Hi ! looking inside some amps even quite powerful ones i see usually discrete diodes for rectification but rarely oversized Actually the contrary
I'm afraid I've developed a fetish for electronic parts to the point that this is my favorite ornament i bought recently
i wonder if there is any problem to take the oversizing route I think that the bigger the diodes the less the resistance ?
i am trying to get the impedance in the power supply rails as low as possible
After all these bridges are quite cheap and also easy to use
Do you prefer discretes maybe ? why ?
I'm afraid I've developed a fetish for electronic parts to the point that this is my favorite ornament i bought recently
i wonder if there is any problem to take the oversizing route I think that the bigger the diodes the less the resistance ?
i am trying to get the impedance in the power supply rails as low as possible
After all these bridges are quite cheap and also easy to use
Do you prefer discretes maybe ? why ?
Many amps put a .01 uf 1000 v ceramic cap parallel each diode in the main bridge. This slows the shut off transient down and causes lower frequency RF to be transmitted to the input stages. If you have to provide pcb space for the capacitors, may as well put the diodes between those pads. Plus individual diodes can be installed by a pick+place machine, and screw terminal and screw mounted bridges have to be installed by human operators. $$
I have a Peavey PV-1.3k with a bridge, but the bridge has UL type fuses in it to limit current when the diode shorts. Plus, the entire transformer & rectifier are inside a separate grounded steel box, to keep the RF of rectifier shut off and 1300 v spikes on the AC input away from the input pcbs. Designed before MOS suppressors were cheap. Now all switcher supplies have a MOS suppressor right after the AC fuse.
Sizing? You assume class AB amp uses twice the electricity of the watt rating. Increase that current the inverse of the step down voltage ratio of the main transformer. That is first guess. Then in prototyping use current probe to determine how much surge current happens at turn on, and later at capacitor charging at low output watts. Ensure rectifier are big enough to handle those surge currents. Bigger amps tend to manage the turn on surge with a timer, relay and resistor between fuse & transformer. Else a NTC resistor like a CL-90 from GE.
I have a Peavey PV-1.3k with a bridge, but the bridge has UL type fuses in it to limit current when the diode shorts. Plus, the entire transformer & rectifier are inside a separate grounded steel box, to keep the RF of rectifier shut off and 1300 v spikes on the AC input away from the input pcbs. Designed before MOS suppressors were cheap. Now all switcher supplies have a MOS suppressor right after the AC fuse.
Sizing? You assume class AB amp uses twice the electricity of the watt rating. Increase that current the inverse of the step down voltage ratio of the main transformer. That is first guess. Then in prototyping use current probe to determine how much surge current happens at turn on, and later at capacitor charging at low output watts. Ensure rectifier are big enough to handle those surge currents. Bigger amps tend to manage the turn on surge with a timer, relay and resistor between fuse & transformer. Else a NTC resistor like a CL-90 from GE.
I think you might mean "MOV" (Metal Oxide Varistor) surge suppressors . They are discussed in this Wikipedia article.
Diodes monolithic bridges _ do you use them in power amps ?
Yes, I do. Generally I use the ones in the "GBPC" package and 1000V peak reverse voltage rating, 35A or 50A continuous forward current rating, whichever is in stock. (example) They bolt to the chassis so they're easy to cool, and they have enormous instantaneous peak current ratings so they don't melt when you use them without an inrush current limiter disc (a bad idea!). Please do include an I.C.L. disc like part number MS22-75004.
how do you size them ?
Largest available voltage and simultaneously, largest available forward current. While keeping the price below USD 7.00 or thereabouts.
The good news is that an optimized C-RC snubber eliminates all transformer secondary ringing, even when using these brutal and twitchy diode-packs that create big reverse recovery snapback pulses.
Yes, I do. Generally I use the ones in the "GBPC" package and 1000V peak reverse voltage rating, 35A or 50A continuous forward current rating, whichever is in stock. (example) They bolt to the chassis so they're easy to cool, and they have enormous instantaneous peak current ratings so they don't melt when you use them without an inrush current limiter disc (a bad idea!). Please do include an I.C.L. disc like part number MS22-75004.
how do you size them ?
Largest available voltage and simultaneously, largest available forward current. While keeping the price below USD 7.00 or thereabouts.
The good news is that an optimized C-RC snubber eliminates all transformer secondary ringing, even when using these brutal and twitchy diode-packs that create big reverse recovery snapback pulses.
