Fender Tweed Twin Reissue JB Edition

[tictac]

Merlin's suggested standby arrangement for SS rectifier if you must use a standby.....

TT

[pdf64]

I install these diodes in just about every modern amp that uses a GZ34 rectifier. The most common being the Vox AC30 CC where both the Sovtek and JJ tubes blow often. Since installing the additional SS diodes the GZ34s longevity has been greatly extended.
Maybe we are talking more about the improved PIV by using the additional diodes thus protection the tube ?...

I think that increased PIV is not the actual benefit here; rather, the GZ34 are probably still failing due to the reservoir's capacitive load being being 'hot switched' with the standby arrangement (thereby likely breaching the plate's peak current rating), but that failure is not readily apparent, as the failure mode is a-k or a-a short, which allows the series silicon diodes take over the rectifier function following those failure modes.

From http://www.geofex.com/tubeampfaq/TUBEFAQ.htm#ampmods see

Put 1500Volt, 1A silicon diodes in series with the two sections of your rectifier tube (if you have a rectifier tube) so that if the rectifier tube shorts, the silicon will save the output tubes, and power and output transformer. The B+ will go up about 50V when (if!) the rectifier tube shorts, so the amp will have a little more power and run hotter. This can still hurt modern manufactured power tubes if it goes on too long, so check the rectifier tube frequently.

Also consider that in a lower HT amp such as an AC30 type, PIV (~1kV at initial power up) is far below the tube's 1500V rating. A Super Reverb would get a bit closer to that limit, maybe 1200V PIV, but the SR uses a non-hot switching standby, hence probably don't have a rep as being hard on their rectifiers, in the way that standby equipped AC30 do.

It's disappointing that amp designers can't break the cycle of implementing such an unnecessary feature (standby) in a way that induces catastrophic failure; especially as the tube manufacturer's then get blamed for the inevitable early tube failure.
It's not as if Korg haven't had ample opportunity to learn from the Marshall designed and built TBX era, in which a hot switching standby was compounded by a wiring error that passed the HT current via the GZ34 heater, ie the HT output was taken from pin 2
https://www.thegearpage.net/board/index ... st-1036657
http://schems.com/bmampscom/vox/Vox_ac3 ... er_amp.pdf
http://www.mif.pg.gda.pl/homepages/fran ... /GZ34S.pdf

If standby were removed, disabled or simply not used, as per real 60s AC30, I suspect that modern GZ34 would give good service life in AC30CC. Of course, even then, series diodes are good mitigation for the inevitable, so please continue to add them, though as RG notes above, 1kV types are often marginal and the use of 1200 or 1500V rated diodes is suggested.

http://schems.com/bmampscom/vox/Vox_AC3 ... )%20SM.pdf
http://schems.com/schematicheaven.net/m ... _schem.pdf

[pdf64]

Merlin's suggested standby arrangement for SS rectifier if you must use a standby.....

TT

But with a tube rectifier, that arrangement is bad; standby should not break the current path between the reservoir and rectifier, see http://www.valvewizard.co.uk/standby.html

[tictac]

Sorry, my bad, you are correct.... here's the correct tube version standby. I've worked on so many Twin Reverbs that when I see the word "Twin" my brain says, "you mean Twin REVERB" right?

[John_G]

Hi,
Aside from manufactures exceeding the MAX filter capacitance, some ignore the minimum HT secondary resistance. ( see Mullard spec) I must measure the VOX AC30 CC secondary resistance.
John

gz34-5ar4-cv1377-valve-rectifier.jpg

[John_G]

This time.............

gz34-5ar4-cv1377-valve-rectifier.jpg

[pdf64]

Bear in mind that it's more expensive to make a transformer with a lower winding resistance than 'normal'; thicker wire requires more copper and larger windows in the laminations are required to accommodate the thicker wire, which requires larger laminations. And the reservoir cap value in the AC30CC looks to be 44uF at most.
So it's unlikely that either of those factors are stressing the rectifier.
Whereas 'hot switching' the reservoir cap on to the rectifier is bad practice that will likely exceed the plate's instantaneous current rating.
I acknowledge that hot switching isn't mentioned as a no no in GZ34 info sheets.
All I can suggest is that it was just so widely understood to be bad that it was 'taken as read'.
And why would it be done anyway? I think that guitar amps are the only receiving tube application that featured standby, and that was probably only introduced so that underrated caps could be used http://www.valvewizard.co.uk/standby.html
See notes regarding hot switching constraints in RCA RC30 http://www.mif.pg.gda.pl/homepages/fran ... /5R4GB.pdf

[a j shaner]

Bear in mind that it's more expensive to make a transformer with a lower winding resistance than 'normal'; thicker wire requires more copper and larger windows in the laminations are required to accommodate the thicker wire, which requires larger laminations. And the reservoir cap value in the AC30CC looks to be 44uF at most.
So it's unlikely that either of those factors are stressing the rectifier.
Whereas 'hot switching' the reservoir cap on to the rectifier is bad practice that will likely exceed the plate's instantaneous current rating.
I acknowledge that hot switching isn't mentioned as a no no in GZ34 info sheets.
All I can suggest is that it was just so widely understood to be bad that it was 'taken as read'.
And why would it be done anyway? I think that guitar amps are the only receiving tube application that featured standby, and that was probably only introduced so that underrated caps could be used http://www.valvewizard.co.uk/standby.html
See notes regarding hot switching constraints in RCA RC30 http://www.mif.pg.gda.pl/homepages/fran ... /5R4GB.pdf

Regarding the reservoir cap value, the stock value is 20uF+20uF @ 600VDC, Sprague Atoms are available in that value but their physical size makes it impossible to mount them in a Bassman and Twin.
As an alternative I have used 2-80uF/450v e-lytics in series with a pair of 220k/1W resistors as the reservoir with 3-20uF/500v caps on a fiber board and use a small blackface cap can.
I have never had a problem with the rectifier on my 5f6-a Bassman. While I agree that hot switching is a design error, why would it not show up in the 5f6a?

