Help with VVR2 for Express that I fried

[bcmatt]

I realized the EL34s in my Express were not holding up anymore... it's ok; they have been in there for like 7 years. Just bad timing because (I realized as I was trying to bias them) 2 out of my 3 backup matched pairs are also not even closely matched anymore!
The one good unused pair burst into flames and blew a fuse when I tried to bias them. I think I accidentally touched an extra socket pin. After getting a new fuse into my amp, I realized the VVR no longer affected the B+. I was able to bias the amp normally with the VVR all the way up. That was thursday.

Well, today I tried swapping out the main Mosfet (NTE2973) that I had from another VVR2 that used to be in my JCM800 2204 build. That seemed to give a good range of plate voltage again, but biasing the tubes could only get between 0 and 5 mA. I tried to steal the smaller Mosfet from my other VVR2 and that was a real gong show trying to get them off their PCBs... probably destroyed them both if they weren't already. Before even trying to bias, I am now only getting a range of 20-60 VDC for plate voltages on the power tubes. I didn't even get to reading bias settings.

Does anyone have some familiarity with VVR and able to tell me what the heck is going on? I also did replace a zener diode near the small mosfet. Perhaps I need to remove the VVR completely and see if the amp is even working properly without it anymore.
Does anyone know if Dana Hall even has these things for sale anymore? I think I'm definitely going to at least need to buy some new components. (if I can find available equivalents- I know nothing about transistors) .. but I think I should probably replace the PCB at this point as well.

I love my Express, but there is no way I could ever use it without VVR to keep the volume under control.

[aceofbones]

I'm not familiar enough with the VVR to help with the problem, but I did speak with Dana about a year ago and he had no more boards, and was not going to be making anymore. You'd either have to salvage what you've got or maybe post a wanted post to see if anyone here has one laying around.

[martin manning]

Perhaps I need to remove the VVR completely and see if the amp is even working properly without it anymore.

That's a good place to start.

I think I'm definitely going to at least need to buy some new components. (if I can find available equivalents- I know nothing about transistors) .. but I think I should probably replace the PCB at this point as well.

I would replace all of the semiconductors. There are subs for the original NTE MOSFETS, IRFPE50PBF = NTE2973 and IRF9620PBF = NTE2372. I would also consider replacing the 12V Zeners with 9 or 10V to limit the max surge current.

[wpaulvogel]

No vvr but I have a Fryette PS-100.

[bcmatt]

I would replace all of the semiconductors. There are subs for the original NTE MOSFETS, IRFPE50PBF = NTE2973 and IRF9620PBF = NTE2372. I would also consider replacing the 12V Zeners with 9 or 10V to limit the max surge current.

Awesome! That is actually a huge help. Now I know what to look for.
I removed the VVR board for now.
I tried to bias the amp for it's full 410VDC plate voltage, but could only could get as high as 23mV across the 1R resisters. That's only 9.4 watts dissipation per tube I guess, but it makes me wonder why my bias rage would be so low. That's as hot as I can bias it right now. I'll take test my other voltages, but it's back in the cab right now because I tried playing it like that. It's actually a more stiff sounding amp than it feels to me at my more moderate plate voltages.... but I'm not used to these extreme volumes... and I won't be able to use it this loud anyways.
I'll try to make sure things are generally healthy voltage-wise, but I don't see any use in changing any resisters to bias hotter if I'm waiting to re-install the VVR2 anyways.
Does anyone else find a stock Express circuit won't allow a range to get biased properly? Just wondering if something else got fried...

[bcmatt]

I finally got all my new semi-conductors. I ended up ordering from Newark. The IRF9620PBFs took a little while extra, but finally showed up yesterday.
I guess I need to sort out the cold biasing issue before I reinstall the VVR. Has anyone had this issue before where there is not really a usable range of biasing? Do you think something has gone vastly wrong?
If it's not that weird, I would assume a would try a slightly smaller value for R31 (the 47K resister that feeds to the bias pot)... like maybe a 33K to start...

[martin manning]

I finally got all my new semi-conductors. I ended up ordering from Newark. The IRF9620PBFs took a little while extra, but finally showed up yesterday.
I guess I need to sort out the cold biasing issue before I reinstall the VVR. Has anyone had this issue before where there is not really a usable range of biasing? Do you think something has gone vastly wrong?
If it's not that weird, I would assume a would try a slightly smaller value for R31 (the 47K resister that feeds to the bias pot)... like maybe a 33K to start...

Getting it working and correctly biased prior to reinstalling the VVR is a good plan. I would play with R29, the 220k resistor that connects the bias rectifier to the HV secondary, increasing it for a hotter bias range.

[aceofbones]

I had a similar issue with my Express build. I couldn't get enough negative voltage bias range, so I had to alter R29 as Martin suggests.

It was suggested I change out the 220k for a 150k, I was able to parallel the 220k with a 1M to get me in the 180k range and that did the trick. With the 220k my bias range was -20vdc to -30vdc. The change to 180k got me -25vdc to -35vdc.

The 150k suggestion would probably be best for my amp but how it is now works.

[bcmatt]

So, yesterday morning before your replies, I did repopulate my VVR board and reinstalled it. Interestingly, it puts the bias possibilities back into range... but also allows a pretty huge range. It still allows a full plate voltage of about 406VDC, so I set bias at about 42mV. 90% of the time I'm running the Plate voltage set down at around 60-80VDC. If I remember correctly, the corresponding bias voltage tracks down to about 6-8ish mV.