• Check out the Annual AK Secret Santa exchange Program, signup before November 23rd in the AK Charities & Christmas forum!

Tube vs. SS rectification: Any significant sonic differences?

Nirky

Active Member
I've been getting differing opinions about tube vs. solid state rectification.

Boiling it down, it appears there are two camps. One says tube rectification is necessary, because if you go with SS, you lose or significantly alter the tube sound.

The other says that the sonics remain, and if anything sonically changes, you actually gain in faster bass response.

All of my valve amps have tube rectifiers, so I can't do any comparisons. I'm curious to hear what others have noticed, if anything.
 
Register to hide this ad
I suppose you would have to have the exact same amplifier, one tube rectified and one SS to do a fair comparison. That being said, I prefer my SS rectified Dynaco 70 easily over any I've heard with tube rectifiers. That's still not a fair comparison though, because with the SS rect. came other mods too. I guess if we were voting, you could put me in the SS camp just because it's so much easier to regulate and such. I don't know though, I like a lot of stuff that's tube rectified as well. :dunno:


Some guys even "roll" rectifier tubes of the same number and claim sonic differences.
 
The thing I don't like about solid state rectification is you've got HV instantly, before everything has a chance to settle down. If you go with silicon rectifiers, add a "standby" switch to keep the HV off the tubes 'til they're warmed up and everything stabilizes.
 
I know of one very good headphone amp that uses SS rect. if I am not mistaken and it is my Yamamoto HA-02. It uses a single WE408a pentode per channelas a driver tube I think. I am not a tube builder so I really don't know much about these things, but I have been reading. I know this amp is faster than my AT-HA5000 mosfet SS amp was when I had it.
 
hey-Hey!!!,
It is difficult to get a SS diode to switch OFF as quickly and thoroughly as a vacuum diode. SS is lower impedance, and immediate. The comparison gets even more nutsy when stuff like Mercury vapour is introduced. There is a lot of value to slow warm-up afforded by diodes like the 6AU4 TV damper.

As far as SS goes, the SiC Schottky variants from Cree are *VERY* nice.
cheers,
Douglas
 
60 second B+ timer.

The thing I don't like about solid state rectification is you've got HV instantly, before everything has a chance to settle down. If you go with silicon rectifiers, add a "standby" switch to keep the HV off the tubes 'til they're warmed up and everything stabilizes.


I installed a home brew 60 second B+ timer for my Sansui 1000A that has solid state rectifiers, the tubes have 60 seconds to heat, then the high voltage is applied.

Mark T.:music:
 
Both

I have both and I think I prefer the diodes although if you ask me tomorrow I might tell you that I like the tube. A bit like ice cream if you know what I mean. :smoke:
 
My complaint with tube rectifiers is the inconsistency! I was always bugged by the variance between 5U4's in my Craftsmen amps. Even when the voltage drop measured the same, I'd hear huge changes in sound (presumably due to dynamic impedance variations and switching spectra) between, say, a Tung Sol 5AS4A (overblown bass, crunchy top), a Tung Sol 5U4GB (hard bass, strange coloration, lack of detail), an RCA 5U4GB (gorgeous lower-midrange, glassy top-end) or a Sylvania 5U4GB (fast, tight, clean, dry). This made me feel like a total audiophile nut. In the end, the best by far were the RCA 5U4G and Sylvania 5U4GB, probably due to better damping with the choke input filter. I swear some even lead to audible filter ringing, but I was too lazy to try and measure this. :sigh:

So, in the end, I'd prefer Solid State for a normal every-day tube amp.
 
Hm, and here I never figured it would make any audible difference. I was figuring that since it was simply a power supply part, all it needed to do was supply clean DC at the proper voltage for the B+. One would think the filter caps and chokes would negate any differences between tubes since all they're supposed to do is supply DC power. Well, like everything with audio, not all differences make sense.
 
You can use fast diodes full wave, and still put a high-current damper diode tube (like 6CJ3) in series for slow B+ turn on. I've never seen it proven with a 'scope, but I think the damper also filters a lot of HF hash/noise caused by some diodes.

Another trick the kids are doing these days is to use a low-inductance low-dcr choke between diodes and cap input, like a choke input but way below critical inductance at 120 hz. Henry Pasternak did a nice analysis of what happens at the ripple wave level when you do this -- and it's not trivial and it's not obvious. You get a flywheel effect that "overshoots" the resulting B+ with each cap charge-up cycle, which smoothes the charge-up and improves regulation, passively.

