Any advantage in keeping selenium rectifiers and old buffer capacitors?

[Djack51]

All,

While reading a (french) forum, I come across an active member over there vigorously advocating to keep the selenium rectifier(s) (as well as old buffer electrolytics) in a vintage tube amp, "because these rectifiers have subjective qualities closer to valves than silicon diodes". To what extent do we have to believe that they would be better?

Now selenium diodes/rectifiers may still do their work, but it would seem that they can fail sooner or later, maybe with really unwanted side-effects.
Is there really an audible advantage over silicon?

As for old electrolytics, even the really high-end ones from 50 years ago are not necessarily better than decent ones from today's production?

I understand one wants to keep old equipment as much as possible in its original state, but when I want to use an amplifier from 1961 on a daily basis, I'm happier if the parts that are bound to fail have been replaced.

In short, I'm probably seeking for opinions confirming that the monsieur from the french forum is exaggerating.

Cheers,

Jacques (not a native french speaker)

 

[whoaru99]

I think keep them, if you're interested in potentially needing to fix bigger problems in the future. If the goal is less problems then replace them.

 

[petehall347]

if you fit a new selenium rectifier it should be ok .

 

[peterh]

All,

While reading a (french) forum, I come across an active member over there vigorously advocating to keep the selenium rectifier(s) (as well as old buffer electrolytics) in a vintage tube amp, "because these rectifiers have subjective qualities closer to valves than silicon diodes". To what extent do we have to believe that they would be better?

Now selenium diodes/rectifiers may still do their work, but it would seem that they can fail sooner or later, maybe with really unwanted side-effects.
Is there really an audible advantage over silicon?

As for old electrolytics, even the really high-end ones from 50 years ago are not necessarily better than decent ones from today's production?

I understand one wants to keep old equipment as much as possible in its original state, but when I want to use an amplifier from 1961 on a daily basis, I'm happier if the parts that are bound to fail have been replaced.

In short, I'm probably seeking for opinions confirming that the monsieur from the french forum is exaggerating.

Cheers,

Jacques (not a native french speaker)

We know that selenium rectifiers will age and change properties. We know that electrolytic caps will age, although slower, but both will when degraded
affect sound quality. Failing they might cost both tubes and transformers.

Replacing them _should_ only be positive to sound, we know that replacing them IS positive for reliability.

 

[wa2ise]

if you fit a new selenium rectifier it should be ok .

Who makes them? Anyway, if you want to preserve the looks of the insides of the equipment, you could leave the selenium rectifier in, and use one, and only one of its terminals to mount a new silicon rectifier. The other terminal should have nothing on it.

 

[Naptown Rob]

Bell Labs was worried about aging and failure in Selenium Rectifiers as far back as 1953 (reading gets good around page 18)

http://www3.alcatel-lucent.com/bstj/vol32-1953/articles/bstj32-6-1469.pdf

Mr. French Person can keep his, if he desires.

http://www.youtube.com/watch?v=zU8C5p7CQ_c

 

[diamondsouled]

New diodes and electrolytics are relatively cheap. Vintage transformers, not so much; reason enough to get rid of 'vintage' diodes and lytics. :thmbsp:

 

[KiM3Ce]

The selenium diode may have a bigger voltage drop than a modern silicon diode. This can be corrected with a resistor in series. Be sure the value and power rating are appropriate. If you have a schematic and voltage chart you can do the calculations.

Both of the last two posts are correct. I would discard the selenium rectifier to prevent some future owner from soldering it back into the circuit. Who cares whether the inside of your unit looks exactly like it did in 1960? Also make sure the selenium doesn't get eaten by kids or pets. You could sell it to the French guy but that would be wrong!

As for subjective qualities, the willingness of people to make baseless pronouncements is amazing. I suspect selenium rectifiers were chosen for cost and availability, not to "sound like" tubes.

FWIW the subjective consensus nowadays seems to favor fast-recovery diodes such as HEXFREDs. Switching from valve rectification in the power supply to solid-state rectification might make a subjective difference.

Look inside the cost-no-object new tube units from Jadis or Leben; I doubt you will find any selenium.

When I was young and dumb I plugged in an old amp at an estate sale and a puff of smoke came out. Replace those electrolytics! Fortunately that piece was repairable; you may not be so fortunate.

Good luck and thanks for posting this question.

 

[nashvillebill]

I'm not aware of any new selenium rectifiers still in production!

However, way back, selenium rectifiers were the first solid state rectifiers to hit the market. They were selected because they were cheaper than tubes at performing the basic task of rectification. However, their performance was relatively poor and reliability was questionable. So when silicon diodes were perfected, they took over and made selenium rectifiers obsolete.

There was absolutely no selection of selenium rectifiers to meet some imagined audio criteria. The primary goal was (and still is) to rectify AC into DC. Selenium did it cheaper than tubes. Silicon does it more reliably and more efficiently than selenium.

 

[Tom_r]

new diodes and electrolytics are relatively cheap. Vintage transformers, not so much; reason enough to get rid of 'vintage' diodes and lytics. :thmbsp:

++1

 

[Flyquail56]

There seems to be general agreement these days that fast recovery diodes subjectively sound better that standard silicon diodes, so from that I might infer that there is a subjective difference between silicon diodes and selenium rectifiers, however small. And which is better would be debatable. And any potential subjective superiority would not be worth the risk of failure and the collateral damage.

Seems like it kind of falls in line with the older-is-better school of thought, i.e. early triode power tubes, field coil speakers, cotton covered Western Electric hook-up wire, etc.

No doubt there is someone who can hear a difference between the two, or from changing one carbon comp resistor to carbon film, or from changing the angle of the lampshade in the listening room. I'm glad that I can't hear that well

Mike

 

[Djack51]

Thank you all for the answers so far. Subjective qualities of selenium, they seem to be something very subjective indeed.

In a different amplifier, I have replaced selenium diodes by 1N4007 and 50 uF buffers by 220 uF. The result was impressive: the amplifier all of a sudden produced that silky, seemingly effortless, detailed wide soundstage. Some friends have listened to it in disbelief, how can such an ugly thing sound so good?
I must say that other capacitors and the tubes (12AX7, 6BQ5) have been replaced, too. The power supply changes made a real difference. The amp may not be original anymore, but the original concept is still there and it sounds better than it did before. There was a certain uneasiness in the bass (100, 150, 200 Hz harmonics?) that has completely disappeared -- may be related to refreshing the tubes as well (cold-war russian 6P14P-EV have replaced the worn-out EL84/6BQ5).

This amp with a circlotron topology has four separate HT supplies with half-phase rectification (one diode) only and it produced a slight hum -- with the new buffers, it is hardly audible anymore. Very hard to install a bridge-style rectifier, one power winding on the transfo tab is directly connected to a heater filament winding (each output tube has it's own heater winding) and no apparent way to separate them.

To conclude, refreshing the power supply has brought back the qualities the amp was intended to have -- a circlotron circuit has it's merits. This one was produced by a local manufacturer, designed by an engineer who had worked at Philips in the late 1950s.

Thank you all for supporting my subjective experiences!

 

[shelly_d]

Keeping those in circuit benifits only the parts supplier down the line when you need to replace the parts damaged when those things fail.

Don't ask me how I know.

Shelly_D

 

[wa2ise]

You could use solid state Damper diodes (these are usually rated around 1000PIV and a few amps) to replace the selenium diodes, and you could mount them similar to the way the seleniums are mounted. Below is what the damper diodes look like inside my SCA35.