1
\$\begingroup\$

I have an AT89C4051 micro (powered with 5VDC) in which the reset pin is pulled down via a 100K resistor (formerly 39K) and pulled up via a 4.7uF tantalum capacitor. For simplicity, let's say a normally-open button is added in parallel with the capacitor even though one isn't there.

I hooked the positive lead of a voltmeter to the reset pin and grounded the negative lead of the voltmeter.

When I hold down the button, the results that were expected appeared (around 4.9VDC), however, when I let go of it, the voltage stays at about 2.6V. I was expecting the meter to slowly count down to 0V but it doesn't unless I remove a voltmeter lead from the circuit.

The only thing I could think of is that the insides of the capacitor are somehow broken, but I can't tell from the outside since nothing blew up on the board. I did measure the resistance across the capacitor and the meter stayed at -1 which suggested to me that no short circuit (yet) exists in the capacitor.

What can I do to correct this issue?

|improve this question
\$\endgroup\$
  • \$\begingroup\$ It behaves as if the capacitor is in reverse with that much pullup leakage against the internal pulldown 50~300 KΩ.. Mind you some Tantalums are rated at max(0.01CV or 0.5μA, ) or about 10MOhms. ESD damage can be another factor. \$\endgroup\$ – Sunnyskyguy EE75 Oct 7 '17 at 2:46
  • \$\begingroup\$ Why did you elect to use 100k or 39k pull-downs when everyone is using 8k-10k resistor for this purpose? \$\endgroup\$ – Ale..chenski Oct 7 '17 at 4:14
  • \$\begingroup\$ the datasheet for an AT89C4051 chip indicates that the pull-down resistance should be between 50 and 300k. \$\endgroup\$ – user152879 Oct 7 '17 at 4:41
0
\$\begingroup\$

Tony is probably correct in his comment about voltage reversal- note that SMT tantalum caps have the polarity bar on the positive side, rather than the negative (which is common with most other electrolytic capacitors), and I have seen that cause confusion before. If the cap has been reversed you should replace it with a new one.

If you do not have a series resistor to the reset input, shorting the MCU power supply will discharge the cap through the input protection diodes, potentially permanently damaging them. In the future, a few hundred ohms is sufficient if this is a possibility.

|improve this answer
\$\endgroup\$
  • \$\begingroup\$ I do have a 47nF ceramic capacitor. If I replaced the tantalum with a ceramic would the reset function properly? or is 47nF too low? Maybe tantalums are just not a good capacitor? Idk \$\endgroup\$ – user152879 Oct 7 '17 at 4:45
  • \$\begingroup\$ RC circuits are crap for resetting a micro and you shouldn't use them at all, but that's not what you asked in the question. You should use a proper supervisor chip which will always generate a nice long reset pulse regardless of the ps rise time or brownout conditions or brief dips. Anyway, 100K and 47nF is a time constant of 4.7ms which is pretty short. Something like 100ms would be better, so a 1-2uF ceramic. \$\endgroup\$ – Spehro Pefhany Oct 7 '17 at 9:54
  • \$\begingroup\$ For the power, I'm using 5VDC regulated (from the output of 7805 IC). Sadly I don't have a ceramic larger than 47nF on hand right now. I'll try 47nF and see what happens. \$\endgroup\$ – user152879 Oct 7 '17 at 21:29
  • \$\begingroup\$ Ok, I did use the 47nF and 100k resistor and things work now. I guess I internally blew up the 4.7uF but the circuit board didn't show it. I think I'm gonna call it quits with tantalum capacitors. \$\endgroup\$ – user152879 Oct 7 '17 at 23:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy