Connecting a capacitor in parallel with each diode in a FWBR? 1

[scoobyrollz]

What is the purpose of adding a capacitor in parallel with each of the 4 diodes in a full-wave bridge rectifier? Does it help with noise rejection, or what? Any help will be greatly appreciated.

 

[Skogsgurra]

They are there to reduce HF noise caused by the reverse recovery current being snapped off when the carriers in the diodes are depleted.

Yes, I know, it could have been better phrased. But that is what they are for.

If you google "reverse recovery" you will find curves showing how the tiny reverse recovery current goes to zero very suddenly - in nanoseconds or less. This sharp current edge produces high frequency noise. It is shorted out by the capacitors.

Gunnar Englund

http://www.gke.org

 

[scoobyrollz]

I am trying to design HV diode stacks for T/R sets. I was wondering about when adding an RC compensation network, if I should match the line frequency or the reverse recovery time of the diode? Also should I use T = RC and then f = 1/T or f = 1/(2*pi*RC) to calculate the frequency characteristics of the RC network? Any help will be greatly appreciated.

 

[Skogsgurra]

There's no connection between line frequency and RC corner frequency. None at all.

The RC network's corner frequency (the -3 dB frequency) is 1/RC when expressed in radians/second or 1/(2*PI*RC) when talking Hz.

If there is no HF pollution concern (might be in R mode) then you do not have to care about capacitors. On the other hand, if RF pollution is a concern, make them one order of magnitude bigger than the diode zero volt capacitance. Use HF capacitors. Ceramic and such stuff. No R, I would say. SCR:s need R but not diodes.

Gunnar Englund

http://www.gke.org

 

[logbook]

You shoud try to use a "soft recovery" rectifier if you need to minimise the reverse recovery spikes. This is a better alternative (if they are available) than creating spikes and then suppressing them.

 

[CarlPugh]

scoobyrollz

Are you designing a high voltage power supply where there are 4 diodes in series in each leg? (If high power, there would be 4 legs/16 diodes if single phase, and 6 legs/24 diodes if three phase-USUALLY-)
Where there are more than one diode in series, capacitors and resistors are used to control the DC and AC voltages across each diode.
The resistance of the resistors depends on the DC leakage current of the diode.
The capacitance of the capacitors depend on the stray capacitance of the rectifier assembly to ground.

Another approach is to use transient voltage suppressors instead of resistors and capacitors.

 

[Skogsgurra]

Yes, what Carl says makes a lot of sense. Capacitors parallel to the diodes in a normal bridge rectifier are for HF reduction. If you have several diodes in series for high voltage, then you need diodes for voltage distibution. The R:s are then parallel to the C:s - not in series.

If this is the case, avalanche diodes are more suitable. So, which one of these applications is it you are designing for?

Also; do not post same question in several fora. It only brings confusion. And it is against site policy.

Gunnar Englund

http://www.gke.org

 

[scoobyrollz]

The application is diode stacks for T/R sets. These convert 480Vac to 50kVdc. So there will be multiple diodes in series, the amount I couldn't tell you since I haven't designed it yet.

My question is really for understanding the purpose of the capacitor and how to calculate the capacitance value to be used in the circuit. Yes, each diode will have a resistor and a capacitor in parallel with it.

According to Carl I should calculate the reactance for the cap. Any explanation of an example for a better understanding would be extremly helpful.

Sorry for the confusion with the posting . . . my mistake.

 

[CarlPugh]

Suppose you have a string of resistors with one end of the string connected to ground and the other end of the string connected to 60 KV RMS, 60 hertz.
There will be stray capacitance from each resistor to ground. If the impedance of the stray capacitance is less than the resistance of the resistor, there will be unequal voltage across each resistor. The resistor connected to the 60 KV will have much higher voltage across it than the resistor connected to ground.
To correct the voltage unbalance, a capacitor with much larger capacitance than the stray capacitance is connected across each resistor.
To really understand what is being described, you should draw the circuit and do some calculations.
The same thing also happens with a rectifier string, except that the calculations are considerably more difficult.
Avalanche diodes as suggested by Gunnar are sometimes used (instead of resistors and capacitors)for low current high voltage rectifiers and the use of transient voltage suppressors (instead of resistors and capacitors) is probably a more modern approach.
You may discover that is is lower cost to purchase a complete rectifier assembly rather than design and build it yourself.
Good Luck, it sounds like a fun job.
Carl Pugh

 

[VEBill]

I've posted my understanding of diode stacks and parallel RC to your other thread "How to calculate RC comp network for diode stack?" [Which is why a single thread is better...].

Long and short of it is to swamp out diode-to-diode variations in the reverse characteristics.