Tube rectifier alterations

[EFK]

Here's a shot from the layout of Ceriatone's JTM45 kit. They use diodes on the GZ34 socket probably as a failsafe. I like the idea of it, but here's an interesting question: what happens if you have the tube rectifier wired up like this, and you decide to plug in one of those diode rectifier replacements?

Will it work?

Also: in the layout diagram shown, why is the cap across the terminals? And why is it 3kV? Does it need to be rated that high?

[JHow]

Is Ceriatone having a sale this month? Because we were just talking about this here:

http://music-electronics-forum.com/t18213/

It looks to me like these diodes are in series with the tube rectifier, so if you used ss rectifier after them, it would be like having two strings of diodes, which should be fine, right?

[Enzo]

Diodes in series with diodes works the same as diodes.

[EFK]

Well this has become interesting. While I have been waiting for a GZ34, being impatient, I figured I'd take the Ceriatone approach and stick two diodes on there so I could run the amp w/o tube rectifier while I was waiting for it. So, I used two 16A/1200 V hexfreds and soldered them up to the inside of the socket, anodes to unused pins 3 and 5 and cathodes to 4 and 6. HT lines from the PT come in on 3 and 5 with the anodes. Then I just made a quick little 3-prong jumper to stick in the socket - jumps 4 and 6 (cathodes of the hexfreds) to pin 8 - so I could run the amp w/ the diodes only while I waited on the tube rec. Runs just fine this way. So today the tube rec arrives, I plug it in, and ... nothing. When the amp is on standby, I get 122 V AC on the power switch and about 445 V DC on the hot side of the standby switch. As soon as I switch the standby switch to 'on,' the VDC drops to almost nothing, then slowly comes back to only about 51 VDC. Suffice it to say the amp will not make sound.

So what is going on here? Is there some reason that I can't use hexfreds like this as per the Ceriatone diagram? I've read about doing this a few times, doesn;t seem to be a problem although most seem to use IN5408 for this. But, I thought I'd go a little heavier so I could run the amp on the diodes alone if need be and two IN5408 don't give me quite enough confidence. Obviously something is wrong.

Anyone?

[JHow]

Hmm. Maybe the experts will weigh in here. To my thinking all we have done is insert two ss diodes in front of the tube ones, which shouldn't prevent the circuit from working. In the ceriatone diagram they take the neg. bias voltage off in front of the ss diodes at pin 7, you have that, right?

[EFK]

The only difference is, for reasons of neatness in wiring my particular socket, I used 3(cathode on 4) and 5(cathode on 6) as soldering points for the anodes instead of 5 and 7. It works flawlessly with no rectifier. My guess is, for some reason which I don't understand, hexfreds don't work properly in this situation [series with tube rec]. ??? I could just stick some IN5408s or IN4007s on there but I'd like to understand the problem first. Also my 3-prong jumper is dangerous as all get out (imagine if someone touched it) and I will not use that as a permanent rectifier.

[JHow]

Yeah, I noticed you deviated slightly from the ceriatone scheme in pin choice, but I don't think it matters which unused pins you start at, only where you end.

Regarding the jumper, it does sound dodgy. Kind of leaving a high voltage connection right up on the top of the chassis where the tubes are. Air is a great insulator, but not a barrier.

Like I said, perhaps one of the gurus will weigh in, because I don't see why diode type would make a difference here.

[Diablo]

Do you have the PT 5V AC winding connected to pins 2 and 8 of your rectifier socket to heat up the filament? If you do and the rectifier still isn't working, you likely have a bad rectifier tube. I can't see how the type of diode would make a difference to the tube, as long as it's a good tube.

[Enzo]

So test your plug in rectifier. It could be defective. Can you plug it into some other compatible amp? Does its heater light up?

Don;t look for esoteric reasons for this not working. REctifiers are rectifiers, whether they are ultrafast hexfreds or slow as molasses whatever. If current comes through the jumper plug, a working tube should handle it fine.

[EFK]

Hmmm. OK, so the type of diode is not a problem. I tried 2 brand new rectifier tubes and they both did the same thing. One thing I don;t understand - if I'm getting appropriate rectified DC on one side of the standby, doesn't that mean the tube is working? So why does it drop to almost nothing when I flip the standby switch? It works just fine with the jumper so it can't be anything downstream sucking up all the power. I'm going to go check this over again...

[Enzo]

Does its heater light up?

DOn't believe I got an answer to this.

When the standby switch is turned on, there is a load on the B+. When it is off, there is not. The tube doesn;t have to do much of anything to put a charge on something not loaded. But it can;t supply the load with current.

IS the tube heater glowing or not? Nice and hot?

You might measure 5VAC or so across the winding, but an open winding can still show a voltage on your AC meter. After all, you can grab the probe tip with your fingers and get voltage.

Power off, no tube in socket, measure resistance from pin 2 to pin 8 of the socket. That will be measuring continuity through the winding.

[EFK]

Thanks for your time Enzo. Double-checked all my connections and made sure the heaters are working. They are working - tube gets good and hot, has 5 VAC on the heaters. Reflowed the solder connections anyway. Cleaned the socket (it's brand new but I cleaned it anyway). I bought 3 new JJ GZ34 tubes and just tried all three: every one does the same thing. Power on, 122.5 AC on the power switch and 444 DC on the hot side of the standby switch. As soon as I flip the standby, the voltage drops and it will only make 51/52 VDC both on the switch and on the main can. If I take out the tube and pop my little socket jumper back in place, the amp works fine.

Could there be any reason a .1uF/630V cap across the standby switch is causing a problem for the tube rectifier?

As I understand it, the whole point of the diodes on the tube rec is backup in case the tube fails. So, even if the tubes were crapped out, shouldn't the amp still work anyway on diodes alone?

I can't figure this one out.

[JHow]

Have you tried it with just the tube rectifier, that is, remove the diodes?

[EFK]

Well I just figured it out and feel kind of silly - it's always something simple. Apparently my little 'plug' opened up the pin 8 socket lug anough to make only an intermittent contact w/ the tube pin. At least, that's what I think it must have been. I noticed if I rocked the tube around with the amp the voltages went up and down. I tightened it back up, stuck one of the tubes in there and presto. Works perfectly.

You may all slap my forehead now.

[JHow]

It always a simple fault in retrospect.