[John_G]

Hi,
Just measured a VOX AC30CC amps HT winding resistance and got 52 ohm and 50 ohm with respect to the centre point.
Looks like I will be installing 47 ohm resistors or there abouts in each leg to the GZ34 Anodes from now on.(maybe the diodes as well)
John

[Roe]

The best alternative on ac30s, imho, is solid state rectifier together with a CLC (Pi) filter. A big choke has some resistance, almost as much as a gz34 (60 ohms circa). and with a 50 ohms cathode resistor on the el84s the current draw at idle is soo high that there is not very much sag in the power supply when you turn up the amp.

I'm using the hammond 5h 300mA choke (the brian may choke is 7h, 260mA, 78 ohms) and a 470R shared screen resistor.

See here for more info: http://ampgarage.com/forum/viewtopic.ph ... ma#p277832

BTW fender also used CLC filters on the earlier tweed amps. I prefer the 5e1 champ to the 5f1. But not all amps sound better with a CLC filter. The ac30 does however. No ghosting and very tight bass

[pdf64]

...I have never had a problem with the rectifier on my 5f6-a Bassman. While I agree that hot switching is a design error, why would it not show up in the 5f6a?

The peak current through the rectifier plates will depend on the total equivalent resistance in series with each plate plus the ESR of the reservoir. And maybe the particular rectifier you've got in there is tough. Also with a 5F6a there's a 47nF film reservoir, perhaps that's fast enough to dump a bit of charge into the main reservoir when standby is flipped, and it may be that all these factors combine to keep peak current just about ok and save the day?
The point with bad practice, not obviously risky behaviour etc, is that generally we get away with it; if it caused a problem every time, most of the time or even just quite often, Darwin's theory means we'd work out ourselves not to do it. Whereas instigating and maintaining good practice requires effort, self discipline, self review and even external audit.
Perhaps the thing to take note of is that Fender stopped using hot switching.

Whatever, being supposedly a beefier amp, the 5F8a may well have lower equivalent resistance in its HT supply than a 5F6a, thereby making the hot switching arrangement more stressful to its rectifier.

...measured a VOX AC30CC amps HT winding resistance and got 52 ohm and 50 ohm with respect to the centre point.
Looks like I will be installing 47 ohm resistors or there abouts in each leg to the GZ34 Anodes from now on.(maybe the diodes as well)...

The PT HT winding of the Woden in my 64 AC30 measures 65 and 62 ohms.
I've had the amp since ~1981 and never changed the Brimar GZ34, for all I know it may be the original.
I've added series 1N4007 just in case it ever goes but it may see me out, even with the 47uF reservoir I've currently got in there.
The HT is still ~320V so I know it's good (ie not shorted out yet)
So I wouldn't bother adding any series resistors in AC30 CC, much better to do something about that knucklehead standby.

...A big choke has some resistance, almost as much as a gz34 (60 ohms circa). and with a 50 ohms cathode resistor on the el84s the current draw at idle is soo high that there is not very much sag in the power supply when you turn up the amp...

FYI the equivalent resistance of a tube rectifier isn't linear, rather it differs according to the current draw on it.
My finding is that comparing a GZ34 to silicon diodes plus a series a sag resistor, when feeding a class AB 2x6L6 amp (idle HT ~420V), at idle a sag resistance of ~80 ohms gives parity, but at full load >200mA, ~20ohms gives parity.

[Roe]

Yes a gz34 is not linear. But ac30s with the 50R k draw so much current at idle that the non-linear effect is fairly slight in this case.

[PaisleyTube]

Many, many Fenders did use very small transformers, mostly the PTs.
See the PT as internal resistance between mains and a thing that amplfiies/creates sound and it kinda makes sense.

Maybe we've to check what tube-rectifier mr K. Rochards is using in his 5F8As on stage?

[Guy77]

This was a really informative read. I am interested in installing a GZ34 rectifier in a future 100 watt JM build, I love the sag that the rectifier gives and have tried to duplicate it with various resistors but it's not quiet the same.

I have read through this post and its stated that the GZ34 will max out at 250ma but the power transformer used in the JB Twin amp I believe can provide 500ma of HT current and the output transformer is 2k primary which I thought will also cause the tubes to pull a lot of current.

Isn't there a danger here that when pushing the amp hard , and the tubes try to reach there 100ma each of potential current draw, that the rectifier will blow? Or is the reasoning here that since the tubes are biased low ( 30ma each for example) they will never reach there 100ma capability?
I may have missed something when reading this and apologise in advance for opening an old post but it's such a great read!

Regarding Chris's question about what rectifier Keith Richards has in his High Power Twins. I graded a screen shot of the back of his amp from the video below and the labelling and red marking look a lot like a Groove Tubes GT 5U4. The glass bottle flares out after the base while the GZ34 does not flare out and is narrower.

https://www.youtube.com/watch?v=0QvftmhE9pw

High power Twin KeithR.PNG

Cheers

Guy

[TUBEDUDE]

This isn't anything truly new. Fender hasn't had a new idea in decades, (not including their forays into sandy state amps). Same old
12a_7's and 6L6's, same goofy interactive tone stacks, poor loops, same same same.