A little choke will also very effectively filter HF noise both ways... It can be something like 300-500mh (half a henry), with DCR in the 10 ohm neighborhood. Triad and Hammond make them, or I've got a lot of NOS ones that are kinda right, and will give hobbyists a couple if they pay a visit.
 
hey-Hey!!!,

As far as SS goes, the SiC Schottky variants from Cree are *VERY* nice.
cheers,
Douglas
Where do you get those and how much do they cost? I looked them up and they not only come in higher voltage ratings than I've seen before, but insane currents... plus they're Shottky's. I just can't figure out how much they cost.
 
Trying to compare the sound difference from SS diodes to tube rectifiers in identical amplifiers by just swaping out the rectifiers is going to net you a very skewed comparison (basically worthless). Both setups can be designed and implemented to perform very close to one another from a Sonic stand point. To get the best out of either type the power supply as a whole would be very different. If you slap SS rectifiers in a amplifier designed for a tube rectifier the overall performance is going to suffer you may see improvements in one area and degragation in another. Place a tube rectifier in an amp design for SS (usually impossible without major changes) and the same will be true.

No real absolute answer exists. I personally prefer tube rectification for many reasons most of which have nothing to do with Sonics.

Craig
 
Where do you get those and how much do they cost? I looked them up and they not only come in higher voltage ratings than I've seen before, but insane currents... plus they're Shottky's. I just can't figure out how much they cost.

IIRC, Partsconnexion has them. They do not behave like P-N junction in a critical fashion. They *CAN* be paralleled effectively, but serious connection does not increase PIV rating.
cheers,
Douglas
 
They *CAN* be paralleled effectively, but serious connection does not increase PIV rating.

No serious connection, eh? Maybe something comic, or just sincere?

But seriously (heh), disregarding the typo you can't put them in series with balancing resistors to increase PIV rating? That IS a whole different way of diode-ing.

Is there a simple explanation of how those things work?
 
AC heating tube, DC heating solid state

Hi,
I think if there is a solid state DC heating you can use solid state components in the HT supply, too.
If there is AC heating directly from the mains transformer it makes definately sense to me using a tubes in the HT supply. This makes the set free from solid state parts.:banana:

Kind regards,
Darius
 
No serious connection, eh? Maybe something comic, or just sincere?

But seriously (heh), disregarding the typo you can't put them in series with balancing resistors to increase PIV rating? That IS a whole different way of diode-ing.

Is there a simple explanation of how those things work?

I wouldn't think series-ing them would increase the voltage rating. They are dropping maybe 2 volts each. 1000 volts goes into 1 of them, 998 comes out. The next diode is STILL having almost 1000 volts applied to it. I thought series-ing components for a higher voltage handling capacity only worked when you ended up with those components with HV on one end and GROUND on the other, like two series filter caps?
Wait a second... is 1000 volts the max REVERSE voltage they can handle, and they can handle infinite (within reason, maybe not a few million) FORWARD volts? In that case, having them in series would help because they are closed when reverse voltage is applied and not acting like a 2 volt drop but a high voltage (infinite) drop.

Hmm... those shottky's cost quite a bit more than some modern production rectifier tubes. Each.
 
I wouldn't think series-ing them would increase the voltage rating.

Think voltage divider if you use resistors around them. That's how the ham radio guys would make full wave bridges with 1000 PIV diodes for kilovolt supplies.
 
What value resistors would I use?
I was actually considering building a large voltage doubler/tripler/quadrupler, etc, just because it would be an interesting project, and high voltage has all sorts of fun applications (coil guns, big sparks, etc) if you're careful with it. I didn't realize that 1000 volt diodes could actually be used about 1000 volts at all.
 
I have seen resistors (100K-200K) used, but I guess that's for safety, in case of diode failure, so it doesn't cascade through the rest...

Check http://w5jgv.com/hv-ps1/index.htm to see how a builder progressed in use of cheap 4007 diodes for a HV supply. Scroll down to HIGH VOLTAGE RECTIFIER: but be sure you get to the photos of the second and THIRD versions!

Here's how I thought I'd want it to look if I build an HV supply:
http://www.caarc.ca/e107_plugins/gallery2/gallery2.php?g2_itemId=2374

That's how I've seen it illustrated, with a snubbing cap and a resistor around each diode.
 
Back
Top Bottom