0
\$\begingroup\$

I'm trying to confirm a fault on an electronic unit of which no documentation is available, I can't even tell which MCU it is so it's largely guesswork so far. It's an automotive control unit and let me describe the circuit as I understand it.

I have a simple switched ground input from the car which runs a control unit. It can be assumed the input pin on the MCU has a pull up. During the normal operation of car, the switch is normally open until the switch is pressed. Now the vehicle fault doesn't recognize when the switch is opened anymore, therefore I can assume the MCU input is connected to ground or thereabouts. (Short circuit?)

The unit has since been removed, and is not powered. When the resistance is checked between the ground rail and input pin - it measures 25 ohms resistance, which would be a low input.

My question is this: Does the 25 ohm resistance indicate anything? Is it typical for a rested state input pin?

Or can I conclude that it has shorted internally providing there is no physical short circuit on the PCB?

\$\endgroup\$
  • \$\begingroup\$ I'm really not able to follow your description of the circuit. Could you make a circuit to clarify, please? The question editor has a circuit editor built in :) \$\endgroup\$ – Marcus Müller Jun 27 '17 at 21:51
  • \$\begingroup\$ Could be the switch. \$\endgroup\$ – Scott Seidman Jun 27 '17 at 22:27
  • \$\begingroup\$ Hmm I can't seem to find circuit editor on mobile, I will try next time i am at a PC. But if it helps, what I mean is the GPIO is just a wire to ground with a switch inbetween? \$\endgroup\$ – Martin Pearce Jun 27 '17 at 22:27
  • \$\begingroup\$ The switch is fine, it was disconnected and still had low input. \$\endgroup\$ – Martin Pearce Jun 27 '17 at 22:28
  • \$\begingroup\$ When you measured with your resistance probe, was the + probe to the pin and - probe to GND? \$\endgroup\$ – Tom Carpenter Jun 27 '17 at 23:33
0
\$\begingroup\$

i've now tested other IO pins on this board that have the same circuit (a switched ground input) and they all have high resistance compared to the 25 ohms of the suspected faulty one.

Good test. Assuming nothing different was attached to the "problem" and "working" pins at the time of the testing, then that different result is important.

When the resistance is checked between the ground rail and input pin - it measures 25 ohms resistance [...] can I conclude that it has shorted internally providing there is no physical short circuit on the PCB?

Yes, after your testing of other equivalent GPIO pins using the same test equipment (I guess a DMM) and getting different results between "problem" and "working" pins, then you can conclude the "problem" pin is "shorted internally" (i.e. an unexpected low-resistance path now exists).

(There are concerns about some DMM test voltages, and the possibility for them to cause problems, or for people to believe there are problems when a DMM makes internal ESD structures conduct. However in your case, the different measurements between "normal" and "problem" pins suggest that the "problem" pin really is different, and this different result is not a side-effect of the measurement process.)

Many years ago, on a device with no additional external ESD protection between a keypad and MCU, I had a similar problem - except in my case, the GPIO pin was a low-resistance to Vcc rather to ground, but it's basically the same situation i.e. low resistance to a power rail. You can read more about it in my answer to the question: "What causes, and can you fix “stuck” gpio pin?"

\$\endgroup\$
0
\$\begingroup\$

My question is this: Does the 25 ohm resistance indicate anything? Is it typical for a rested state input pin?

Not really. It does demonstrate that the pin is probably still connected to the chip internally.

enter image description here

Figure 1. There's a lot going on behind a GPIO pin. Source: Mosaic Industries.

\$\endgroup\$
  • \$\begingroup\$ Thanks, that helps me understand the internal circuit a bit better, although I seem to get different results on other IO pins. \$\endgroup\$ – Martin Pearce Jun 29 '17 at 17:35
  • \$\begingroup\$ Try and find a diagram of the actual transistor arrangement. It's much more complex than my little neat boxes suggest. In any case, there's enough in my block diagram to show that a resistance measurement is as much use as a chocolate teapot. \$\endgroup\$ – Transistor Jun 29 '17 at 18:23
  • \$\begingroup\$ I will do thanks, if for no other reason than I would just like to know how it works. On the other hand, i've now tested other IO pins on this board that have the same circuit (a switched ground input) and they all have high resistance compared to the 25 ohms of the suspected faulty one. \$\endgroup\$ – Martin Pearce Jun 29 '17 at 19:49
  • \$\begingroup\$ I've also found that the MCU is a Renasas M306N4FGTFP \$\endgroup\$ – Martin Pearce Jun 29 '17 at 19:51

Not the answer you're looking for? Browse other questions tagged or ask your own question.