Maker Pro
Maker Pro

Potentiometer Burnout

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
Hey!

So, yesterday I had two 10k pots burn out at [I think] the exact same time, but I don't understand how/why and was hoping some of you might have some insight.

The project is a midi controller based on the Teensy 3.1, powered by USB. I have two banks of five 10k pots (L/R) that are being read by a 74HC4067. There is a ceramic cap [.1uF] between the Teensy and the IC [and each of the banks] across +3.3v > Gnd to provide a small bit of isolation. All are connected to +3.3v and AGnd on the Teensy.

Each side of the controller also has an illuminated 3-pin rocker switch [Vin, Vout, Gnd] which I have set up to apply power to the base of an NPN when it is activated. The NPN connects a Teensy pin with a pullup resistor to ground. I am unfortunately failing to remember clearly offhand if I have the switch in the same loop as the 5 pots, or if I separated it so it is connected directly to the +3.3v and Gnd (not AGnd) of the Teensy. I will check that tonight to be certain.

Anyways, I had everything hooked up and put together and all inputs were reading properly and with the full range of values for about two weeks. I had started into the programming phase of the project on Friday and was playing with the controller all weekend just working on getting the software to behave as I want and at one point I noticed that two of the pots were only reading in the range of 0-3 instead of 0-127. I don't know if they went out at exactly the same time or just very close in time, and because I was working on unrelated code I wasn't checking the pots' output each time I uploaded a modified sketch- so I don't know at exactly what point they went out.

I can easily swap both of them out with spares tonight, but I'm hesitant to do so without knowing the cause of their failure. I don't want to end up with four dead pots instead of just the two I have now. Apart from manufacturer defects, the only way I know of burning out a pot is to run current into the wrong pin or run too much current through the correct pins. I don't see how I could have done the former, since the related code has been working consistently as expected this whole time and returning accurate values, and all the wiring was thoroughly tested and then hot glued together to keep anything from shifting or shorting. But I also don't see how it could have been the latter, either, because there are three more pots on the exact same rails connected millimeters away and they're all fine- so if there was some sort of unexpected overload, wouldn't they be burnt out too?

The reason I mentioned the LED rockers is both because they may be in the same power loop and are located directly next to the pots, and also because early in development one of the first rockers I was using would cause the Teensy to reset when the switch was toggled on. The weird part was that it was dependent on how you turned the switch on. If you flipped it fast and it had more of that 'snap' into the on position- reset. If you eased it carefully from off to on, no reset. I chalked that up to defect since it was only that one switch and just replaced it- which solved the reset problem.
As a side note, I measured and found that the LED in the switch pulls 6mA when hooked up to a 6v battery (meaning R = 1000ohms; 3.3mA = 3.3v/1kohms ...right?) so even if the switch was causing a spike it wouldn't be very large, would it?

The attached schematic isn't the full circuit but just wanted to give a small visual on the parts I've mentioned.


Any thoughts / links / suggestions / help very welcome. Thanks!
 

Attachments

  • potschem.PNG
    potschem.PNG
    10.4 KB · Views: 207

shrtrnd

Jan 15, 2010
3,876
Joined
Jan 15, 2010
Messages
3,876
Can only speculate with what I read.
If the pot is actually 'burned-out', is it possible the current through them was excessive the entire time you were experimenting, and it just happened to finally reach the point of pot failure?
I mention this, in case you're running excessive current through all of the pots, and in replacing the two you lost, you may wind up losing all of them.
Hopefully someone else may have some insight here, or a follow-up question that will help isolate your problem. I, would recheck the circuits carefully, as I would think from the small picture
shown, that more than just your two 'burned-out' pots are at risk here.
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
Can only speculate with what I read.
If the pot is actually 'burned-out', is it possible the current through them was excessive the entire time you were experimenting, and it just happened to finally reach the point of pot failure?

You make a very good point- that definitely could be the case, and is something I had never considered. When I get a chance (hopefully tonight) I will check and see what the current is both through the dead pots and through one or more live pots.

If the current does seem to be excessive, would the solution then be add a series resistor between the pots and the power source, and then also connect that to the AREF pin so the Teensy maintains an accurate reading?
 

Rleo6965

Jan 22, 2012
585
Joined
Jan 22, 2012
Messages
585
Basing on your circuit. 10k should not be burnout unless its center tap was connected to low resistance to ground or connected to higher dc voltage source.
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
Basing on your circuit. 10k should not be burnout unless its center tap was connected to low resistance to ground or connected to higher dc voltage source.
Yeah, see that's what I thought...

