Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: rdrgtr on October 12, 2010, 08:56:39 pm
-
I guess I read to much. One guy claims that all 1N4007 diodes need to be replaced with UF4007 fast diodes. He says you will hear noticable improvment (smoother sound). Has anyone out there made the change? In theory what should the results be? I know it is a simple mod, I am just not sold that the change will be worth taking my amp apart.
Thanks.
-
All I use is UF4007. Do I hear a difference - naw, but what the heck. I spose if I was a gunnin metal player I might.
-
Total nonsense. The slowest diode on the planet switches 100,000 times faster than the 60 Hz of a transformer secondary. Whatever difference this could possibly make would result in a miniscule fraction of perhaps one volt difference in the output B+ of the supply. If you think you can hear that, you're a better man than I.
On the other hand, it's absolutely critical to change any blue or green pilot-light jewels you may be using to red ones.
-
I have installed lots of both types (I buy 100 at a time) but never heard any difference.
-
In theory what should the results be?
In theory, the faster diode, once forward biased, should transit from non-conductive to conductive faster. It should also de-conduct, eg, shut off faster. What means "faster"? All the diodes "know" to conduct when their anodes climb about .65 volts above their cathodes (bars)
So, on a rising half-cycle of AC, the faster diode should conduct infinitesimally sooner. Aha, you say, that means more electrons = higher current, and higher volts get through to the filter.
But by the same token, on the falling half cycle of the same waveform, the fast diode shuts off faster. So, less current, less volts get through to the filter.
Once the filter caps are charged up in perhaps a dozen cycles of input AC or less, I myself can't see this making any kind of difference...by that I mean mere fractions of a volt in something being charged up to 350-450 volts. If there is some difference that's worthwhile paying attention to in this picture, then one should be likewise paying attention to line voltage variations, temperature coefficients in coupling caps and lots of other minutiae.
But I'd be happy to be proven wrong.
-
But I'd be happy to be proven wrong.
thats the mistake right there. You scientific types, its always "if you cant prove it, it doesn't exist". Pish. Come over to the dark side, use your ears, feel it...
:grin:
-
Either the AX84 or 18-Watt folks convinced me to use UF5408 (http://www.mouser.com/catalog/specsheets/uf5400.pdf) "Soft Recovery Ultrafast Plastic Rectifiers". Same guys who insist that you need a diode/resistor network any time you put rectifier diodes in series to eliminate "switching noise". The 3 amp current rating probably does make sense in something like a Super Reverb.
FWIW here's the data sheet on the UF4007 (http://www.fairchildsemi.com/ds/UF/UF4007.pdf)
Can I hear a difference compared to 1N4007 rectifiers? Can't say without a proper double blind test, but I really doubt it.
OTOH there are only 2 or 3 rectifier diodes in an amp, they all cost less than 4 bits, and I figure it doesn't hurt to over-build in this area.
YMMV
Chip
-
Actually, I prefer chicken entrails. :huh:
Dave
-
FWIW I have found a difference in a 6G15 type unit that had FW SS rectification (not bridge) with two 1N4007 diodes in series on each side of the rectifier, when I put a .01uF cap in parallel with each diode (that's 4 diodes and 4 caps in total). I undid and redid this just to be sure, and the diodes definitely did snub extraneous noise. It may well have been because it was a noise-sensitive situation, but there was definitely an improvement/noise reduction with the snubber caps.
I have had a few skeptics poo poo me on this and try and tell me I don't know what I'm on about etc etc, but I'm just relaying my experience. Also I haven't tried fast switching diodes, but a few of the guys at ampage (who are much more savy than me) have said that fast switching diodes can also make a difference in this regard. It boils down to, you won't truly know unless you try it for yourself. No amount of listening to other opinions is any real substitute for direct experience.
-
One guy claims that all 1N4007 diodes need to be replaced with UF4007 fast diodes. He says you will hear noticable improvment (smoother sound).
Fast diodes are less noisy, this can be measured easily. Is there a subjective difference in proper DBTs? AFAICT, none whatsoever for properly designed and built toob geetar amps.
As fast diodes cost only marginally more than std versions you may use them for new builds - I do - but retrofitting them to perfectly working amps isn't worth the bother.