7
\$\begingroup\$

If I feed DC into a bridge rectifier, it still outputs DC right? Also, it seems like it should make the DC inputs polarity agnostic, right?

I'm asking because a friend has a programable timer/relay device that takes 12/24V AC or DC as input, the rest of the controller board is likely operating on DC. So this got me wondering what circuit would allow this AC/DC input.

\$\endgroup\$

5 Answers 5

Reset to default
16
\$\begingroup\$

Output polarity of the bridge rectifier remains unchanged irrespective of the input polarity.

enter image description here

Hence it may be used as a 'reverse polarity' protection device while powering a circuit, should the 1.4 V drop introduced by it be permissible.

\$\endgroup\$
1
  • 4
    \$\begingroup\$ [Off-topic] Altough this could definitely be a reverse polarity protection, there are better/cheaper ways of doing it. You could use germanium or schottky diodes to reduce voltage drop. If "polarity agnosticism" isn't needed, you could employ a single diode, or use a MOSFET for the protection. It will all depend on the reverse voltage and forward current. \$\endgroup\$
    – Filipe Nicoli
    Commented Feb 21, 2022 at 14:13
9
\$\begingroup\$

If it is a conventional bridge rectifier, 4 diodes with RC network. You are supposed to get the same DCV-1.4V(considering constant 0.7V per diode drop) at the output. Only two of the diodes shall be always ON and the C behaves as a open circuit for DC steady state.

\$\endgroup\$
2
  • 9
    \$\begingroup\$ And those two diodes will be conducting all the current instead of the half on average for AC, so the diodes' current rating needs to be sufficient. \$\endgroup\$
    – Andrew Morton
    Commented Feb 20, 2022 at 11:08
  • 4
    \$\begingroup\$ Note, there are diodes with different drop than 0.7 so always be sure you're buying what you think you are, or that the rest of your circuit does not care. This note is especially important for student's projects, trust me. \$\endgroup\$
    – Mołot
    Commented Feb 20, 2022 at 14:47
5
\$\begingroup\$

Using a bridge rectifier in a supply line allows it to be powered from AC or DC.

Sometimes this is used for convenience, sometimes as input protection. The argument is:

  • the electronic item can work unharmed if the customer/installer accidentally connects the power supply backwards

Unfortunately, there are plenty of downsides to that with a DC supply.

  • Drops in power diodes are typically 1.1 V at reasonable currents, not 0.7 V. So 2.2 V of supply voltage is lost. And (2.2 V x Iload) W power is dissipated as heat in the bridge rectifier.

  • The item's DC ground (0 V) is now raised by 1.1 V, so ground loops are easy to create. Like where item connector grounds going to other equipment that runs off the same PSU and using PSU's 0 V as ground.

  • Noisy inductive loads, such as DC motors, driven through supply diodes then produce much worse EMI through harmonics. It is considerably harder to get such equipment through approvals.

Other protection methods (e.g. inline polyfuse, reverse diode across the supply) make it clear to the customer/installer that it's wired backwards but incur hardly any losses when powered correctly.

In commercial equipment, it is pretty rare for a customer to ongoingly connect lots of units backwards. So it's not worth hitting the cost and power efficiency of every unit made. But I had a hard time getting that argument across to clients when I've seen them do it.

\$\endgroup\$
4
\$\begingroup\$

I'm asking because a friend has a programable timer/relay device that takes 12/24V AC or DC as input, the rest of the controller board is likely operating on DC. So this got me wondering what circuit would allow this AC/DC input.

A bridge rectifier is the most likely circuit used to do this, though a single diode could also do it if half-wave rectified AC is sufficient (the DC input would then need to have the correct polarity).

Any device that works on DC internally must have a rectifier to operate from AC. In your friend's case the device is specified to work on both AC and DC so it must be designed to do so. However in the general case where operation from DC is not specified a device could have problems because:-

  • On AC the peak voltage is ~1.4 times the rms value, so the DC voltage may need to be up to 1.4 times higher.

  • On AC the current is shared between all 4 diodes in the bridge so each one only has to handle half the average current. On DC the current only goes through 2 diodes, which should both be rated for the full continuous load current.

  • Some devices generate different DC voltages by feeding the AC wave through capacitors before rectifying it. This might be done to produce positive and negative supplies, or to boost the voltage. On DC this this won't work because there is no wave.

  • Some devices use the mains frequency for timing purposes.

\$\endgroup\$
2
  • \$\begingroup\$ A recent Q&A has answers pointing out various possible issues with the opposite problem: feeding DC peak voltage to devices designed to from AC: Using 336V HVDC at home including current going through only one pair of diodes, and more importantly fuses and PFC issues. Some of which aren't as relevant for devices running from lower voltage, although current handling for the rectifier diodes may be more relevant at lower voltage (and thus more current for the same power). \$\endgroup\$
    – Peter Cordes
    Commented Feb 21, 2022 at 1:57
  • \$\begingroup\$ @PeterCordes Wow! That's spectacularly dumb. My answer relates only to low voltage devices that are known to run on DC internally. I left out the possibility of having a transformer or induction motor because that is unlikely, and of course switch and fuse arcing aren't generally a problem at low DC voltages. Diode current handling is just as relevant at any voltage, since the designer probably wouldn't specify an 'overrated' (for AC) diode that cost more. Anyway as always, if in doubt - don't! \$\endgroup\$
    – Bruce Abbott
    Commented Feb 21, 2022 at 4:48
3
\$\begingroup\$

Connecting DC to diodes bridge makes the circuit indifferent to supply polarity. Makes the connection easier. But keep in mind if connecting AC, voltage should be lower 1.4 times.

\$\endgroup\$
4
  • 2
    \$\begingroup\$ Lower 1.4 times? I think you mean 1.4 volts. If you mean 1.4 times to get RMS voltage from an AC circuit, you cannot have an output reduced by 1.4 times (which would be 140% reduction). I think you meant, reduced by 0.3 times (that is, 70% of the peak voltage). \$\endgroup\$
    – GT Electronics
    Commented Feb 20, 2022 at 18:57
  • \$\begingroup\$ I think ‘reduced by a factor of 1.4’ would be the least confusing way to put it. When I was at school ‘lower 1.4 times’ didn’t have a well-defined meaning, that may have changed. Nevertheless, 12VAC rectifies to around 15 or 16V, but it seems the equipment in question is rated to operate at 12VDC. \$\endgroup\$
    – Frog
    Commented Feb 20, 2022 at 19:17
  • \$\begingroup\$ If the AC RMS voltage is X, then AC peaks are 1.414 X or more accurately, X * sqrt(2). If you want to invert that and call that 0.707, sure. \$\endgroup\$
    – Harper - Reinstate Monica
    Commented Feb 21, 2022 at 1:38
  • 2
    \$\begingroup\$ Depends what you're powering, whether the AC RMS or AC peaks were relevant. For a switching power supply, the AC peaks are more relevant, so yes, lower AC RMS voltage than the DC. A recent Q&A has answers pointing out various possible issues with the opposite problem: feeding DC peak voltage to devices designed to from AC: Using 336V HVDC at home including current going through only one pair of diodes, and more importantly fuses and PFC issues. \$\endgroup\$
    – Peter Cordes
    Commented Feb 21, 2022 at 1:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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