By the way- these are the switches I mentioned: http://www.amazon.com/10PC-Round-Rocker-Toggle-Switch/dp/B007B87RIC/

Which from what I can tell, I believe have a built-in resistor for the LED although 10-15 minutes was not able to verify 100% conclusively. All forum posts I found [non-merchant, end users, I think] seemed to agree that there was either a series resistor built in already and/or no additional resistor was needed. BUT I have no issue adding an additional one anyways as a safeguard.

Haven't had a moment to sit down and do any tests this evening but I'm hoping I'll still have a half hour to check it out before I crash tonight.
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Did you see / smell any smoke? If not, maybe they did not burn out, but failed mechanically.


Bob
 

KJ6EAD

Aug 13, 2011
1,114
Joined
Aug 13, 2011
Messages
1,114
My guess, and it's only a guess, is that the I/O lines attached to the wipers were inadvertently changed to outputs in software long enough for a low impedance path to ground or Vcc to create a dead spot in the pot element as Rleo alluded.

For this to happen, the wiper would need to be near end of travel so only a small portion of the element was dissipating all of the energy.

I would not expect the I/O on the μC to fare well under such a load either.
 
Last edited:

shrtrnd

Jan 15, 2010
3,876
Joined
Jan 15, 2010
Messages
3,876
Plus, we don't know the rest of your circuit.
If the circuit allows an excessive increase in current flow through some pots, and a decrease in current flow through the other pots, balancing total current flow trough the circuit:
it's possible some pots could become damaged while others were not.
You keep referencing the LED lighted switches and possible associated resistors. So I assume you're aware that they may be causing issues.
Good luck checking this out.
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
Thanks everyone for your suggestions! Definitely have given me a lot more things to check than I could think of on my own. I should have time to investigate tonight and will report back with an update.
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
Ok, so I finally found some time to do some testing this weekend. It seems that the pots were still functioning as they should have been- meaning they weren't actually burned out as I had thought. They provided accurate results when I checked their resistance with my multimeter and when attached to a different slot on the Teensy it pulled the full range of values.

I removed the IC and checked the connections between the chip socket's pins and the header pins and everything there seemed fine, too. SO the conclusion I've drawn is that somehow those two pins on the 74HC4067 became damaged and are incapable of reading the inputs properly. I have replacements but I wanted to check in here first to avoid potentially frying another IC.

Attached is a rough schematic. Pretty disorganized but should be accurate.

For reference:
-All pots (2 slide, 4 rotary, and 4 more on the joysticks) are 10k
-Transistors at the top controlling the RGB LEDs are all 4k7 NPN BRTs
-Couldn't find an existing part for the two illuminated rockers so I just diagrammed out the components
-The Teensy3.1 part object in Fritzing didn't have any of the bottom pads so I had to edit them on [the really awkward pins on the bottom end of the Teensy]



Any thoughts appreciated!
 

Attachments

  • RabbitBox-v1_schem.png
    RabbitBox-v1_schem.png
    241.6 KB · Views: 208

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
That schematic is too small to be readable...
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I find it hard to follow that schematic... too much like a grid of wires, and not many components. Maybe you could compact it a bit, and move components around so there's not so much zigging and zagging? And group wires together by using different spacings? And keep important signals (AGND for example) as a long horizontal line with no other horizontal lines nearby? I think that would all help the readability.

Also, can you draw in the schematic of the joysticks? That would be a big help.

BTW, the rabbit in your icon looks pretty fierce... I think it has myxomatosis!
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
BTW, the rabbit in your icon looks pretty fierce... I think it has myxomatosis!

Haha, maybe that and/or rabies...

I find it hard to follow that schematic... too much like a grid of wires, and not many components. Maybe you could compact it a bit, and move components around so there's not so much zigging and zagging? And group wires together by using different spacings? And keep important signals (AGND for example) as a long horizontal line with no other horizontal lines nearby? I think that would all help the readability.

Yep I can definitely try to rearrange things so there's less crossover. The connections to the board got all funky when I was actually soldering it together due to very limited space, so I switched which inputs went to which Teensy pins so there was less physical crossover. The result is a very disorganized input/pin distribution as you can see :D

Also, can you draw in the schematic of the joysticks? That would be a big help.

