2
\$\begingroup\$

I want to run a ~2W, 12 VDC fan in an EMI-sensitive environment.

The schematic is shown below. Due to the spiky current draw, I intend to place C1 and C2 right by the fan connector, so high frequency current draw across the board's power distribution network is reduced. I don't want large currents in the frequency band from about 1 kHz to 1 MHz drawn across the board because it could interfere with sensitive analog circuitry, and obviously I want to contain >10 MHz currents very locally due to EMI.

When transistor M1 is suddenly and fully turned-off while the fan is still spinning, a back emf current will flow through D1, taking the fan 'out' node and C1 voltage to about -0.7 V.

C1 is a polarized Al electrolytic with 0.5..1.5 Ohm ESR. Is this reverse voltage across C1 a problem for it? That situation can occur multiple times per day.

schematic

simulate this circuit – Schematic created using CircuitLab

I have studied capacitor datasheets of several Al electrolytic capacitors that could fit the role of C1 from several reputable manufacturers. In datasheets of parts from Nichicon, Rubycon, Cornell, United Chemicon, I could not find any discussion of reverse voltage situations. A Panasonic datasheet at least contained a very short hint under 1.3.(1), namely: "[...], do not apply the reverse voltage. Find the correct polarity before insertion.". I am not able to draw any informed conclusion based on these datasheets. If such a situation would always lead to permanent damage, wouldn't all the manufacturers feel obliged to add a bold hint in their datasheets ?

\$\endgroup\$
9
  • \$\begingroup\$ You haven't spec'd what type of capacitor C1 is. If it's a ceramic, for example, then there's no problem. \$\endgroup\$
    – TonyM
    Commented Oct 21, 2021 at 7:40
  • \$\begingroup\$ Change D1 to Schottky? \$\endgroup\$
    – winny
    Commented Oct 21, 2021 at 9:26
  • \$\begingroup\$ You need to check the datasheet for the capacitor but lower negative voltage should be lower stress overall. If the datasheet limit is 0.5 V than you move from over stress to within limit. \$\endgroup\$
    – winny
    Commented Oct 21, 2021 at 9:38
  • \$\begingroup\$ @winny Is there some important threshold that makes -0.4 V much safer than -0.7 V? I have added more information found in several datasheets. It wasn't of much use tbh. \$\endgroup\$
    – tobalt
    Commented Oct 21, 2021 at 11:40
  • \$\begingroup\$ You have a bigger problem, your 100µF capacitor is surging through your mosfet at turnon. You don't really need it anyway \$\endgroup\$
    – Lorenzo Marcantonio
    Commented Oct 21, 2021 at 12:51

1 Answer 1

Reset to default
-1
\$\begingroup\$

This reverse voltage under 1 volt is not a problem for capacitors. I already make a test on 16 V capacitor with one volt reverse and result is no current flow and the capasitor not affected. Usually one and half volt is limit of capacitors.

\$\endgroup\$
9
  • 1
    \$\begingroup\$ Thank you. Could you detail a bit what "my experience" means ? Have you used capacitors in this manner for an extended period of time? \$\endgroup\$
    – tobalt
    Commented Oct 21, 2021 at 8:28
  • 1
    \$\begingroup\$ This is an opinion, not an answer with a definitive explanation. Downvoting until improved. Might be a new user innocently using 'Answer' on this Q&A site like 'Reply' is used on discussion forums, so could be deleted and reposted as a comment. \$\endgroup\$
    – TonyM
    Commented Oct 21, 2021 at 8:39
  • 1
    \$\begingroup\$ I see a lot of circuits use diode like this to protect capacitor and i already test it for you i put 1 volt in reverse across 16 volt capasitor for five minutes and nothing happen \$\endgroup\$
    – Omar ibrahim
    Commented Oct 21, 2021 at 8:41
  • 1
    \$\begingroup\$ That's still opinion, I'm afraid, and doesn't answer the question. A subjective test on one part for a few minutes doesn't remotely prove a part will work within specifications for its life across full temperature and humidity and at full voltage range. It just shows 1 part didn't have a catastrophic failure. Walking across a road with your eyes shut ten times doesn't prove roads are safe. Your answer should specify capacitor technology with reference to its datasheet. Good design is not designing something that works - that's relatively easy. It's designing something that never doesn't work. \$\endgroup\$
    – TonyM
    Commented Oct 21, 2021 at 9:02
  • 2
    \$\begingroup\$ "Usually one and half volt is limit of capasitor" Is this a reliable figure due to the involved chemistry or something ? \$\endgroup\$
    – tobalt
    Commented Oct 21, 2021 at 9:08

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