VVR3 (fried the MOSFET ... not thinking)

[doveman]

VVR3 mounted in Dr Z Route 66 ... working great (see other post and clips here at ampgarage). Discovered a missing guide pin on one of my KT66 tubes. Decided to replace with Golden Lion KT66s ... rated at 30w (not 25w). Set it up ... rebiased and maxed it out ... VVR started acting funny. Bottom line ... looks like I blew the MOSFET. Dumb thing to do ... as I would probably never have maxed it out like that with the VVR on zero. But I think it pointed out a weakness in where I mounted the MOSFET. I need to remember that Dr Z underestimates watts ... I have to treat this as a 50 watt amp ... at the top of the VVR3 range.

Right now as the VVR3 pot turns, it scales the BIAS but not the B+. Went through the testing procedure ... and the troubleshooting for this issue and it seems to confirm a blown MOSFET.

So I plan to:

• Replace the MOSFET (need a mouser part number) Maybe 526-NTE2973
• Recommended replace Zener diode (need mouser part number & which diode is it?)
• Maybe relocate MOSFET via 3 wires and install heatsink (need a mouser part number for heatsink)

I have a note into Dana but thought I'd ask here too.

Here's where I have the MOSFET mounted now:

[IMG:639:423]http://i104.photobucket.com/albums/m168 ... C_0299.jpg[/img]

I can probably find a better place on the underside of the chassis ... or back leaving room for a heatsink on the opposite side. But is there a recommendation for heatsink for this MOSFET ... and maybe a mouser part #? I'll need to see the heat sink first.

Another option would be just to replace the MOSFET and diode but avoid stressing the amp at full bore ... especially with the more powerful tubes. Just a dumbass move.

Can you guys help with Mouser part numbers? Traveling next week and would like to order the parts and get them on the way. Emailed and PM'd Dana too.

[markr14850]

Are you saying that the MOSFET blew when the VVR was maxed out? If so, that's odd, as there should be very little power dissipated by the MOSFET under that condition (similarly at very low volume). The toughest state for the VVR is at 50%.

[doveman]

Are you saying that the MOSFET blew when the VVR was maxed out? If so, that's odd, as there should be very little power dissipated by the MOSFET under that condition (similarly at very low volume). The toughest state for the VVR is at 50%.

What I mean by maxed out ... full operation
- treble 10
- bass 10
- volume 10
- VVR 0

Amp wide open but VVR allowing the least volume. That would create the most heat possible in my mind. Sounds like you are saying that low heat with VVR at 0 and 10 ... most heat at VVR at 5. Is that correct? Not sure what moment it blew but I would have had it at many settings including the one I described. If what you say is true the 5 setting would be the one I would probably use the most ... maybe not the amp controls dimed. But that would mean I'd have to find a heatsink and relocate. Am I reading this right? If that is true ... good to know.

[Colossal]

Sounds like you are saying that low heat with VVR at 0 and 10 ... most heat at VVR at 5. Is that correct?

Hi Doveman,

Yes, dissipation by the MOSFET is greatest at about 5 (or half way down).

[doveman]

Sounds like you are saying that low heat with VVR at 0 and 10 ... most heat at VVR at 5. Is that correct?

Hi Doveman,

Yes, dissipation by the MOSFET is greatest at about 5 (or half way down).

Well that is good to know ... that means I definitely have to consider relocation/heatsink options. I wonder if the additional power of these tubes was just enough to take it over the top. It had been working flawlessly and almost always at the middle settings ... so most heat from what you are telling me.

The new tubes were in only minutes before I noticed this ... so I wonder ... it will give me a problem and solutions to consider.

I think I did find the MOSFET number at Mouser ... added it to original post. Still need to know which diode on the board, diode part number and heatsink suggestions. I hope Dana weighs in this weekend.

Thanks this is good information ... sounds like I would not have avoided my problem if I had stayed away from these settings.

[Reeltarded]

I think KT88 pair biased where they sound good is like 20% higher output than you think it is when it's cranked.

[doveman]

I think KT88 pair biased where they sound good is like 20% higher output than you think it is when it's cranked.

Maybe the same for KT66? I know this is the loudest 32 watt amp I've ever heard. Dr Z amps are known for that ... but I really hope I can get this operating again and avoid overheating ... it really is SWEET when the VVR is in operation.

[Reeltarded]

66s and El34s make around 50 watts where in the same situation 88s might push 70, apples/apples.

In a JTM45 setup a pair of 66s make about 40 watts. In a similar SS rec amp, about 50+ a little. I have to speak in Marshalltudes, because I know that language enough to cuss.

[doveman]

I'll do a few more searches ... maybe someone else has mentioned ordering these specific diodes here before. Maybe discussed the specific heat sink?

[doveman]

Dana contacted me and worked out a pretty fair deal for replacement MOSFET and diode. I did see a couple of different discussions on heat sink ... still not sure what to do there. I just think I'm still at risk ... if the highest stress on the MOSFET is at 5. Hope Dana might weigh in on this last issue.

Maybe something like this aluminum heat sink on the other side of chassis ... after relocating the MOSFET to another location.

[doveman]

Maybe move the MOSFET like this ...

From here to here ...
[IMG:639:423]http://i104.photobucket.com/albums/m168 ... C_0309.jpg[/img]

Top side view ...
[IMG:639:423]http://i104.photobucket.com/albums/m168 ... C_0311.jpg[/img]

with one of these on top side of chassis ... or maybe a couple to drain off heat from all (mosfet, transformer, rectifer) and leave it in the current spot.
[IMG:150:150]http://i104.photobucket.com/albums/m168 ... atsink.jpg[/img]

Entire chassis top side for reference ...
[IMG:639:423]http://i104.photobucket.com/albums/m168 ... C_0310.jpg[/img]

[doveman]

Bump ... for heat sink ideas. Traveling this week but would like to order some parts if I could figure out a practical heat sink.

[Reeltarded]

The best (and sticking out) place is the rear of the chassis..

Looks like you should pick the coolest place though.. huh? Keep it away from HEAT.. like an output tube. I don't see an ideal place for one. As exposed to air as possible, and as far away from external heat source.

Consider the sink is trying to move heat away, but being swamped.. yeah?

[Jana]

Yeah, what Reeltarded says--I think, lol. With the heat sink on top by the rectifier tube, it might be soaking up more heat than it is dissipating. Unless, maybe you had a little muffin fan blowing cooler air right on the heat sink.

[doveman]

These are good points ... if it's too close to the rectifier it would probably be drawing heat towards that area of the chassis. I could run it off the back near preamp tube V1. I'm wondering if it might create noise in that area of the chassis?

My original thought was that the MOSFET was located right in the middle of the side panel. Maybe locating a heatsink off the inside of the chassis right next to the current MOSFET location. I may only need a bit since it worked fine for many days at the settings that were a bit tougher than I would normally run.

That "stressing the VVR at 5 setting" knowledge would have been helpful information too. I ran it through some 10-10-10 settings with the VVR at several setting including 5 and 0 ... for no real reason.