Rectifier Tube Arcing - Has Rectifier Tube Quality Declined?

[Grantorino]

For those experienced - have you observed rectifier tubes arcing when powering up, at the moment the standby switch is flicked?
I have observed a 5AR4 rectifier consistently arc on startup. Two flashes, a second apart. The second arc sends a signal through the speaker. see the gif.

Perhaps I would not be otherwise concerned, but this amp previously suffered a failed rectifier tube. A modification was made to reduce stress on the rectifier. Swapping the C1 (47uF) and C2 (22uF) positions.

Does arcing inherently damage a tube?

-update- The rectifier tube had failed. The manufacturer recommends bypassing the use of the standby switch to avoid current spikes. They have also suggested that the quality of rectifier tubes has declined considerably in recent years.

Thank you.

[johnfromcyrene]

I put in a stand by switch and had arcing on mine. I had to remove the stand by switch and replace the tube. Not sure why this happens or how you prevent this.

[Lekky]

Not sure why this happens or how you prevent this.

It could definitely happen if the standby switch was wired between the rectifier tube and the reservoir cap. Putting the standby switch after the reservoir cap, or removing it completely, is a better solution with tube rectification.
Schematics and info on Valve Wizard's page about standby switching: https://valvewizard.co.uk/standby.html

[acheld]

If you post a schematic, we can give better advice.

Arcing is always a problem.  The question is -- is it the tube, the circuit design, or a fault in the circuit?

Since this is not the first time, there is a problem somewhere.

[Grantorino]

The Sovtek tube is toast. It was in place four weeks.

A newly fitted JJ rectifier tube does not arc.

 

Arcing is always a problem.  The question is -- is it the tube, the circuit design, or a fault in the circuit?

Since this is not the first time, there is a problem somewhere.

Thank you. I am inclined to agree.

[pdf64]

It's weird how people carry on using a standby switch even when with a moment's reflection, it would be clear that its use is causing significant problems.

[Grantorino]

It's weird how people carry on using a standby switch even when with a moment's reflection, it would be clear that its use is causing significant problems.

Thanks PDF - this is a salient point, as although the manual for the amplifier recommends the usual start-up procedure (1 minute warm-up in standby), having now spoken to the manufacturer, he recommends bypassing the standby switch altogether, leaving it always on.

[Grantorino]

Not sure why this happens or how you prevent this.

Putting the standby switch after the reservoir cap, or removing it completely, is a better solution with tube rectification.
Schematics and info on Valve Wizard's page about standby switching: https://valvewizard.co.uk/standby.html

Putting the standby switch after the reservoir cap will help prevent the current spike at switch-on, but will not prevent rectifier failure during operation. This would require changing the first filter cap to handle over 500V and i'm told would result in an audible 'pop' through the speaker when engaging the standby switch.

[Grantorino]

-update- The rectifier tube had failed. The manufacturer recommends bypassing the use of the standby switch to avoid current spikes. They have also suggested that the quality of rectifier tubes has declined considerably in recent years.

What is the consensus here on the quality, robustness and consistency in rectifier tube manufacturing in recent years? The manufacturer states that since the conception of the amplifier (approximately 8-10 years ago), the quality of rectifier tubes has declined considerably and this is the reason for repeated rectifier failures in this amplifier.

They have floated the idea of replacing the rectifier altogether with a transistor-based plugin device, such as the Copper Cap by Weber.

Any thoughts on this subject would be interesting to me.

Thanks again.

[stratomaster]

Reply #2 identified the issue and the remedy. 

Has rectifier tube quality declined over time? Probably, maybe even most likely, but that's irrelevant as there are plenty of recently produced rectifier tubes operating just fine in better implementations of standby.

[Grantorino]

 
Has rectifier tube quality declined over time? Probably, maybe even most likely, but that's irrelevant as there are plenty of recently produced rectifier tubes operating just fine in better implementations of standby.

I would say it is relevant to a customer without technical knowledge, that has purchased an expensive amplifier to play, where the manual tells him to warm up the amplifier in standby before every use.

When considering the cause of a failure in such an amplifier, whether it be the circuit (including standby position), or a weak rectifier tube, this does become a consideration.

I don't disagree about bypassing use of the standby switch altogether.

The quality of rectifier tubes is relevant to me and I have an interest in any thoughts on the matter, having now spoken to tube suppliers and tube amp manufacturers about the issue.

If the Sovtek factory is in Saratov, Russia, which just yesterday was being hit with drone strikes, the factory is certainly at risk. Not to mention the men from the factory floor being seconded into the military.

If, for example, a Vox amplifier blows four Sovtek rectifier tubes in quick succession, but runs well with a JJ tube, what does that suggest to you? (an anecdote relayed to me by a tube suppler).

In the specific case of the amplifier I am working on, I think it would be prudent to measure the heater and HV secondary pins of the rectifier, to see exactly the voltage demands placed on the tube.

I think it's all relevant...

