Metropoulos Forum

Hello, I searched for threads, but could not find any on the GZ34 with two diodes & one .01uf cap mod. I have seen this mod on some schematics and I have read that it's fail-safe. Is this mod recommended and does it compromise performance/sound?

Here is an image...

Just saw this thread n thought I'd add my 2Cs, I asked a similar Q in the tube section not long ago n my Q seemed to kill the thread. I've been using the diodes n cap on two of my clones without a problem, no idea how it effects the feel of the amp, didn't notice any change when I swapped over the first one after a JJ GZ34 died but since then I've put old used tubes in both so I have no idea if the diodes n cap offer more reliability but I sorta have a bit more peace of mind with them in. Blind faith I'm sure

Hi, I asked about this on DIYaudio & Kevin O'connor advised against putting the diodes on the valve base as the unused pins on the rec r 4 "voltage isolation". He also advised against putting the cap across the diodes, saying NEVER. Also said diode switching hash is negilible. Cheers

ivan H wrote:Hi, I asked about this on DIYaudio & Kevin O'connor advised against putting the diodes on the valve base as the unused pins on the rec r 4 "voltage isolation". He also advised against putting the cap across the diodes, saying NEVER. Also said diode switching hash is negilible. Cheers

Thanks for the heads up Ivan, I just went over there n found the post you referred to and will add it here for anyone who cares. Sounds like a couple of 5w resistors in series with the HT is a better way to go.

If you look at the tube manufacturer's data for tube rectifiers, they recommend placing series resistance for each plate. This acts to limit the initial charging current of a noncharged capacitor.

A cap looks like a dead short at the instant power is applied. The only thing that limits the charge current is circuit resistance, comprised of the DC winding resistance of the plate winding of the PT, resistance of wiring and connections, dynamic resistance of the rectifier, and any deliberately added resistance used to create low-pass filtering for the supply.

Suppose B+ is 400V and there is 1R of total resistance in the charge path. The initial charge current is 400A. If you add a 10R resistor feeding the cap, initial current is limited to 40A. The data sheets for most tube rectifiers suggest values of 135R, which for the example limits the current to just under 3A. The problem with such a high value is that it also limits the operating current for the amp and introduces serious supply sag under load.

However, adding just a few tens of ohms will greatly improve tube rectifier life.

Hi, I just went there & re read it 2 make sure I got it right. It actually doesnt say NEVER put he cap across the diodes, it says it should be avoided. Yeah, sounds like the resistors r the go. I just asked him if he thinks protecting the OT & tubes from AC is needed if the GZ34 fails, or if the PT's HT secondry CT fuse is enough. We'll c wot he says. Ths forum is also a great place 4 learning. Cheers

Hi all. Ok, the consensus seems to be; to take th load of the GZ34 @ swtch on, a couple of current limiting resistors between the PT & the GZ34 plates, like 47 ohm/5W (400V). U could add 1between the rec & filter (up to 135 ohm tube manufacturers reccomendation), but it'll introduce considerable more sag. To protect the filter caps, OT & tubes from AC if the GZ34 fails, a single diode in the B+, straight off pin 8. Cheers