SS rectifier snubber networks

[Esquirefreak]

Been reading about the use of snubbers and caps to dampen noisy rectifiers.

I have just the slightest 100hz buzz in my recent 6G15 clone and thought that I'd try to get it quiet.

Some people use a cap across each diode or a single cap in parallell with the HT AC leads. Then there's the version with series R+C, in parallell with the HT AC.

I have a couple of questions:

× Is one better or worse? If so - why?

× When putting either C or R+C in parallell with the HT AC, is there any trouble when the PT has a CT?

/Max

[shooter]

I've used them in HiFi builds, never in a guitar amp
they are typically used to suppress transient spikes that diodes seem to throw into circuits
hiss is typically a resistor noise, or possibly a sensitive signal getting to close to a B+ line and coupling the 120hz into the signal path

[kagliostro]

Ciao Max

Simply put an UF5408 diode between the bridge and the first filter capacitor


Franco

[Esquirefreak]

Yeah, I get that it is more commonly used in the hifi world.

Just a regular UF5408? With the anode/cathode facing the same direction as the rectifier? How does that work, I mean in a little more technical terms? :)

/Max

[pdf64]

Ciao Max

Simply put an UF5408 diode between the bridge and the first filter capacitor


Franco

Have you got any further info / a rationale, for this?
It’s just that I don’t see how it can do any snubbing 

[Esquirefreak]

That's kinda what I was thinking. I too wonder why/how it would work.

[dwinstonwood]

I read somewhere that it's diode switching noise that it addresses. Using ultra-fast switching diodes like the UF4007 or UF5408 instead of 1N4007's is supposed to suppress switching noise. If I've ever heard "switching noise" I'm not aware of it, and don't know what it sounds like.

[shooter]

^^^
that's my understanding also, used for transients aka switching noise
IF you hear them, they're sorta like the dust "pops" you hear in a record, or the "pop" you hear switching channels.
tubes mostly don't care at audio frequency, since most transients are too fast for tubes to respond, in a SS amp, they get noticed, can cause chip fails by exceeding max volts on inputs, PS regulators

[pdf64]

I think the noise may be due to interaction between the diode's switching characteristics and the transformer winding inductance and parasitic capacitence.
When switching, the diode conducts reverse current momentarily, then switches that off suddenly. That excites a resonance in the PT circuit.
Tim Robbins explains stuff really well, see p15 onwards of https://dalmura.com.au/static/Power%20supply%20issues%20for%20tube%20amps.pdf

[Esquirefreak]

Last reply is my understanding as well. Capacitive coupling between the PT and recto. So as long as the caps in parallell with the diodes have a higher capacitance then the diodes themselfs, the "switch off" smooths out.

I've ordered UF4007 and UF5408, I also have some KBPC3510FP (chassis mount) laying around. I thought that I'd build a couple of rectifiers and experiment a little. It's cheap enough for the sake of learning.

Interesting nonetheless :)

Repeating my second question:
Does implementing a R+C in parallell with the secondary HT present any trouble when there's a CT involved? Drawing borrowed.

[pdf64]

So rather than a full wave bridge, you’re referring to a 2 phase / push pull rectifier, with a winding CT connected to circuit common?
If so, I think the snubbing should be from each leg to circuit common.

[Esquirefreak]

RC between both AC HT leads and center tap?

(HT -> SNUBBER -> CT <- SNUBBER <- HT)

Or in parallell with each diode?

[shooter]

maybe some quality time with an oscope will help, don't believe there's been a hiss issue I can recall that lead back to rectifier diodes, now noisy resistors, signal coupling, you can find those all day every day in past threads

[kagliostro]

I'm sure the one fast diode trick is from Mr. Merlin (Blencowe) but I learned it long time ago and I don't remember if he shared the trick on his web pages (they vary on the time) or in one of his brooks (or both)


Franco

[Esquirefreak]

While it is buzz rather than hiss, it may indeed be related to something else, like resistors. It could also be due to a less than perfect grounding path. I will re-do my grounding (since I did a bunch of homework on that part on another build recently...) before any further testing. I can't hear any significant 50hz hum at all, though it may be drowned by 100hz noise.

If that won't do it and I know my grounding scheme is alright I'll probably experiment with the rectifier. Easy enough to find out if one/two snubbers or a UF5408 between rectifier and reservoir cap dampens the buzz or not. And if it does, we'll know a cheap fix.

/Max

[shooter]

side-track ahead.
HiFi info injected.


these are shots from the amp, using the PS I posed earlier.
If there were switching problems, they'll typically show up, indicated in red
you can see both in's n outs are clean to 48W
I followed the TI schematic, sans the protection stuff


go hunt down your bug, play guitar til the tubes glow

[kagliostro]

I did a search

The single fast wsitching diode trick was in the book Designing Power Supplies for Valve Amplifiers, Rectification chapter

Fig. 2.14 b

"A single fast-switching" diode can be used to obtain a short reverse-recovery time, which reduces the amplitude of ringing"

I think that the price of a diode worth the effort to give a try

Franco

[pdf64]

Thanks for the reference!
I had a read through it and yes, shock horror, with a FWB rectifier Merlin's right, that would work!
It didn't help that my grasp of FWB rectification had become a bit fuzzy.
However I'm not sure whether it would also work the same on a push pull / 2 phase rectifier; as I see it, in that case, the reverse current creates a momentory short across the full HT winding, which a further diode after the rectifier won't eliminate 

Whatever, as a 6G15 only uses a half wave rectifier, a single UF diode will suffice anyway 
But then the buzz would have been 50, rather than 100Hz?

https://el34world.com/charts/Schematics/files/Fender/Fender_reverb_6g15.pdf

[Esquirefreak]

I'm actually using a full wave rectifier. Since my PT was 125 - 0 - 125, I had to do this in order to get the desired B+ voltage. Also using a hum block looper/elevated signal ground circuit.

So building a couple of rectifiers or mounting a bridge recto for experiment is just a couple of minutes work.

/Max