 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Microcontroller and PC projects : Destructive PIC32 Testing
| Author | Message | ||||
Geoffg Guru  Joined: 06/06/2011 Location: AustraliaPosts: 3282 |
Out of interest I tried some extreme tests on a PIC32MX470 (64-pin) to see if it could be damaged by incorrect voltage inputs. Supply Voltage: The chip worked fine up to 5V (as far as I wanted to go), the program kept running and the chip did not draw any extra current. With negative voltages it drew a lot of current which I limited to 200mA and it survived that. Input Pin (both 5V tolerant and not): It worked up to 5V with no current drawn by the input pin however with a negative input the pin drew a lot of current which I limited to 200mA. Input Pin (intermediate voltages): I also experimented with a wide range of input voltages between 0V and 3.3V to see if there was any additional current drawn at intermediate logic levels (ie, neither on nor off) and there was nothing. Input Pin with Vss and Vdd shorted: This was a test to see if applying a voltage to an input would cause damage when the chip was not powered. No current was drawn by the input pin and no damage was caused. Output Pin shorted to Vss or Vdd: No damage was caused. Vusb Pin with Vss and Vdd shorted: This was why I started the tests in the first place - to see if plugging an unpowered MX470 into a USB host would cause damage. The result was that no current was drawn via the Vusb pin and no damage was caused. The bottom line is that my MX470 is still working despite some significant abuse. A Microchip engineer once told be that they put a lot of effort into the "input cell" as they called it and it seems to have paid off. Geoff Geoff Graham - http://geoffg.net |
||||
mikeb Senior Member  Joined: 10/04/2016 Location: AustraliaPosts: 174 |
Thanks for the feedback Geoff. What about 'divide by zero'. Grogster won't be happy. No Smoke. 240V AC should fix that. Gizmo's turn now with the bigger hammer. Zonker can flash his LED no matter what. Thankyou for a sensational product AND your continued efforts. There are 10 kinds of people in the world. Those that understand binary and those that don't. |
||||
TassyJim Guru Joined: 07/08/2011 Location: AustraliaPosts: 6269 |
They don't like lightning. 
Should I make a warranty claim? Jim VK7JH MMedit |
||||
Grogster Admin Group  Joined: 31/12/2012 Location: New ZealandPosts: 9590 |
Yeah, never thought of that - I will hook one up to the mains and let you know what happens.  Smoke makes things work. When the smoke gets out, it stops! |
||||
| mikeb Senior Member Joined: 10/04/2016 Location: AustraliaPosts: 174 |
@geoff Please allow the banter without thinking we are making light of your efforts. It was great information and much appreciated. @TassyJim Lame. I expected more. @Grogster Just for you - Let the smoke out  There are 10 kinds of people in the world. Those that understand binary and those that don't. |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9590 |
....it's a pretty orange....  Smoke makes things work. When the smoke gets out, it stops! |
||||
| piclover Senior Member Joined: 14/06/2015 Location: FrancePosts: 134 |
Be sure to warn your neighbours first... You never know how many blocks of houses will be impacted by mains power experiments...
On a more serious note, it's an interesting experiment you did there... Kudos to you... and to the Microchip engineers who indeed did a great job. Note that, about letting the input pins internal clamping diodes take care of over/under-voltages (e.g. for "dumb" (resistors only) serial interface with the RS-232C port of a computer), this can result in inaccuracies when a diode is solicited while the PIC performs an ADC conversion on another pin (I think that's mostly because the internal clamping diodes are in fact using the chip substrate, introducing a voltage offset to the "ground" when the diodes are solicited). That why, in my designs, I always clamp such "abused" pins with external Schottky diodes (BAT85 or equivalent). |
||||
| mikeb Senior Member Joined: 10/04/2016 Location: AustraliaPosts: 174 |
@piclover I've come across the problem before when the clamping diode rectifies 'RF' noise on the input pin and adds a bias to the signal. I've always used filter components on my analogue inputs to prevent this noise getting there in the first place. These are usually in the form of an encapsulated, 3 pin 'T' filter made up of two ferrite beads, in series, with a capacitor to ground from the centre tap. 
