SS diodes in front of rectifier tubes: fad or good idea?

[Rhodesplyr]

I've read about several people on DIY sites placing diodes in series with tube rectifiers such that the tube only receives the positive voltage pulse, the theory being that not seeing the negative supply swing is somehow easier on the tube rectifier.

Does anyone think that there's any merit to this idea, or is it just a pointless exercise that sounds like a good idea?

.

[Enzo]

That is not why it is done. Reverse voltage is not stressful on a rectifier tube - they are made for it.

The diodes protect everything else from the tube. If a rectifier tube shorts, you have blowing fuses and possible circuit damage. Transformers could be stressed, AC could reach the filter caps, etc. A series diode means that if the tube shorts, the diode is there to continue the rectification.

[R.G.]

Didn't we thrash this out in depth here sometime back? R.P., weren't you involved with this?

For the record, as Enzo states,
- it's a good idea in terms of disaster protection from rectifier tube failure
- it's cheap and easy
- where have you read this? They're getting the idea hosed up.

[soundguruman]

I've read about several people on DIY sites placing diodes in series with tube rectifiers such that the tube only receives the positive voltage pulse, the theory being that not seeing the negative supply swing is somehow easier on the tube rectifier.

Does anyone think that there's any merit to this idea, or is it just a pointless exercise that sounds like a good idea?

.

This idea is moronic...

Not seeing the negative swing:
all that does is reduce the output of the rectifier tube.
(maybe they wanted that on purpose)
are you gonna re-design the amp for lower B+?

Protection against short circuit:
If the tube is shorted, the current goes thru the diode, regardless.
I mean WTF?
why don't you just install a FUSE?

[Enzo]

No, the rectifier tube itself blocks the negative half of the wave form. Only positive current passes through the tube. So it doesn't matter whether the tube itself rectifies the AC by itself or if only positive going pulses reach the tube in the first place, either way the exact same full wave rectification occurs. The only difference would be that the diodes themselves will drop something between a half a volt and a volt from the result. Do you think reducing the B+ by 1 volt requires redesign? The rectifier tube is a pair of diodes, just not silicon. Having a silicon diode in series with a tube diode is not much different from having two silicon diodes in series. Yes, there is a much larger voltage drop across a tube than silicon.



When a rectifier tube shorts, that means a plate and a cathode short together. A tube alone will then wind up with its other side dead across the AC winding - fuses blow. A series diode thus protects the rest of the circuit, because when that plate shorts to that cathode, there still is the diode there to rectify. So instead of blowing a fuse, the circuit continues to work. Only downside is that B+ will rise some - the amount normally lost through the tube.


In fact, with added silicon series diodes, one could replace the rectifier tube in the socket with a jumper plug.

[EFK]

In fact, with added silicon series diodes, one could replace the rectifier tube in the socket with a jumper plug.

Yes, kind of a nifty option to have. I have done exactly this in a pinch, while waiting on a new rectifier, and only needed to temporarily adjust bias to compensate for the higher B+.

[Enzo]

And I can recall a couple times when I didn't have a rectifier tube at hand, so I bent a couple 1N4007s and stuffed them into the holes of the rectifier socket so I could continue a repair.

[R.G.]

This idea is moronic...

Well, how about just misguided? As I said, the source of the idea has the motivation wrong. Putting in back-up rectifiers doesn't help or hurt a fully working vacuum rectifier tube. It's what happens when the rectifier tube fails that gets interesting, and worth putting in the $1.00 or less of silicon.

Not seeing the negative swing:
all that does is reduce the output of the rectifier tube.
(maybe they wanted that on purpose)

Well - doing some parsing; reducing the negative swing isn't a big deal. However, reducing the negative swing does not materially reduce the output of the rectifier tube. A vacuum rectifier has a hard time getting under 40V for a forward drop. Adding a silicon diode which increases this forward drop all the way up to 41V doesn't seem like too much of a hit, expecially when the output voltage is a few hundred volts, and the ripple on the output voltage is large too. It would be hard to reliably *measure* the drop in B+ from this.

are you gonna re-design the amp for lower B+?

Fortunately, that's not needed. Losing a volt or two off B+ will be hard to measure, let alone cause problems.

Protection against short circuit:
If the tube is shorted, the current goes thru the diode, regardless.
I mean WTF?

Yes - and that is precisely the idea. If the tube shorts, the diode *prevents" the tube short from letting current flow backwards, saving wear and tear on things like the filter caps and power transformer. It turns a big failure into a soft fail. When the vacuum rectifier shorts, the B+ rises about 40-50V, and the amp keeps on playing.

So the right question is "are you gonna re-design the amp for HIGHER B+?" and the right answer is "no, but I will check my filter cap voltages to be sure the higher voltage doesn't harm them, because I *like* the amp just going on playing until I can get a replacement rectifier tube".

why don't you just install a FUSE?

Fuses have a really poor track record at letting you go on playing. A fuse turns one failure (the rectifier tube) into two failures ( the fuse AND the rectifier tube) in the hope that the second failure will prevent third through N failures.

[R.G.]

Yes, kind of a nifty option to have. I have done exactly this in a pinch, while waiting on a new rectifier, and only needed to temporarily adjust bias to compensate for the higher B+.