I'm not sure I understand your request. The joysticks are already included in the first [visible] schematic, located on the far left. Physically they look like this- http://www.karlssonrobotics.com/car...3EojvR2_3CRFTw3LAdgh1l6hAjTuswBlK1Z35Q5PD_BwE

Is that what you're asking for, or...... ?

In the meantime I'll see about un-cluttering the rest of it.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Haha, maybe that and/or rabies...
Maybe it's Stephen King's pet rabbit :)
I'm not sure I understand your request. The joysticks are already included in the first [visible] schematic, located on the far left.
I see. I think it would be useful to show the two potentiometers and the pushbutton inside the rectangle that's marked "JOYSTICK". It's pretty obvious from the pin names, but a diagram wouldn't hurt, if it's possible to do that.
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
I think it would be useful to show the two potentiometers and the pushbutton inside the rectangle that's marked "JOYSTICK". It's pretty obvious from the pin names, but a diagram wouldn't hurt, if it's possible to do that.

Absolutely possible! Although as far as I know the only way to be to manually edit the existing part object files and that's really annoying to do... so to make things easier I'll just replace each joystick with 2 discrete pots and a button, and just label them appropriately. Seems like a much more simple solution than editing part schematics and properties :]

I've made some small progress on cleaning up the overall schematic by detaching the input/output pins from the body of the Teensy and instead just attaching the wires to clearly-marked single pins which are located much more locally to the actual inputs/ICs. I'll keep working on that and see if I can get it done in the next hour...
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Absolutely possible! Although as far as I know the only way to be to manually edit the existing part object files and that's really annoying to do... so to make things easier I'll just replace each joystick with 2 discrete pots and a button, and just label them appropriately. Seems like a much more simple solution than editing part schematics and properties :]
Good idea. Can you add a box or a dotted box around them all to indicate that they're part of the same unit?
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
Took me a bit longer than I expected... BUT here is the updated version:

https://drive.google.com/file/d/0Bx0dzEZFL5_VRHR1TWxOSVctc2M/edit?usp=sharing

The blue boxes indicate which components are the combined joystick units, and the green boxes indicate which components are the illuminated toggle units [blue LED, if curious].

As mentioned before, the only way I could figure out how to really clean up the design was to disconnect the inputs from being pinned directly to the Teensy, so now they're each attached to a single pin with the Teensy pin name. Hopefully I didn't mislabel anything, but either way I'm pretty confident that those are actually wired the way they should be because when I was testing with the coding all the inputs were responding. But hey- I wouldn't have needed to start this thread if I really knew what I was doing haha
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Wow, what a huge improvement! It's like you've weeded the garden!

I think your 74HC4067 has its GND and -EN pins connected to +3.3V!

You might want to keep the pushbuttons separate from the analogue stuff - use separate feeds from +3.3V and GND from the Teensy for them, and keep them physically separate, otherwise you might find that pushing the button causes a tiny change in the values you read from the potentiometers.

Other than that, it all looks good to me. I can't suggest why you're having problems, unless it's a construction issue. If you can lay out the construction as tidily as you've laid out that schematic, you should be fine!
 

whiterabbit

Sep 14, 2013
90
Joined
Sep 14, 2013
Messages
90
DERP! When I shifted everything around a couple of the bend points got hidden and I didn't realize...

Also realized that I have the pushbuttons on +3.3V not +5v, so I fixed that too.

https://drive.google.com/file/d/0Bx0dzEZFL5_VZjA0cm5JWDB3R2s/edit?usp=sharing

Honestly the thing's a monstrosity... I'm considering rebuilding it altogether, I just won't have any time the next little bit here because I'm moving, so I was hoping I could get it functioning and move on. If the schematic looks good then either I physically built it with alternate connections, OR I built it the way the schematic is laid out- and just did a poor job on the construction. Entirely willing to admit to either of those, heh. Or just a defect in the chip, but that seems less likely.

If the replacement IC fries I guess I'll know for sure I need to just rebuild the whole thing! At least then it'll be easier now that I've had the experience with this first version.

Also, if anyone is interested- this is what it actually looks like:

https://drive.google.com/file/d/0Bx0dzEZFL5_VNi1sTENHM1Nsb0U/edit?usp=sharing
https://drive.google.com/file/d/0Bx0dzEZFL5_VUWJ0QVdSNVFVMG8/edit?usp=sharing
 
Top