diytube.com

2 JJ Gz34 died within 3 months here. With a lightshow, ok, but I'll never buy JJ again.

I'm going with a couple 1N4007 diodes in series to the plates of the GZ34 to add some PIV protection. Should solve the stability problem of the current production GZ34 tubes. According to what I've been reading, all current production 5AR4/GZ34 tubes aren't as stable as the 20 year old NOS ones. It's just a time delayed rectifier.

You don't have to purchase 20 year old tubes for stability. The Sovtek 5AR4's work fine. Stay away from the JJ's and Chinese 5Ar4's and you will be fine. Do the diode mod and you won't go thru another rectifier tube.
I first performed this mod around 2002 in a pair of Quicksilver mono block amps I had. They are still working and with the rectifier tubes I had installed in them in 2002.

Hi TY,
Sorry for the dumb question!! Is the photo in your link above a modified portion of the ST70 wiring diagram?
Thanks, Vince

Hi Vince, No it isn't. But this is a shot of the backside of my modified rectifier. The fine folks at Antique Radio Supply got me my new parts rather quickly and just finished the soldering about an hour ago. The red wires just moved clockwise one tab. Diodes I chose were the 1N4007 types. I chose two in parallel because they are 1A rated and two would match the 1.9A rating of the 5AR4.

DC voltage at the first cap is now at 423V compared to 508V with the 5U4GB. Although the forward drop of a 5AR4 is less than a 5U4GB and would seem to have a higher output, the internal resistance is greater hence the lower voltage. Warm-up time is about 20 seconds which is greater than the warm-up time of the EL34s which is an excellent match.

http://answers.yahoo.com/question/index?qid=20080920183834AAIAv4W

A UF5408 diode would be my preferred choice but if it works with two diodes in parallel........of course I am the same person who has paralleled current limiters

Yup, your right about runaway, but has little effect here. Emitter resistors in transistor power amps with multiple output devices is the same thing. I'm just matching to the 5AR4 the device not the circuit. Worth measuring, but I don't think average draw ever exceeds 400mA. Idle is around 180mA if I'm not mistaken and the amp runs in class-A rather deeply. Actually, one is fine, but thought why not a pair cause I had them.

davygrvy wrote:Hi Vince, No it isn't. But this is a shot of the backside of my modified rectifier. The fine folks at Antique Radio Supply got me my new parts rather quickly and just finished the soldering about an hour ago. The red wires just moved clockwise one tab. Diodes I chose were the 1N4007 types. I chose two in parallel because they are 1A rated and two would match the 1.9A rating of the 5AR4.

DC voltage at the first cap is now at 423V compared to 508V with the 5U4GB. Although the forward drop of a 5AR4 is less than a 5U4GB and would seem to have a higher output, the internal resistance is greater hence the lower voltage. Warm-up time is about 20 seconds which is greater than the warm-up time of the EL34s which is an excellent match.

Hi, you suggestion was great and I have used single IN5408 between the plate and HV of the TX winding. Earlier to this modification, my JJ 5AR4 use to arc, blow fuse and now no problems. I have also put 0.01 mfd /2KV polyester cap with parallel to these diodes. I am using this in my AN 300B Parallel Monos. My question are as follows:
1. Whether it will slow down the voltage flow or lose linearity after inserting the diodes
2. Is it better to use 2 NOS UFIN4007 in parallel instead of a single 3 amp diode

6bg6ga wrote:You don't have to purchase 20 year old tubes for stability. The Sovtek 5AR4's work fine.

Yes, they do! Sovtek's are great.

As for diodes, I use MUR4100 in series and don't bother with the slower rectifiers that need bypass caps.

Cheers!

Geek wrote:

6bg6ga wrote:You don't have to purchase 20 year old tubes for stability. The Sovtek 5AR4's work fine.

Yes, they do! Sovtek's are great.

As for diodes, I use MUR4100 in series and don't bother with the slower rectifiers that need bypass caps.

Cheers!

Hi, Thanx for your suggestion!! How one do without these diodes if it slows down!!

abhijit wrote:Hi, Thanx for your suggestion!! How one do without these diodes if it slows down!!

You're welcome!

But sorry to say that I did not understand the second part of your question?

Cheers!

I was reading a few different forums on the topic of running a pair 1N4007 diodes to a GZ34 Rectifier. This is done in the photo on this post.

I did the mod the other night.

I read one post where someone said with the 1N4007 diodes you'd be dealing with the problem of "Switching". The post advised using the UF4007 in place of the 1N4007 as mentioned in this post and in the photo. The person was talking about the benefit of the Ultra Fast Diode in place of the 1N4007.

I was thinking of re-doing the mod with the UF4007... Thoughts?

abhijit wrote:Hi, Thanx for your suggestion!! How one do without these diodes if it slows down!!

"Slow" and "fast" or "soft" here refers to diode recovery.

Some more info here -

http://en.wikipedia.org/wiki/Schottky_diode

http://www.diyaudio.com/forums/solid-st ... iodes.html

http://www.diytube.com/phpBB2/viewtopic.php?p=10663

mantha3 wrote:I was reading a few different forums on the topic of running a pair 1N4007 diodes to a GZ34 Rectifier. This is done in the photo on this post.

I did the mod the other night.

I read one post where someone said with the 1N4007 diodes you'd be dealing with the problem of "Switching". The post advised using the UF4007 in place of the 1N4007 as mentioned in this post and in the photo. The person was talking about the benefit of the Ultra Fast Diode in place of the 1N4007.

I was thinking of re-doing the mod with the UF4007... Thoughts?

Now that you have the diodes in place you can just jumper out the GZ34 and toss it.

mantha3 wrote:I read one post where someone said with the 1N4007 diodes you'd be dealing with the problem of "Switching". The post advised using the UF4007 in place of the 1N4007 as mentioned in this post and in the photo. The person was talking about the benefit of the Ultra Fast Diode in place of the 1N4007.

I was thinking of re-doing the mod with the UF4007... Thoughts?

It is true that UF4007 has a faster switching time than the 1N4007. For this particular application (solid state diodes in series with the plates of a vacuum tube rectifier) I do not believe the switching time of the solid state diode is relevant. Regardless of how slow the diode switches, there is no possibility of reverse current flowing through the diode. You cannot pull electrons backwards through the vacuum tube. They simply will not go from plate to cathode.

It is the flow of reverse current through the solid state diode that causes the "switching noise". If you've already built it with 1N4007, I wouldn't bother changing it at this point.