[pdf64]

Adding back up silicon diodes stops arcs and hence shorts developing, keeps the valve rectifier operational,  perhaps until it loses emission.
https://www.premierguitar.com/the-immortal-amplifier-mod
The silicon diode voltage rating needs to be higher than RG Keen advises though, 1kV is insufficient unless the winding voltage is <550V AC.

[Grantorino]

Adding back up silicon diodes stops arcs and hence shorts developing, keeps the valve rectifier operational,  perhaps until it loses emission.

Thank you PDF.

 I had discussed this with the manufacturer. Their concern was that the if the rectifier tube shorts, the diodes allow continued operation of the amplifier - the protection diodes essentially become the rectifier -  but with a lower voltage drop, resulting in a large increase in plate voltage.

Their preferred fix was actually to fit a solid state plugin rectifier device, such as the copper cap by weber.

[stratomaster]

 
Has rectifier tube quality declined over time? Probably, maybe even most likely, but that's irrelevant as there are plenty of recently produced rectifier tubes operating just fine in better implementations of standby.

...
I think it's all relevant...

If you have a device with a design flaw that causes rapid battery drain when it's supposed to be powered off it doesn't matter if you put Duracells or Energizers in it. One might last marginally longer than the other, but still not as long as it should.  The issue isn't the battery; it's the design flaw.  Your expensive amp has a bad implementation of standby. Fix that, then it'll work with any rectifier tube you wish.  Or simply stop using the standby.

You shouldn't have to burn through rectifier tubes or purchase Golden Era NOS tubes to find one that can take the punishment of a hot switched standby (which was the real problem with the AC30 your tube supplier alluded to).

[Grantorino]

Thanks everyone for taking the time to reply. My mind is now settled on this one.

Speaking to the amp’s designer, he confirmed the standby switch inrush is the main mechanism behind most of the rectifier failures he sees.

I’m not sure i’d frame it as a “design flaw” specific to this amp, this standby switch topology -switch between rectifier and first filter cap - is seen across decades in tube guitar amps, Fender, Marshall, Vox, and most boutique builders included.

Whether that means the standby topology was always somewhat marginal, or whether it’s only become a problem as tube quality has changed, I honestly can’t say with any authority, wasn’t around in the glory days of production to know how these tubes originally held up in practice.

What has changed, per the designer and by my own valve supplier, is the margin modern production 5AR4 tubes have to absorb that inrush safely.

Practically, I am in agreement with the general consensus

Leaving the standby switch permanently engaged and using only the power switch removes the current spike entirely, at no cost and no modification.

Thanks again for all the contributions. I do appreciate the independent thoughts offered here.

[pdf64]

... I had discussed this with the manufacturer. Their concern was that the if the rectifier tube shorts, the diodes allow continued operation of the amplifier - the protection diodes essentially become the rectifier -  but with a lower voltage drop, resulting in a large increase in plate voltage.

With a GZ34, the increase in HT voltage should be marginal, normal mains voltage fluctuations typically cause more difference.
But nevermind, the whole rationale generally turns out to be incorrect.
Shorts in the valve rectifier are the result of arcs sustaining, which deposits conductive material from the electrodes along surfaces near the arc.
Whereas with the series silicon diodes, arcs can only last a half cycle, they can't sustain themselves beyond that.
So shorts in the valve rectifier don't form, the concerns turn out to be hypothetical, not actual.

Their preferred fix was actually to fit a solid state plugin rectifier device, such as the copper cap by weber.

Weber WZ34 have a higher resistance than a GZ34, so at full load drop far more voltage, than a real GZ34.
Plus the outer copper cover can become detached and expose the high voltage innerds (obviously valve glass envelopes can breaks and present the same hazard.

[Grantorino]

Weber WZ34 have a higher resistance than a GZ34, so at full load drop far more voltage, than a real GZ34.
Plus the outer copper cover can become detached and expose the high voltage innerds (obviously valve glass envelopes can breaks and present the same hazard.

For clarity, the amp designer did not mention specifically the Copper Cap by Weber, but generally a plugin solid-state device. I was using this as an example, but appreciate your thoughts on the Copper Cap and its limitations. I think you may be correct in suggesting that the magnitude of the concern regarding voltage drop was perhaps overstated. It's true that i've seen mains voltage here often drift by around 15V, depending on time of day etc. It's a good point.

On the diode point, I appreciate the detail. For what it's worth, my tube suppler supported installing protection diodes and I have done this once before on a JMI era AC30. I would say it isn't a settled question among those with experience. I'm not in a position to comment beyond that, except to say the designer of the amplifier advised against it. Perhaps he was thinking of a scenario where the rectifier develops a short for some other reason than reverse voltage

Personally, I have not seen a tube-rectified amplifier with protection diodes installed suffer failure of the rectifier tube.

Regardless, everyone seems to agree on the standby switch being an issue!

The owner of the amplifier is not interested in any further modification and is likely looking to sell. The amp runs well (for now) and sounds fantastic.