Damn modern rectifier tubes

[Randall]

A 66 Super Reverb came in with the rectifier tube socket converted to diodes. I wired the socket back correctly and put in the new TAD rectifier, and Spark/Pop! Dead fuse. Tried it again under different scenarios, same thing, lightning in the bottle. It even did it on a cold power up of the GZ34 while standby switch was on. Blew fuses with two different pairs of 6L6s. Runs fine on a used USA made GZ34, putting out a full 40 watts with a pair of GE 6L6s the guy had, standby switched several times while servicing. So this discussion of hot switching or not is moot to me. It's a AB763. The tube is faulty, full stop.

Now, what to do about this going forward?

[Jon Snell]

No mention as to the value of the tank capacitor.
Maximum value for a GZ34 is 60uF for the 1960s Philips/Mullard type or 50uF for newer builds.
Too large a value will destroy the rectifier.

[pdf64]

I retrofitted a 1N4007 in series with each 5AR4 anode in my son's AC30 CC2 combo as the simplest and best way (imho) to alleviate poor valve rectifier life due to the standby switch.

That’s a great method to protect the amp from the rectifier valve shorting. The amp seamlessly carries on working, albeit with a higher HT voltage.
But it would be better, especially considering an AC30’s already high idle dissipation, if the rectifier valve didn’t fail. And the most likely contender for the root cause of that seems to be the hot switching standby.

[Pedro Vecino]

Another detail, in addition to not exceeding the recommended capacity consists of minimum resistance measures in the transformer secondaries that many current manufacturers do not meet. In some amplifiers I have installed resistors in series to complement it.

[pdf64]

The stock Vox AC30-6/TB I think is a good example of Hot Switching. The heater of the 5AR4 tube is heated, but the plates have not yet been closed....the Standby switch accomplishes that. This has traditionally led to premature failure of that rectifier tube in those amps. Just had one fail on me today, changed the wiring so the plates and heaters are in circuit when the amp is powered up, so the tube warms up while slowly charging the HT supply caps. S/B mute now done by shorting the grids of the power tubes together.

The Marshall designed built AC30TB of the 90s wins my gold award as the worst standby ever, as it combines hot switching with a wiring error, apparently long standing, that sent HT current via the GZ34 rectifier heater (DC output taken from pin 2 rather than pin 8). So the start up current surge stressed the rectifier heater, occasionally leading to the unusual dead heater failure mode.
Of course, taking the DC output from pin 2 only really matters with an indirectly heated type such as a GZ34.
Edit - just seen in the other thread that you’re already on top of this issue

[Leo_Gnardo]

I retrofitted a 1N4007 in series with each 5AR4 anode in my son's AC30 CC2 combo as the simplest and best way (imho) to alleviate poor valve rectifier life due to the standby switch.

Did much the same in that Super Reverb that just got picked up by its owner. In this case, two 1N4007 in series instead of a single, on each anode. A bit more voltage standoff & less possibility of failure in case the tube rectifier shorted. I've done this a couple times. In the case of this Sup Rev because the amp seemed jinxed, to the point its latest return was due to power transformer failure - primary had gone open circus in spite of having proper 2 amp fuse installed.

FWIW I mount a tie strip near the rectifier socket, soldered directly to chassis with the big 80W Weller "pencil" iron.

May have to do this in every 5AR4/GZ34 amp I repair as a matter of course in the future.

[pdf64]

I hope / expect the several BVA (British Valve Association) GZ34s I’ve got for my own amps see me out.
I’ve not had to buy one for a customer’s amp for a while, the Sovtek GZ34 I got a few years ago for a Super Reverb is still fine, as far as I know.
Of course, I fitted protection back up diodes, so if it did fail, without checking, no one would be any the wiser

[trobbins]

I'd suggest there could be a few factors at play here with respect to the differences between vintage and new rectifiers, and how the collective experience is that new rectifiers are failing (which I would assume is via internal arcing).

Less uniformity/quality of the cathode can be a weakness that becomes noticeable with hot switching, as the peak current requirement from the cathode is at its highest and localised zones on the cathode could experience stress beyond the average capability - which can lead to cathode surface damage from such an arc, which then likely gets worse and worse if arcing is allowed to keep occuring.

