Anyone try fast recovery diodes in the power supply of their 'wreck clone? Using FREDs was promoted by Victoria Amps and Tonequest for a while as a tone enhancer but seems to be considered as bogus by a lot of folks. The Trainwreck Pages mentions that you should not use them on the output sockets to suppress spikes but I can't find any mention of them here otherwise. I think the fast recovery is part IN4009. I bought some FREDs from Digikey a few years ago and used them to rebuild a Traynor Bassmate as a kind of half-assed copy of a Dr. Z with an EF86 preamp; the amp sounds real good although without the beef of the Z(I kept the original trannies), but I never compared it without the FREDs so who knows if they were any improvement. I am thinking about using the FREDs or the IN4009 in the power supply when I finally start the Express project, but maybe someone here has some input why I wouldn't.
Any thoughts?
fast recovery
Moderators: pompeiisneaks, Colossal
Re: fast recovery
I played with FREDs a few yrs ago and here's my conclusion, YMMV:
FREDs, snubber caps, etc can be effective at quieting hashing noise in a SS rectifier- IF you have a noise problem with the rect to begin with. The 120hz rect buzz can manifest itself as a harshness in the high freq response of the amp. Therefore, if you lower the noise it can, perceivably, "soften" the high freq response of the amp.
Most rect noise problems can be solved by proper layout, grounding, etc, but in some cases it may be more convenient to use a snubber or FRED. In my case, it was a very small chassis that I was trying to shove everything including the kitchen sink into. The rect spikes were radiated like RF and picked up by a nearby grid circuit, then modulated by the guitar audio signal. You could hear the buzz riding on top of the guitar sound which added a slight harshness to it. The FRED was effective at removing that.
No mojo, no voodoo... Just a specific solution to a specific problem that happened to work. If you don't have a problem with rectifier noise (and you shouldn't with proper layout & grounding in a new amp) I wouldn't bother with it.
IMO...
Doug
FREDs, snubber caps, etc can be effective at quieting hashing noise in a SS rectifier- IF you have a noise problem with the rect to begin with. The 120hz rect buzz can manifest itself as a harshness in the high freq response of the amp. Therefore, if you lower the noise it can, perceivably, "soften" the high freq response of the amp.
Most rect noise problems can be solved by proper layout, grounding, etc, but in some cases it may be more convenient to use a snubber or FRED. In my case, it was a very small chassis that I was trying to shove everything including the kitchen sink into. The rect spikes were radiated like RF and picked up by a nearby grid circuit, then modulated by the guitar audio signal. You could hear the buzz riding on top of the guitar sound which added a slight harshness to it. The FRED was effective at removing that.
No mojo, no voodoo... Just a specific solution to a specific problem that happened to work. If you don't have a problem with rectifier noise (and you shouldn't with proper layout & grounding in a new amp) I wouldn't bother with it.
IMO...
Doug
Re: fast recovery
Is that here somewhere? Some time ago, Rick Erickson drove a thread on ampage to the conclusion that 1N4007's are probably not fast enough in that position and that UF4007s would be a good replacement. They're not the FREDs people usually mean. They're just faster than the 1N400x series. I would guess that people might think they could use 1 FRED instead of 3 1N4007's, which wouldn't get the voltage threshold high enough. I don't know why there would be any other issue.arfy wrote:The Trainwreck Pages mentions that you should not use them on the output sockets to suppress spikes but I can't find any mention of them here otherwise.
Re: fast recovery
Fooled around with FREDs for rectification in my Express clone yesterday. This is what I found:
With FRED diodes (BY500/800) the normaly bold low end mushed out, which surprised me. FREDs are generally said to bolster lo end. Anyway, I changed B+1 caps from 47µ and 22µ to 2x 22µ and the bottom was tighter again contradictionary to common knowledge (more cap = tighter lo), yet somehow muddier sounding than with 1N4007s so they went right back in along with that 47µ cap.
Timo
With FRED diodes (BY500/800) the normaly bold low end mushed out, which surprised me. FREDs are generally said to bolster lo end. Anyway, I changed B+1 caps from 47µ and 22µ to 2x 22µ and the bottom was tighter again contradictionary to common knowledge (more cap = tighter lo), yet somehow muddier sounding than with 1N4007s so they went right back in along with that 47µ cap.
Timo
Re: fast recovery
might be mis-remembering, but I thought the reason KF didn't think they were good for spike suppression was that they had too much capacitance and bled off too much highs ( from the plate AC signal). The only time I've noticed a changed w/those suppress.diodes (using 1N4007s) is on a boogie .22+--it noticably bled off highs (due to the greater gain of the EL84s? = more negative fdbk.?).jaysg wrote:Is that here somewhere? Some time ago, Rick Erickson drove a thread on ampage to the conclusion that 1N4007's are probably not fast enough in that position and that UF4007s would be a good replacement. They're not the FREDs people usually mean. They're just faster than the 1N400x series. I would guess that people might think they could use 1 FRED instead of 3 1N4007's, which wouldn't get the voltage threshold high enough. I don't know why there would be any other issue.arfy wrote:The Trainwreck Pages mentions that you should not use them on the output sockets to suppress spikes but I can't find any mention of them here otherwise.
Re: fast recovery
Part of my confusion
is that I can't find any info about The Trainwreck Pages after the late 80's. I don't recall FREDs being around back then.