Hello.I have an old radio that i want to change the 4 rectifier diodes and the 4 caps in parallel with each diode due to oxidization.
Original diodes were DS135D and caps 22nF ceramic.
I will replace the diodes with 1N4007 but for the caps i have 10nF/63V and 100nF/63V ceramic type.
The transformer secondary is 9V AC. Is the voltage rating of my 63V capacitors sufficient? can i use one of those two values instead of 22nF?
Original diodes were DS135D and caps 22nF ceramic.
I will replace the diodes with 1N4007 but for the caps i have 10nF/63V and 100nF/63V ceramic type.
The transformer secondary is 9V AC. Is the voltage rating of my 63V capacitors sufficient? can i use one of those two values instead of 22nF?
Thank you. I will use 10nF then.And what about voltage rating? Is 63v ok to use?
Edit:I found 22nF in my parts but they are also 63V rated.
Voltage rating is my only concern now.The old ones were 100V rated.
Edit:I found 22nF in my parts but they are also 63V rated.
Voltage rating is my only concern now.The old ones were 100V rated.
Last edited:
Thank you. i will use 63v then.
Any info on the voltage of the capacitor that i should generally use?
I understand that this is a simple form of snubber network for minimizing diode switching noise/electromagnetic emision when diode is reverse biased.
Any info on the voltage of the capacitor that i should generally use?
I understand that this is a simple form of snubber network for minimizing diode switching noise/electromagnetic emision when diode is reverse biased.
Unless there's a L input filter, the snubber caps' voltage rating needs to be sqrt 8 times the AC voltage at least.
Best regards!
Best regards!
snubber caps should never be put across the diodes, they will conduct high freq noise from the toroid to the amp
snubbers only before or after the bridge like explained in the quasimodo jig tread
snubbers only before or after the bridge like explained in the quasimodo jig tread
For 50/60Hz power supplies, snubber caps across diodes are simply not needed. marconi118 has it right. HF ringing as the diode cuts off simply isn't an issue. It's either non-existent, or too small to worry about.
Double check how much surge current your filter bank will pull. 1N400x does not have a high peak current capability. Model your supply in Spice and you may be surprised. 30A sounds like a lot, but do the simulation.
I believe Mr. Pass still suggests caps across the diodes to slow them down. Quite the opposite of the current fad for stupid fast diodes. Personally, after reading all I can find and then benching my latest supply, I'll spend my money on better filter caps. I guess if the rectifier was remote from the filter bank, switching could cause some RF, but I like my rectifier and filter bank as close as I can make them. If you are concerned with transformer ringing, then do a snubber at the transformer as was mentioned above, but the only improvement will be a little less noise out the AC line. Not a bad thing.
I believe Mr. Pass still suggests caps across the diodes to slow them down. Quite the opposite of the current fad for stupid fast diodes. Personally, after reading all I can find and then benching my latest supply, I'll spend my money on better filter caps. I guess if the rectifier was remote from the filter bank, switching could cause some RF, but I like my rectifier and filter bank as close as I can make them. If you are concerned with transformer ringing, then do a snubber at the transformer as was mentioned above, but the only improvement will be a little less noise out the AC line. Not a bad thing.
The application is an old ss radio with 9VAC secondary - there is no way that such a supply will push peak diode current anywhere near 30A limit, and the original diode is effectively an 1N4002.
Go forth blind with your logic if you wish. Maybe OK, maybe not. Your filter cap is probably dried out and has ESR in Ohms, so yea, maybe OK. I prefer to model and test.
If designing for release or production, no excuse not to build the test fixture and get correct components. Actually when I look at current production units, most have not bothered which is an indication of their actual care about the customer. Not.
But for the typical DIY one off- way too much work for no audible gain. One can at least toss a scope on it and make some reasonably close guesses. If you GOOGLE a bit, you can even see how to build a cheap DIY RF sniffer to test for EMI.
More distressing is the short supply of mu-metal boxes to shield the transformer or your boards. Open torroids, E-cores without even a band. No wonder and 120 Hz show up only 60 or 70 dB down.
They are not snubbers. They are shunts. Radios need diodes across diodes in a power supply because diodes modulate conductivity to mains wires. Mains wires work like antennas, and radio has a high sensitivity to all possible sources of a signal.
So any rectifying diode in a radio power supply has to be shunted by a capacitor (about >=1 nF) to get rid of this parasitic modulation.
RC-snubber across the secondary may be possibly used too, but it can't work fine without those shunting caps.
Last edited:
I said so times and times again; good to see someone else aware of the issue.
The combination of the wiring around the rectifier and the PIN diode effect of rectifiers results in a beautiful, random-switched array of antenna's and transmission lines modulating EM fields (inbound or outbound) at 100Hz (or 120Hz).
The level of disturbance introduced by diode commutation is dependant on many factors - no one remedial action fixes all situations/designs/equipment.
Although products in the past may have added capacitors in parallel to diodes to pass compliance requirements, or suppress noise ingress to other circuitry/detectors, that action became the simplest remedial fix, but is not the only path that could have been taken.
Just taking the simplest remedial path imho completely misses the root cause for disturbances, and hence what other actions could also lead to the same outcome, or turn out to be a better path to take for particular situations.
Although products in the past may have added capacitors in parallel to diodes to pass compliance requirements, or suppress noise ingress to other circuitry/detectors, that action became the simplest remedial fix, but is not the only path that could have been taken.
Just taking the simplest remedial path imho completely misses the root cause for disturbances, and hence what other actions could also lead to the same outcome, or turn out to be a better path to take for particular situations.
Last edited: