Electronics > Beginners

Why usually there is no diode across a DC choke filter?

(1/8) > >>

RoGeorge:
I don't remember seeing any flyback diode across big DC inductances (when a big series L is used as a DC choke filter), but I'm tempting to add a 1N4007.  In my case the L is 1mH and 0.5\$\Omega\$, and the DC can be disconnected suddenly (the connector can be manually unplugged).

Is there any reason for not adding a snubber diode, any side effect that I am missing, such that adding a snubber diode across a DC choke filter would be a bad idea?

PGPG:

--- Quote from: RoGeorge on November 11, 2024, 12:36:43 pm ---Is there any reason for not adding a snubber diode, any side effect that I am missing, such that adding a snubber diode across a DC choke filter would be a bad idea?

--- End quote ---
In my 12V supply I have 18V transil at input. During Surge pulse (I assume 25A, 50us) transil limits pulse to about 25V. So at choke one side I have for a moment 25V while at the other still 12V. Surge is short enough to not make 12V to rise more than 1..2V (thanks to L and C). So surge is filtered out.
Diode across choke will perfectly ruin filter functionality so adding it is certainly very bad idea.

I have never met a situation like you describe when current in filter choke can be suddenly disconnected. If someone located choke in one device while capacitor in other so they can be suddenly disconnected than I would say it is very bad design. Instead of adding diode across choke add capacitor on the choke side of disconnection point.

RoGeorge:
This one is for a PC amplifier.  The audio amp is powered from the PC's power supply, so the DC is already filtered.  I've added the choke because lately the CPU background noises become too strong, enters into amplif, and it become audible in the speakers.

By adding the 1mH/0.5\$\Omega\$ coil in series with V+ (outside of the PC case, between +12V from the power supply and the +V of the power amplifier), the CPU noises became inaudible, which is good, but the coil is quite big, made on the core of a miniature transformer.  I've added a diode anyway, in parallel with the coil, for just in case I forget and unplug the amplifier while powered.

zilp:

--- Quote from: RoGeorge on November 11, 2024, 10:00:53 pm ---By adding the 1mH/0.5\$\Omega\$ coil in series with V+ (outside of the PC case, between +12V from the power supply and the +V of the power amplifier), the CPU noises became inaudible, which is good, but the coil is quite big, made on the core of a miniature transformer.  I've added a diode anyway, in parallel with the coil, for just in case I forget and unplug the amplifier while powered.

--- End quote ---

Can you explain what bad thing you think would happen without that diode?

RoGeorge:
When disconnecting a wire, there is some contacts chattering.  Because of the coil, that disconnecting chattering can produce spikes of very high voltage.  In practice, the spikes can be of many hundreds of volts, or even more, depending of how fast the current turns off (v=L*di/dt).

Since choke filters with a diode are rather rare, did a simulation out of curiosity.



Circuit 1 is without diode, and circuit 2 is a copy, but with a flyback diode added across the choke, both circuits side by side, to easily compare their behavior.  Rload represents the audio amplifier that is powered from PC, and Rhand is presumably the resistance of the hand that grabs the wires during unplugging the power cable of the audio amp.

Notice how the voltage spikes without a diode can get very high (Vhand1 is in the megavolt range in this simulation), while the spikes with the freewheeling diode attached (Vhand2) are less than 50V.

Navigation

[0] Message Index

[#] Next page