VVR and Red Plating

[yodermr]

Hey guys, posting over here, cuz I found more folks talking about the VVR...

So I built a Metro JTM45 a few years ago. Worked great, no problem. Loud as hell, year later, I installed a VVR (Dana) between the first cap and the choke and between the Bias trim and the 220K resistors. Worked great, no problem. Zoom forward about a year...

Loud hum, then back to normal. Switched it off. Back on and watched the tubes, spark shower storm inside one tube. Shut her down. K, Figured it was time for new tubes...

Got a pair of Gold Lion KT66s, installed. Going from suggestions on the Metro Amp board, tried to bias at .033 (65%) but I didn't have the needed range to go that low (got down to .039). So I change the bias resistor.

Biased her in at .034mV. Sounded great. Here is where it got weird on me.

So I'm playing along with the bias meter on, letting it get warmed up to readjust the bias. Turned down the VVR and it didn't drop the volume like it used to. Umm, strange...then playing along, Red Plate (on both tubes). Looked at the bias reading, it was .157. Shut it down (I didn't recheck the plate voltage & bias voltage when it was down...assumed, as designed, they track together).

So now, I turn it on with the VVR all the way up, bias in at .034, sounds great but after a few minutes I get this toxic smell happening (thought maybe something on the new tubes, but the smell get strong) and I hear a high pitched buzz in the amp (not from the speakers). Not even going near turning the VVR down at this point

Check all voltages, seem in range. My novice brain is tapped

Help?

Thx
Mark

[UR12]

Hey guys, posting over here, cuz I found more folks talking about the VVR...

So I built a Metro JTM45 a few years ago. Worked great, no problem. Loud as hell, year later, I installed a VVR (Dana) between the first cap and the choke and between the Bias trim and the 220K resistors. Worked great, no problem. Zoom forward about a year...

Loud hum, then back to normal. Switched it off. Back on and watched the tubes, spark shower storm inside one tube. Shut her down. K, Figured it was time for new tubes...

Got a pair of Gold Lion KT66s, installed. Going from suggestions on the Metro Amp board, tried to bias at .033 (65%) but I didn't have the needed range to go that low (got down to .039). So I change the bias resistor.

Biased her in at .034mV. Sounded great. Here is where it got weird on me.

So I'm playing along with the bias meter on, letting it get warmed up to readjust the bias. Turned down the VVR and it didn't drop the volume like it used to. Umm, strange...then playing along, Red Plate (on both tubes). Looked at the bias reading, it was .157. Shut it down (I didn't recheck the plate voltage & bias voltage when it was down...assumed, as designed, they track together).

So now, I turn it on with the VVR all the way up, bias in at .034, sounds great but after a few minutes I get this toxic smell happening (thought maybe something on the new tubes, but the smell get strong) and I hear a high pitched buzz in the amp (not from the speakers). Not even going near turning the VVR down at this point

Check all voltages, seem in range. My novice brain is tapped

Help?

Thx
Mark

It souunds like you have a blown mosfet and when you turn the VVR control you are reducing the bias but not the B+. Remove the power tubes and make sure both your B+ and Bias are tracking properly.

[yodermr]

Thanks Dana, that was it. Bias tracks but the B+ doesn't.

So as a follow up...Bad tube lead to killing the mosfet..OR..bad mosfet took out the tube?

Is there any change to consider to prevent this from happening again?


Thx
Mark

[stoo]

How are you heatsinking that mosfet? Should be bolted to the chassis with the appropriate insulators and thermal compound if using mica.
Stew

[UR12]

Thanks Dana, that was it. Bias tracks but the B+ doesn't.

So as a follow up...Bad tube lead to killing the mosfet..OR..bad mosfet took out the tube?

Is there any change to consider to prevent this from happening again?


Thx
Mark

Like Stoo said, I would guess that it was probably caused because the mosfet is getting too hot. With a 50 watt amp you need to add an additional heatsink in addition to bolting the mosfet to the chassis. I might have a heat sink laying aroundI can send ya if you want to PM me.

[yodermr]

Umm, Its bolted to the chassis, with the mica insulator - no thermal grease though.

Yes Dana, I will pm you.

Thx
M

[mcrracer]

Dana,can you list what an appropriate heatsink would be for a 50 watt installation? With part numbers if you could. Thanks

[dehughes]

Dana,can you list what an appropriate heatsink would be for a 50 watt installation? With part numbers if you could. Thanks

I too would like to know this.

As well, would one need to use thermal grease if you have one of the plastic insulators? I think not...but then I've yet to install the VVR on my AC30.

[Structo]

Isn't the tab on those MOSFETS hot with voltage?

So the mica insulator is keeping the B+ off of the chassis?

[soma_hero]

On a TO-220, usually the tab on the back is connected to the drain so it needs to be insulated.

On newer packages that are fully plastic with no tab, you don't need an insulator.

[UR12]

It's hard for me to recommend a particular heatsink because it depends on where the mosfet is located and where the heatsink will be added to the outside of the chassis. Some amps may require a larger one than others. The ones I have are about 5 inches x 2.5 inches and 1 inch tall. They can be cut down to a smaller size if needed to fit. I get them out of obsolete conputers.

[UR12]

Isn't the tab on those MOSFETS hot with voltage?

So the mica insulator is keeping the B+ off of the chassis?

Yes, the back of the mosfets are connected to the drain and that is why an insulator is required. The mosfets I use come with the insulator from the manufacturer. You can by mosfets that are called IsoPaks that you can bolt directly to the chassis but they are very expensive in the $25 range. I don't like using the thermal grease because some of it is conductive and some isn't. If you use the conductive kind you need to be carefull it doesn't provide a short to the chassis.

[FYL]

what an appropriate heatsink would be for a 50 watt installation?

As a rule of thumb, worst case dissipation will be 80% of max output power. You need to dissipate 40 watts max, the Mosfet shouldn't be allowed to go above 90° C or so (it's internal junction temperature shouldn't go above 150°), ambient temperature inside the amp is app. 40° C at best.

Calculating the required heatsink is quite easy : 50° / 40W = 1.25 C/W, quite a large beast, even larger if you take into account thermal resistance between the Mosfet, the insulator and the heatsink.

If the heatsink is thermally coupled to an aluminium chassis, you can use a 3 or maybe 4 C/W compact model, if the chassis is steel or if the heatsink isn't coupled, you need a large heatsink or you'll have to add a small fan (a 12V DC 60 or 80 mm ball-bearing pancake working on rectified 6.3V is fine, it'll work nicely and quite silently).

BTW, max dissipation occurs in a DC VRM (voltage regulator module) at mid settings, say between 10 and 2 o'clock on the VRM pot dial.

Oh, and please don't use mica insulators: there are less messy (no thermal grease needed) and much more effective solutions (better thermal coupling) using ceramic/silicon compounds or phase change materials - google Keratherm or Thermaphse for some examples.

[FYL]

You can by mosfets that are called IsoPaks that you can bolt directly to the chassis but they are very expensive in the $25 range.

IsoPak or TF series show quite poor thermal behavior because of the added epoxy-like layer with at least 1 C/W added to the thermal equation. For instance a Mosfet requiring a 3.55 C/W heatsink in standard TA package needs a much larger 2,55 C/W model when in TF form.

[yodermr]

Hey just letting everyone know...got it all up and running. Replaced the Mosfet, added a nice heat sink on it, biased it up and away we go.

Thanks for all the help
M