USB, and when I start MKtool the board appears to be working normally right down to the gyro traces on the scope when I move the board around. After a minute or so the recoms are hot to the touch and with the board disconnected I measure a short to ground across both 5 volt output lines from the recoms, so whatever blew doesn't appear to be a critical component but its enough to make the board useless until it's fixed.

Ken

 

[Bartman]

ken,
are the recoms still putting out their rated voltage? did they survive the smoke incident? it sounds like the traces on the board cooked and shorted.
bart

 

[RTRyder]

i'm trying to follow this conversation but, is a RECOM another word for voltage regulator? do the RECOM's reside on the FC board? is that what is putting reduced voltage from our 4s packs to the FC?
thanks,
bart

Yes, the recom is the voltage regulator that takes battery input voltage and drops it to 5 volts. They're the large rectangular components on the underside of the board, the one next to the capacitor is for the controller circuitry and the one on the opposite side of the board from the gyros is for the servo outputs.

Ken

 

[Bartman]

can someone please link to the FC schematic so we have it for future reference. i'm surfing the .de site and can't find it.
maybe i'll start a thread and just list links to common Mikrokopter board pics/schematics.
thanks,
bart

 

[RTRyder]

ken,
are the recoms still putting out their rated voltage? did they survive the smoke incident? it sounds like the traces on the board cooked and shorted.
bart

There are no burnt traces and I don't know if the recoms are working or not, I haven't connected battery power to the board since the incident, I've only applied 5 volts to the other side of the recom circuits through a USB connection.

There is no sign of anything burnt anywhere on the board and I've spent probably close to an hour looking at it under as much magnification as I have available. So far I've checked the servo driver IC and it appears to be OK so far though without powering up the board and an O'scope to look at signals that's only a preliminary evaluation meaning I haven't measured any shorts to ground on pins that shouldn't have ground potential.

This has me baffled at the moment and if I really want to dig into it I'm going to have to setup the proper diagnostic tools to do it right. Not sure it's worth my time and the aggravation right now, I may just ship it off to the fatherland and let them poke at it.

Ken

 

[RTRyder]

can someone please link to the FC schematic so we have it for future reference. i'm surfing the .de site and can't find it.
maybe i'll start a thread and just list links to common Mikrokopter board pics/schematics.
thanks,
bart

Under wiring diagram near the bottom of the page... http://mikrokopter.us/MKDocs/FlightCtrl_ME_2_1.htm

 

[Bartman]

is it feasible to desolder the recoms and bench test them individually? i'm pretty clueless on this stuff.
bart

 

[RTRyder]

is it feasible to desolder the recoms and bench test them individually? i'm pretty clueless on this stuff.
bart

I've already taken the servo recom off the board but haven't tested it. Yes, it is possible and actually fairly easy to test a component like a recom out of circuit and later I may take the other one out and see if there's still a short to ground. If the short goes away then it's likely the 5 volt recom causing it though I can't see the how or why of that being the problem but it's easy and worth a shot.

Ken

 

[DennyR]

For what it's worth I am using 7.4 volts to my servos on the camera mount (DJI system). You need as much speed from the servos as you can get with high res servos. Also I changed to direct drive from the geared mount. This is quite different from my own camera stabiliser which can work with 4-1 drive. The DJI camera outputs can be set to give a very good resolution to the disturbance but the response speed needs to be faster.

I have under development a direct drive torque motor system which will kill off servo drives for good. Much faster response and much more accurate. 50,000 possible positions in 360 degrees. High sensitivity gyros are built into each separate axis. The first planned production unit will be based on a tiny 3 axis mount for the GoPro2.