PM645 VXi with Fast and soft recovery rectifier diodes

[wkloppen]

Hello Gents,

I replaced the rectifier diodes with these ones. Don't know whether I could hear the difference. I doubt it to be honest but while reaplacing all the caps I thought I'd give it a go.
ON SEMICONDUCTOR RHRP860 Fast / Ultrafast Diode, Soft, 600 V, 8 A, Single, 2.1 V, 35 ns, 100 A

Opinions and experiences apprecciated...

 

[solderdude]

Fast recovery diodes are intended for SMPS. The only advantage they have in a normal power supply is that the DC voltage rail is approx. 0.3V higher when using these diodes. The diodes will also heat up less during higher current draw (at the same current) because the dissipated power is lower because of the lower voltage drop.
The dV/dt of the mains frequency is very very low. Normal rectifier diodes are already so much faster than that so the 'high speed' of the Schottky diodes is not used.

Because the diodes are 'high speed' (for SMPS essential) most 'modders' that use them hear the 'greater speed' in the sound. Just like they hear more 'brightness' with silver wiring.

After the rectifier all there is is DC with a ripple. That ripple will be the same with both types of diodes and the capacitors will provide the same peak currents to the amp if needed.
Better (often higher capacitance) reservoir caps that are usually replaced along with the 'diode modifications' are more likely to make a difference that could be potentially audible than the gain of 0.3V DC voltage.

 

[wkloppen]

Fast recovery diodes are intended for SMPS. The only advantage they have in a normal power supply is that the DC voltage rail is approx. 0.3V higher when using these diodes. The diodes will also heat up less during higher current draw (at the same current) because the dissipated power is lower because of the lower voltage drop.
The dV/dt of the mains frequency is very very low. Normal rectifier diodes are already so much faster than that so the 'high speed' of the Schottky diodes is not used.

Because the diodes are 'high speed' (for SMPS essential) most 'modders' that use them hear the 'greater speed' in the sound. Just like they hear more 'brightness' with silver wiring.

After the rectifier all there is is DC with a ripple. That ripple will be the same with both types of diodes and the capacitors will provide the same peak currents to the amp if needed.
Better (often higher capacitance) reservoir caps that are usually replaced along with the 'diode modifications' are more likely to make a difference that could be potentially audible than the gain of 0.3V DC voltage.

Now thats an answer with solid info!! I indeed read about the low frequency (ie: 50/60hz) is a walk in the park fro every diode. They are in now so it cant hurt I guess.
The small capacitors around the diodes are 0.01 uF. To what value could / should I raise them?

reg

willem

 

[wkloppen]

Now thats an answer with solid info!! I indeed read about the low frequency (ie: 50/60hz) is a walk in the park fro every diode. They are in now so it cant hurt I guess.
The small capacitors around the diodes are 0.01 uF. To what value could / should I raise them?

reg

willem

Or do you mean the big 10.000 uF / 63 volts caps?..I've undestood you cant rais them by too much as the surge current when firing up the amp might take too long too fill them which could be lead to damage to eith the transformer or other things....

 

[solderdude]

The snubber capacitors (10nF) are supposed to suppress the switching glitches of the diode when it switches off. These 'slow' diodes even more completely undoing any 'fast recovery' aspects the diode has.

You should leave them in there but it would be better to put small resistors (say 10 ohm) in series with those capacitors.

Reservoir caps (smoothing caps) are indeed the big ones right after the rectifier.
Increasing the values also increases the inrush current a bit but the transformer is the actual current limiter.
A bigger capacitance, on start up, will make the inrush current a tiny bit higher and last a bit longer. The rectifier diodes experience a bigger momentary (peak) current. When diodes are used that barely make it under normal circumstances these might give up the ghost a bit sooner.

Increasing the capacitance won't increase average drawn power during usage but can provide a bit higher momentary output power (when the amp circuit is not the limitation) when playing music, which then on average will slightly increase power consumption of course.

Upping the capacitance 10 fold (if possible at all regarding sizes) isn't the wisest thing to do. A 50% increase, in general, will not cause problems but also won't increase momentary power much.

Merely putting in new caps (when replacing really really old caps that haven't seen much action for years) or ones that leaked or bulged can have an audible impact but that is because those caps already failed and the circuit was repaired rather than 'improved'.

 

[AnalogSteph]

The snubber capacitors (typically anywhere from 10 nF up to 100 nF) serve two functions:
a) bringing down secondary transformer L-C resonance
b) keeping the rectifier diodes from modulating AM radio signals making it into the unit via mains, which when reradiated generates hum in nearby receivers - obviously quite important back in the day

Part b) is why finding a good power supply for a portable LW / MW / SW receiver could be a bit tricky, as a lot of general purpose linear supplies didn't bother with these and would require modification.