A higher level of intrinsic gas in the tube is also plausible, due possibly to less rigorous pump-down during manufacture, and/or poorer anode/glass preparation to minimise out-gassing during normal life, and/or poorer performing getter.

The use of series 1N4007 stops the onset of any arcing that could occur, and hence is likely to allow a new valve to perform adequately for a fair lifetime, as it is not degraded at the first instance of stress/arcing. To me that keeps the amp performing as it was intended (ie. the standby switch can be used if desired).

The valve diode can be tested for reverse leakage using an insulation resistance tester - but not everyone has that type of tool.

[Leo_Gnardo]

The use of series 1N4007 stops the onset of any arcing that could occur, and hence is likely to allow a new valve to perform adequately for a fair lifetime, as it is not degraded at the first instance of stress/arcing. To me that keeps the amp performing as it was intended (ie. the standby switch can be used if desired).

I notice Fender has been using the series silicon rectifiers for some time now but I don't know if they are in all their currently made tube rectifier amps. Mostly I see DRRI's here maybe 6 to 10 a year and they all seem to have them. Fender's not using 1N4007, instead some recto with globe shape body - off white color - look like old GE rectifiers from the 1960's. For now I'll stick with what I got, and series them. Must have 1000 or so 1N4007 in stock. Way cheap with quantity discount. I'm pretty sure I got mine from Mouser maybe 5 years ago.

I'm not so impressed with Fender's method of mounting the helper silicon rect's right on the octal socket, IMHO the electrodes are close enough there's danger of arcing - but I could be wrong. That's why I use a tie strip to mount them when adding to a vintage amp. "An abundance of caution" - a little extra effort prevents any mishap. I hope.

[Pedro Vecino]

Fender uses BYD33V diodes (1400V) at least in newer Princetons. I think it is advisable to exceed 1000V of a 1N4007 in this function. I use BY269 (1600V) for it.

They keep the entire length of pins. Possibly to reduce mechanical stress and keep them away from the heat source.

[Helmholtz]

Thanks guys, this thread helped me to learn a few things (one of the main reasons why I'm here):

1) For some reason switching the hot rectifier tube to a discharged reservoir cap is termed "hot switching".
It was clear to me that this king of switching puts a lot of stress on the rectifier as the discharged cap acts as a short.
I didn't know the term though.

2) Wiring the switch after the reservoir cap as in the Fender circuit avoids the inrush current surge and is not called "hot switching" .

3) The reliability of current production GZ34s seems to be rather poor.
If a new tube fails in a Fender type circuit, there's no excuse for the tube/manufacturer.
I would claim and return the tubes.

4) Main tube failure mode seems to be shorting caused by internal arcing.
Arcing is facilitated by poor vacuum and driven/sustained by reverse current. Sustained arcing destroys the tube.
Reverse current can be blocked by adding series diodes.

[Leo_Gnardo]

Fender uses BYD33V diodes (1400V) at least in newer Princetons. I think it is advisable to exceed 1000V of a 1N4007 in this function. I use BY269 (1600V) for it.

They keep the entire length of pins. Possibly to reduce mechanical stress and keep them away from the heat source.

Big thank you Pedro for clarifying. Also, you mentioned in your previous post placing resistors in the high voltage feeds to the rectifier tube. What resistance value? From your photo it looks like they are 5 watt wire wound. TIA for your answer.

[Pedro Vecino]

I don't remember the exact AC30 model. In that case they are 68R / 5W. One for each side. A bit tight in terms of power dissipation but stable.

[Leo_Gnardo]

I don't remember the exact AC30 model. In that case they are 68R / 5W. One for each side. A bit tight in terms of power dissipation but stable.

OK, I'll stock up some of that or similar value/wattage. And for extra points, what is the secondary resistance requirement for 5AR4/GZ34? Again TIA.

[Helmholtz]

And for extra points, what is the secondary resistance requirement for 5AR4/GZ34? Again TIA.

For a 2x400VAC transformer it would be 125R per side minimum.

https://frank.pocnet.net/sheets/030/g/GZ34.pdf

The Rt consists of the secondary DCR per side plus the reflected primary DCR.
I found around 25R to 30R for the reflected primary DCR with JTM 45/50 PTs.