Current Production GZ34

[Littlewyan]

My JTM50 uses a GZ34 and I was thinking of putting in protection diodes to keep the amp going if it shorts out. My only reason for not doing this is I may not know for a long time if the GZ34 has blown as the amp will keep working.

How long do current production GZ34s last? Is it really worth putting in protection diodes?

[tubeswell]

Protection diodes are only a couple of cents each. So yes, its worth putting them in.

[Littlewyan]

Would 1 x 1N4007 be ok on the output of the GZ34?

[pdf64]

My JTM50 uses a GZ34 and I was thinking of putting in protection diodes to keep the amp going if it shorts out

That would be a very good idea; cheap protection, uninterrupted operation - what's not to like?

My only reason for not doing this is I may not know for a long time if the GZ34 has blown as the amp will keep working

I don't see why that should be a barrier to installing them?

Would 1 x 1N4007 be ok on the output of the GZ34?

I don't think so; a plate to plate, or plate to cathode short would result in a significant fault current in the HT winding.
A series protection diode per phase is necessary.
The HT phase-to-phase voltages may be in excess of the 1kV Vrrm rating of a single 1N4007 per phase (eg if >320-0-320); series pairs, or a higher rated protection diode, would be a good idea.

[tubeswell]

Re; what pdf64 said about the Peak Inverse Voltage rating of the diodes.

If using a single SS diode on each side of the HT winding, you want the diode voltage rating to be 2.8 x the VAC of the HT winding. This would mean that if you used a 1N4007 (with a 1kV rating), that would be good for an HT winding that delivered 357-0-357 VAC (or 714-0 VAC) - assuming that the wall voltage and load conditions on the amp are such that the HT doesn't exceed that supply voltage.

Otherwise find some 1.5kV or 2kV diodes

[Littlewyan]

Ok i'll install diodes on each anode of the GZ34.

So a 1N4007 should be ok? Does putting diodes in series increase the voltage handling?

[martin manning]

So a 1N4007 should be ok? Does putting diodes in series increase the voltage handling?

Yes and yes. For this application a single 1N4007 in each leg is fine.

[Stevem]

If you want to keep the recto sag, yet have much longer lived recto's then all you need is the chassis room for two 250 ohm 25 watt resistors and then you need to scoop up four 1N5399 diodes !

[Stevem]

Also with this set up if a recto tube does go south at a gig then all you need to do is pull it, stuff in a new fuse and your playin again!

[pdf64]

For this application a single 1N4007 in each leg is fine.

Could you explain your thinking behind this?
My concern being that whilst 1N4007 do seem remarkably resilient, if the rectifier tube is shorted and the silicon diode/s are handling rectification, under power up and standby conditions their 1kV rating may be exceeded, especially if the mains voltage is running at its upper limit.
It would be unfortunate if both tube rectifier and the silicon protection diode were to fail short!

[Littlewyan]

I've just thought of an issue with adding these diodes. I'll need to put the wire for the bias circuit behind them as obviously it needs the AC.

I could use higher rated diodes so I only need 1 on each phase and mount them on the valve socket by using the unused pins, but I've heard this shouldn't be done as those unused pins are needed for voltage isolation.

[stretch2011]

My only two questions concerning this have been stated. But more answers would be nice.

1.) Will the diode protection reduce sag?

2.) If in the event the TR fails, the diodes would be providing a different secondary voltage. Which means in a fix biased amp the bias might need changed.

[Littlewyan]

1. There will still be sag as the current is still being drawn through the GZ34. Until it shorts then there probably won't be sag.

2. In my case the GZ34 I have has the same voltage drop as 2 x 1N4007 diodes in series on both windings. So won't make any difference. Depends on the GZ34 as some drop more voltage than others. Worst case it'll only be until the end of the gig that the amp is running at a slightly higher bias.

[martin manning]

For this application a single 1N4007 in each leg is fine.

Could you explain your thinking behind this?

The 1N4007's would be running at their rated PIV if the voltage across the outer legs of the PT reached 707VAC. Even amps with relatively high B+ won't exceed that voltage (even unloaded) on the secondary.

[Littlewyan]

What if the centre tap fuse blew?