 

[wkloppen]

The snubber capacitors (typically anywhere from 10 nF up to 100 nF) serve two functions:
a) bringing down secondary transformer L-C resonance
b) keeping the rectifier diodes from modulating AM radio signals making it into the unit via mains, which when reradiated generates hum in nearby receivers - obviously quite important back in the day

Part b) is why finding a good power supply for a portable LW / MW / SW receiver could be a bit tricky, as a lot of general purpose linear supplies didn't bother with these and would require modification.

thx ...well i dont have a receiver as Im only using web radio's. I thought the soft recovery diodes also reduced the ringing or is it both the small caps ND the soft recovery diodes adding to that?

reg

willem

 

[sq225917]

The caps on their own won't do that, to snub ringing in the transformer you need a C-CR filter between diode and traffo, and this needs to be tuned to the circuit.

In the case of the phonostage I build a 10n- 10n- 82r filter totally kills any ringing in about 5 cycles, vs about 50 without. I doubt it's audible, but it's best practice.

See the quasimodo thread on diyaudio for details.

 

[AnalogSteph]

I mean, by bringing down the LC resonance you have already reduced common-mode coupling to the mains via parasitic transformer capacitance. A portable AM radio (ideally a full-range LW/MW/SW set) makes a pretty good diagnostic tool, you can literally pick up the buzzing generated by reversal glitches nearby. Fast recovery diodes should reduce these glitches a fair bit.

 

[sq225917]

A cap 9n it's own doesn't do it anywhere near as effectively as a C-RC filter.

 

[wkloppen]

The caps on their own won't do that, to snub ringing in the transformer you need a C-CR filter between diode and traffo, and this needs to be tuned to the circuit.

In the case of the phonostage I build a 10n- 10n- 82r filter totally kills any ringing in about 5 cycles, vs about 50 without. I doubt it's audible, but it's best practice.

See the quasimodo thread on diyaudio for details.

A cap 9n it's own doesn't do it anywhere near as effectively as a C-RC filter.

This is the transformer with rectifier scheme and the small capacitors.

Secondary AC is 74 volt (2 x37)
DC unloaded is +48, -48 volt

I don't see the resistors vor de current limit to the 0,01 uF (10nF) capacitors. Is this more or less trial and error or could also start using 82R for instance?

 

[restorer-john]

A portable AM radio (ideally a full-range LW/MW/SW set) makes a pretty good diagnostic tool

They sure do. I have my Sangean ATS-803A permanently on my bench (150kHz-29999kHz continuous on AM).

One of these:

 

[solderdude]

I thought you meant the diodes were in parallel to the diodes (which one often sees)

I wrote something about snubbering

The capacitors form a resonance circuit to lower the glitches from the diode switching on and off and making it back onto the mains via the transformer.

Basically C35 and the transformer form an LC circuit with small switches that go on and off near the maximum voltage swing of the input.
and just very slightly make small wiggles as the load changes because of it. It is gradual for these diodes. It is very slow motion in the eyes of the diode as it were.
This has worked before so I would just leave it as it is unless somewhere in the house you could notice the amp was on or off by other gear being affected.

When the smoothing caps needed replacement just replace them. It really does not matter which diodes are used. The only thing that changes is the unregulated DC voltage being slightly higher.

My diagnostic tool... you can even see where the soldering iron landed on by accident.

 

[wkloppen]

Gents,...thx for the insights...much obliged. @solderdue...read your snubber article. Learnt a few things.....The amp sounds really good so probsbly no need to add the resistor in series with secondary mains capacitor (c35)

 

[sq225917]

Well chosen c1 and R1 will get you most of the way there, an additional C parallel with them is the icing on the cake.

I have a small square wave generator that you plug in where c1r1 are in the diagram, attach a scope and adjust the component values until you get the lowest ringing, with least overshoot and fastest settling time.

https://flic.kr/p/PePkAc
25va 24v traffo with common mode chokes, showing snubbed on underside.

https://flic.kr/p/QS8HgN
Snubbed and unsnubbed. This really won't be audible, it's just quest for perfection stuff.

 

[wkloppen]

Thx gents for the info.....one small question regarding noise. When I don't play anything, no music, no nothing and with volume to minimum...I hear a a very soft "sssssssssss" through the speaker only audible with me ear very close to the speaker.
When I turn up the volume to maximum,..again with no music playingand with my ear somewhat IN the speaker, I also hear a very soft and low "brom" with I guess the power supply frequency.
Is that normal and more or less typical with a solid state amp?

regards

willem

 

[AnalogSteph]

This does sound fairly typical. Amps with the tone controls in the feedback loop tend to not be the very lowest-noise ever (as relatively high impedance meets the input bias currents of a reasonably hot input stage), this one also seems to have a total gain of 45 dB which is on the more generous side for a 70 wpc amplifier and a 100k volume pot.

And whatever you do, better leave the "phase-correct loudness" off altogether. This circuit seems to have been plagued by insufficient hum rejection right from the factory.