Hi ! thank you very much for your kind and very helpful reply
My main concern was about the rating of the diodes used As i said i have seen in amps diodes good for low currents
I would not consider diodes lower than 10A Some 20A bridges are very cheap
What could be the cons of using oversized diodes between the mains transformer and the caps ?
i do not see any
i see the power supplies as something more electrical (high current high voltage) than electronic (low current low voltage)
i would love to use industrial parts for the amps power supplies
Hi ! thank you so much for your precious advice One big problem solved
Thanks ! i wonder what issues can cause these snapback pulses Will they happen only at the switch on ?
The “reverse recovery snap back pulses” happen every single time a diode turns OFF. That is, 120 times a second. They cause the transformer inductance to ring like a bell on each pulse - and are therefore far more audible than any hum caused by normal supply ripple. Sometimes the circuit just gets lucky and the loading conditions and supply filtering take care of it. Just as often it doesn’t.
Hi ! thank you very much for this very valuable advice If i understand right some kind of cap are always needed across any rectifying diode ?
this is a very important point
About the best value it has been suggested 0.01uF/1000V Why not more uF ? and why so high V ?
in the end there will the the secondary voltage across the diodes Not a very high voltage
This complicates a little the circuit I have seen already these bypass caps on some amps Very interesting indeed
However i am leaning to powerful integrated bridges They look very robust
Maybe their impedance could be lower than less powerful diodes ? i am obsessed by possible bottlenecks in the power lines coming from the transformer and going to the speaker outs
I read that these are normal weak points in commercial amps My main reference for how to do amps are the british amps
They take care of power lines and wirings (i like thick power wires)
this is a very important point
About the best value it has been suggested 0.01uF/1000V Why not more uF ? and why so high V ?
in the end there will the the secondary voltage across the diodes Not a very high voltage
This complicates a little the circuit I have seen already these bypass caps on some amps Very interesting indeed
However i am leaning to powerful integrated bridges They look very robust
Maybe their impedance could be lower than less powerful diodes ? i am obsessed by possible bottlenecks in the power lines coming from the transformer and going to the speaker outs
I read that these are normal weak points in commercial amps My main reference for how to do amps are the british amps
They take care of power lines and wirings (i like thick power wires)
Simple, no-math transformer snubber using Quasimodo test-jig
helpful analogy:
helpful analogy:
... oscillatory ringing of the L-C resonant circuit comprising the transformer secondary's leakage inductance, and rectifier + transformer capacitance. The resonant circuit is the "bell" and the diode reverse recovery is the "hammer" which strikes the bell & stimulates oscillatory ringing. To eliminate ringing ... add a C+RC snubber circuit across the transformer secondary, whose resistance has been tuned using the Quasimodo test jig and an oscilloscope.
Thank you sincerely again To be honest all this is way beyond my ability to understand
You mean that a snubber placed between the transformer outs and the bridge could be the better solution ?
if not very complex i thing i could build one
That would be interesting because there is space between the transformer and the bridge
Anyway i am shocked by knowing about this very critical issue
My power amp does not even have caps across the diodes !
i do not have access to a scope and i am thinking to buy one in the next future
Moreover i have read of issue with these caps related to not adequate rating
Strangely cheap ceramic caps are recommended against plastic ones
I wonder if all this power transformer ringing will show up at the amp output
You mean that a snubber placed between the transformer outs and the bridge could be the better solution ?
if not very complex i thing i could build one
That would be interesting because there is space between the transformer and the bridge
Anyway i am shocked by knowing about this very critical issue
My power amp does not even have caps across the diodes !
i do not have access to a scope and i am thinking to buy one in the next future
Moreover i have read of issue with these caps related to not adequate rating
Strangely cheap ceramic caps are recommended against plastic ones
I wonder if all this power transformer ringing will show up at the amp output
Last edited:
Snubbers in my equipment go across the rectifier diodes, not across the transfomer output. Transformer rectifier shutoff noise can often be heard as a buzz at twice the power line frequency. One expensive way of suppressing this is wrap the EI transformer in copper, then a grounded steel box around transformer and rectifiers.
Ceramic caps are inductive only like 2 wires of same length. Inductance of wound film caps negates the ability to absorb vertical voltage edges which fourier transform runs to dozens or hundreds of megahertz. Impedance of an inductor =sqrt((2*pi*f*L)^2+R^2) where L is inductance. The higher the frequency, the higher the impedance of inductance.