Even better, you could keep a tube base on hand with a power resistor between the pins so that the combination of silicon diodes and resistors closely faked the B+ response of the original vacuum rectifier. This is in fact what's in the highly-thought of Weber "copper cap" rectifier replacements.

[Diablo]

I've put a pair of SS diodes on the socket pins of each of my clone amps. Peace of mind in case of rectifier tube failure.
And my golden ears can hardly hear the presence of the SS diodes mucking up the tone of the vacuum tubes.......

[loudthud]

There are reports of rectifier tubes shorting plate to plate. This is more common with the low quality of current production but I have a GZ34 from the 70's with practically no hours on it that has the plate to plate short. It was that way right out of the box.

[g1]

It turns a big failure into a soft fail. When the vacuum rectifier shorts, the B+ rises about 40-50V, and the amp keeps on playing.

Dang! Now we need another indicator light to tell us when the rectifier has failed?

Seriously though, if it shorts are you going to notice the difference in the sound?
And do the diodes in any way affect the sag when the tube rectifier is good?

[trobbins]

I've happily put those protection diodes in to a restoration - a very easy addition to make.

The question of what happens when say a valve diode does go short is valid. The rectifier balance gets worse, so there would be long term increased ripple exposure to the filter cap - that in itself may be noticeable, and certainly measureable, and hence be the easiest indicator that the valve diode needs replacement. So checking the ripple voltage would added to the maintenance list (ie. along with checking output stage tube bias current levels). The increased ripple may show up in the audible output of the amp (with no signal input) - whether as an increase in low rms voltage, or via a spectrum analyser test - which would be less hassle than opening up the amp and probing ripple voltage level inside the amp.

It's probably a bit pointless to add a resistor-LED-zener type circuit across the valve diode to act as a continuous indicator of correct operation.

[R.G.]

Dang! Now we need another indicator light to tell us when the rectifier has failed?
Seriously though, if it shorts are you going to notice the difference in the sound?
And do the diodes in any way affect the sag when the tube rectifier is good?

Those are remarkably astute questions.

I do know that (1) silicon diodes and a power resistor to simulate a tube rectifier's forward drop do a remarkably good job of emulating the tube in the EE/analytical sense, (2) this *is* what's in a Weber tube emulator, minus the addition of a thermistor to soften power-on inrush, and (3) some very picky people of my acquaintance can indeed hear when there's a copper cap versus a tube rectifier in an amp... about 50% of the time. A couple of them accused me of deliberately faking the test, leaving the tube in all the time to fake them out.

I'm not going to try to convince you that diodes plus the right power resistor is as good as a vacuum tube rectifier - your ears will know if you do a fair test of sound only - but given how few of the remarkably picky people I've seen try it can even say when it's a solid state versus vacuum rectifier, it brings up the pertinent question of ... why not leave the diodes and resistor in there all the time? 8-|

That sure solved the indicator light problem!

To do a more appropriate answer to your questions,

Seriously though, if it shorts are you going to notice the difference in the sound?

You sure could. The B+ goes up a bit, and the power supply impedance goes down, so the amp is just a fraction more powerful. This might be noticeable.

And do the diodes in any way affect the sag when the tube rectifier is good

Everything affects everything, but I have a hard time trying to think how I would even measure this. The rectifier tube causes sag, along with the PT wiring resistance, because the higher DC current through it makes B+ sag a little, and this gets to be a kind of compression if you know where to listen for it. Within the range of normal operation, the vacuum tube rectifier's internal resistance causes B+ to be about 40-60Vdc lower than it would otherwise be because of the tube's internal resistance. Tubes don't have noticeable internal inductances to slow current down at these frequencies and levels. The drop caused by the vacuum rectifier is maybe 50V, the diode adds about 1V, and that drop is *constant* with changes in power. You could ask the same thing about lowering B+ by a few volts by putting diodes in the CT return; they're both in series with the rectifiers.

So - do they affect in in any way? Sure - but I think you'd need ears like a deity to hear it. And I'd *love* to run the blind test to see if people could hear it.

Afterthought: there's a simple way to determine these answers: get a high voltage rated DPDT switch, like a high-voltage-rated mains switch, and hook it up so you can flip the switch and short or unshort both silicon diodes at the same time. The voltages involved will be hard on the wiring, and you'll have to be @#$%$ careful rigging this up, but you can switch the diodes in and out and listen to the differences.

Same thing, another way, in that you can rig a switch to simulate a rectifier tube short. I think the rectifier tube and silicon diodes will survive both of these. Don't quote me on that.

[Alan0354]

For the fun of it, I look at the current vs voltage drop across 5U4GB
http://www.mif.pg.gda.pl/homepages/f...93/5/5U4GB.pdf

It is not a straight line, so it's going to be different from the behavior of resistor. BUT, this is not a signal, this is just amount of sag. The effect should be secondary or even lower. Sounds like if some want to get rectifier tube effect but want to design a cheap amp, diode and resistor is the ticket!!

I tend to think putting diodes on top of the rectifier tube is more troublesome. It will change a little of the sound if the tube blows. For you to know what happen, you have to have an indicator, that makes it a lot more complicated than it's worth. Just let the amp blow and fix it!!! I am surprised to see so many tubes gone south in tube amps when helping people trouble shooting in the forum, this is just one of the many tubes. You want tube amp, it's going to be unreliable no matter what, don't want it to fail, you have much better chance with a SS.