There are 10 kinds of people in the world. Those that understand binary and those that don't. |
||||
| Phil23 Guru Joined: 27/03/2016 Location: AustraliaPosts: 1667 |
Oh, Come on! Surely there's better than that. At least something like this.... Cheers. |
||||
| Phil23 Guru Joined: 27/03/2016 Location: AustraliaPosts: 1667 |
Not dismissing the original post at all. It's good to see these results, as in our/my enthusiasm, frankly, I/we make stuff-ups. It's good to know that the chip will tolerate 5v if it's plugged into the wrong rail, or a non 5v tolerant pin is used when it shouldn't be. Do things like disconnect the power when other powered inputs are still connected. Thanks for the effort Geoff. That said, Some peoples ideas of fun is interesting. Cheers. |
||||
| Geoffg Guru Joined: 06/06/2011 Location: AustraliaPosts: 3282 |
You assume clamping diodes are used for both over/under-voltages. If you look at the test results you can see that they use a diode for negative voltages but not for positive voltages. I applied 5V to an input pin with Vdd tied to ground and measured no current into the pin (actually less than 1mA as the meter has a resolution of 1mA). This means that they cannot be using a simple clamping diode for over voltage. Geoff Geoff Graham - http://geoffg.net |
||||
| Geoffg Guru Joined: 06/06/2011 Location: AustraliaPosts: 3282 |
Why? I tested for that and nothing happened. What I am saying here is that a lot of commonly held beliefs (like that) do not appear to apply to modern chips like the MX470. Geoff Geoff Graham - http://geoffg.net |
||||
| piclover Senior Member Joined: 14/06/2015 Location: FrancePosts: 134 |
True. As far as I could experience it myself so far, only under (negative) voltages (seen on a "dumb" RS232C interface, the voltage from the PC sometimes reaching -12V there, and the clamping diode having to sink the current that goes through the resistor protecting the pin) seemed to cause errors in ADC readings on other pins. I guess I could spare the external over-voltage clamping diode, but I don't feel comfortable to apply +12V (or even +15V, if to strictly follow the RS-232C specifications) to a pin which nominal max voltage is +5V, even if Microchip did a great job, obviously. |
||||
| Chris Roper Senior Member Joined: 19/05/2015 Location: South AfricaPosts: 280 |
One pin you do have to be careful of is the reset pin. In one of our designs we were having boards fail at random with a dead CPU. We finally tracked it down to the reset pin. It was being controlled by the power supply to keep the controller in standby till the supply had stabilized and the design was faulty. Even though the rest of the device survived, you can't do much with it if it won't release from Reset due to a blown pin. Cheers, Chris http://caroper.blogspot.com/ |
||||
| Phil23 Guru Joined: 27/03/2016 Location: AustraliaPosts: 1667 |
"Do" was meant to be "Doing" What I mean't, is we do these things that you tested for by accident, the wrong connections. Then are often left wondering if these simple mistakes have damaged the chip, when things aren't working the way we expect. The suspicion is that a chip has been damages, when in fact our problems are generally related to other inadequacies in our hardware & software construction, which are the true cause of any issues. Cheers |
||||
| Phil23 Guru Joined: 27/03/2016 Location: AustraliaPosts: 1667 |
Did my own Destructive testing yesterday. Inadvertently. Had the Teenager with phone attached move the ceiling heat controller to a smaller bread board. Checked his work & appeared Ok, & it fired up immediately. 10 Minutes later I noticed it was frozen.... 5V Reg on the PS rather hot, & the MX170 hot too. Resets didn't help, it appeared it was dead. Rechecked his circuit & found he'd connected pin 13 to 5V instead of 3.3V. Gave it 5 minutes to cool down & it came back to life. Any YES, it was the Teenager attached to the phone, not passing the buck. Even though I did "Check". Cheers. |
||||
bigmik Guru Joined: 20/06/2011 Location: AustraliaPosts: 2949 |
Lads, I have a 28pin 170 that I plugged upside down when I went to program it.. Got VERY hot, I let it cool down and plugged it in and programmed it OK, I have tested it with my MuP-Test board and it passes but I don't dare use it on anything.... it has masking tape on it with warnings until that dark day that I run out of chips. Kind Regards, Mick Mick's uMite Stuff can be found >>> HERE (Kindly hosted by Dontronics) <<< |
||||
 Print this page |
 |
The Back Shed's forum code is written, and hosted, in Australia. | © JAQ Software 2025 |