AC lines are subject to 1300 v spikes from shutoff of motors as in refrigerators and A/C units. Also lightning strikes far away on the AC line. This can be transmitted at lower value out the secondary of a transformer. Heard as a pop in the speaker. Before MOV became affordable about 1990, a secondary capacitor was the cheapest way to suppress this. Now MOV rated at 150% or 300% of nominal AC voltage can be purchased or salvaged and installed across the line after the fuse and before the transformer. If a transformer equipped amp has one of these, the diode snubber cap can be rated at only the voltage the MOV supresses the spikes to. When I installed MOV on my 1966 design amp, lightning strikes which usually caused a pop in the speaker became silent. A MOV is the blue thing near the inlet fuse in dead switcher supplies, which most modern dead appliances have.
Note other diode snubber designs have both a capacitor series a resistor. These are supposed to be more effective at reducing diode shut off noise. I don't worry about noise at 110 db below signal, but 72 db down pops and buzzes can be annoying.
Ceramic caps are inductive only like 2 wires of same length. Inductance of wound film caps negates the ability to absorb vertical voltage edges which fourier transform runs to dozens or hundreds of megahertz. Impedance of an inductor =sqrt((2*pi*f*L)^2+R^2) where L is inductance. The higher the frequency, the higher the impedance of inductance.
AC lines are subject to 1300 v spikes from shutoff of motors as in refrigerators and A/C units. Also lightning strikes far away on the AC line. This can be transmitted at lower value out the secondary of a transformer. Heard as a pop in the speaker. Before MOV became affordable about 1990, a secondary capacitor was the cheapest way to suppress this. Now MOV rated at 150% or 300% of nominal AC voltage can be purchased or salvaged and installed across the line after the fuse and before the transformer. If a transformer equipped amp has one of these, the diode snubber cap can be rated at only the voltage the MOV supresses the spikes to. When I installed MOV on my 1966 design amp, lightning strikes which usually caused a pop in the speaker became silent. A MOV is the blue thing near the inlet fuse in dead switcher supplies, which most modern dead appliances have.
Note other diode snubber designs have both a capacitor series a resistor. These are supposed to be more effective at reducing diode shut off noise. I don't worry about noise at 110 db below signal, but 72 db down pops and buzzes can be annoying.
Last edited:
Hi thanks for the kind and precious reply I think i have found a schematic Very complex
i have never seen these snubbers physically in a commercial power amp
I am confused now What a MOV has to do with diodes spikes noise ? it acts as a surge protector for the power supply
I guess a MOV can be added quite easily Not a problem Its the diodes switching noise that seems very nasty and critical to deal with
I will read everything better I am clearly missing something
The diodes near power inlet in my old televisions and 1966 Dynakit amp were 1000 v rated ceramic. This is because they were subject to power line spikes of 1300 v, or after transformer, 650 v spikes. A MOV reduces the spikes directly to only ~300 v so the ceramic snubber cap does not have to withstand half of 1300 v or whatever.
Many switcher supplies use 400 v electrolytic caps directly coupled to the power inlet, so they use both an MOV at the inlet after the fuse, and a parallel X or Y rated film capacitor. These X or Y capacitors are labeled for just above the inlet voltage (240 vac in Siemens AC areas) and the ability to withstand 1300 v spikes is understood from the rating agency tests.
The complex device shown in post 12 may include both a X or Y rated cap (left) and a RF absorber built into the IEC socket (the inductor and resistor next to the right). This RF absorber may be required to make the circuitry inside the appliance not emit RF howls to other equipment outside the case, as switcher supplies are prone to do. Corcom initiated these RF blocking IEC sockets in the US market.
Many switcher supplies use 400 v electrolytic caps directly coupled to the power inlet, so they use both an MOV at the inlet after the fuse, and a parallel X or Y rated film capacitor. These X or Y capacitors are labeled for just above the inlet voltage (240 vac in Siemens AC areas) and the ability to withstand 1300 v spikes is understood from the rating agency tests.
The complex device shown in post 12 may include both a X or Y rated cap (left) and a RF absorber built into the IEC socket (the inductor and resistor next to the right). This RF absorber may be required to make the circuitry inside the appliance not emit RF howls to other equipment outside the case, as switcher supplies are prone to do. Corcom initiated these RF blocking IEC sockets in the US market.
Last edited:
Quite often, the snubbers are omitted in solid state POWER amplifiers. Lightly loaded power supplies often do not have the diode/transformer ringing. At heavy load it very well might show up, but be 70 dB below the 100 dB music in the room, and I challenge anyone to hear it. If it happens at background music (or silence) levels it is painfully obvious. High bias AB amps, class A, and preamps tend to showcase it because the relative loading (compared to max ratings) in the transformer is high at idle. Tube amps are usually either class A or high bias AB - AND the transformers have high leakage inductance values (due to the large # of secondary turns) so it shows up in them all the time.
Hi and thanks again It seems that a MOV is mandatory I will check if there is one
I found a schematic
i see that is important that the fuse is on the L wire So i have to check also for phase ?
Maybe i could use one of these capacitors also in a linear power supply ? i will look for a schematic
thank you very much again I will try to get some schematics and also look to some service manuals of power amps
My original question was triggered bt seeing some quite big amps using tiny diodes for rectification This cannot be good
Expecially in high current power supply i like an industrial approach Big parts even if oversized
a 30A bridge cost almost nothing No reason to save money on it
Hi good morning ! thank you very much indeed Very very helpful I guess that i could check the bias and lower it a little
I do not like hot amps at idle at all I know that is possible to get nice sound even from low bias power amps Then why keep it high ?
now i have another reasone to keep it low Thank you very much again
I’ve got an experiment you can try. Fuse the neutral instead. Then plug it in, remove the fuse. Put one hand on your grounded work bench, and touch any part of “Protected Circuit” with the other. After that little experience, you’ll know exactly why you’re supposed to fuse the L wire.
It’s like when some dumb electrician puts the light switch on the white wire and you go to change out the light for a ceiling fan. You might never come in contact with live electricity, but when the black wire hits the box, you end up on the floor with the ladder and a ceiling fan on top of you.
O.P. lives in Italy the flag says. Italy is in the Siemens electric area, and has no neutral. Both sides of the 220-240 VAC are hot.
MOV are not mandatory but do reduce pops from lightning and A/C shutoff in transformer driven amps even if the transformer is not ringing from diode shutoff. Disc caps used in after transformer snubber are subject to 1300 v*(transformer stepdown ratio) spikes unless a MOV or X or Y cap is across the inputs to the transfomer. With MOV or X or Y cap protection, the secondary snubber capacitor can be ~400 v rated or less.
How do I know spikes are 1300 v? I was responsible for test equipment in a refrigerator factory. One test involved interrupting power to a running refrigerator, and observing that when the compressor stalled, the overload protection device cut off the AC to the compressor. Every 15 seconds, a new refrigerator produced a 1300 v spike on the contactor used to interrupt the AC power. I could see it on an oscilloscope.
MOV are not mandatory but do reduce pops from lightning and A/C shutoff in transformer driven amps even if the transformer is not ringing from diode shutoff. Disc caps used in after transformer snubber are subject to 1300 v*(transformer stepdown ratio) spikes unless a MOV or X or Y cap is across the inputs to the transfomer. With MOV or X or Y cap protection, the secondary snubber capacitor can be ~400 v rated or less.
How do I know spikes are 1300 v? I was responsible for test equipment in a refrigerator factory. One test involved interrupting power to a running refrigerator, and observing that when the compressor stalled, the overload protection device cut off the AC to the compressor. Every 15 seconds, a new refrigerator produced a 1300 v spike on the contactor used to interrupt the AC power. I could see it on an oscilloscope.
Hi already done with a power strip with a single pole switch The swith clearly was on the neutral line I touch the live wire
the safety breaker saved my life since then only dual pole switches
The problem is that schuko plug is simmetric It can be plugged in two ways I have to mark the live end on the power strip and on the amp
it would be nice to have a little led that lights on when the Live of the wall socket is connected to the Neutral of the amp I guess it could be done
i see Better use always a safety breaker and test it periodically I should keep always a tester at hand doing things Or use rubber gloves
In the end i will only replace diodes with bigger ones Some stock diodes looks flimsy
as an example these can be found on a mighty SAE 2200 power amp I do not understand why they saved on this important parts
Mains transformer and ps caps are impressive
i love big diodes at least 10A parts Better if integrated
The current at idle will be always low No waste of energy My system should be sustainable
Thanks a lot again for the precious advices
i have to check but i think that is L (230VAC)-G-N the phase finder lights on only on one pin and nothing on the others
i was using a power strip with a single pole breaker I plugged a plug with open contacts and touched the live wire the safety breaker kicked on
Very very interesting You have a really great competence and expertise as i said the industrial applications are a very great training also for power amps design and construction where voltages and currents can be quite high
I should not play with power i know Now all my power strips have dual pole breakers that cut L and N wires
I will just replace the diodes with bridges And some bad looking caps