How hot is too hot?

[Bob M.]

I'm rebuilding an old home brew that served me well for 20 some years plus because it's time for an update and refurbish. When I built this amp in the late 70s, initially upon completion I had alot of problems with the amp (and chassis) heating up and blowing fuses and not being heat-stable. The fix at that time was to add alot of extra heat sinking to the chassis and the amp became stable and worked well for a very long time. I think I even added a fan later at one point but decided to eliminate it now because of the massive dust buildup fans cause in and about the chassis.

I'm rethinking things now (this was an inelegant solution, I think) and it appears there is no real reason that I need such over-the-top heatsinking. The chassis seems of sufficient mass to dissipate the heat buildup from the compliment of components I've chosen. So I've eliminated this extra heatsinking and taken great steps to insure good chassis mounting and grounding of the power transformer, which was the original heat-generating culprit. The amp seems fine now but it does get hotter than before but it's not blowing fuses or anything like that. My question now becomes how hot is too hot for a power transformer? We all know anecdotal measurements (if you touch it and you can't keep your fingers there more than 3 or 5 seconds, then it's too hot) but I'm wondering if there's any hard scientific data on transformers and temperature; what's safe and when does a transformer enter the breakdown area because of excessive heat?

The chassis is similar in size to a Deluxe Reverb amp and the power transformer is a Vibrolux Reverb transformer. I'm driving 2x 6L6s into 8 ohms. The amp is fantastic but I'd like to get on top of this heat situation before I consider the update completed. Any comments and expertise are appreciated.

Thanks in adv,

Bob M.

[Jazz P Bass]

Too hot is when you smell burning varnish.
I think that you will need the transformer manufacturers data to answer this one.
Here is a link that may help explain why.
It has to do mainly with the insulation class used on the windings.
Link: The basics of transformers, part 2.

[MWJB]

How hot are you running the power tube plates? a s long as there isn't an excessive draw on the heaters, this is where most of your heat is generated.

[km6xz]

Do you know the specs of the transformer? Is it a common model transformer where you can look up the data sheet for? If it is a true Vibrolux Reverb its only rated at 35 watts. Reissues are claiming more. There is lower plate voltage in the original than other Fender 6L6GC amps of the era. If the amp is running well, as you say, and not pulling too much current focus on the transformer itself.
Sounds like excess losses, what temperature rise do you get when set with power on but in standby. Something is not right. Does it stay cool if the tubes are pulled but on and out of standby?
You can measure the AC mains current, and each secondary to calculate efficiency during operation? If you have a AC clamp-on current meter available no wires need be lifted to make the measurement.

[überfuzz]

If I were in your shoes I'd start with some probing to see if the tubes seems to be running properly. What voltages do you get? Let's say do find them to run fine and dandy and the amp is still getting seriously hot. (I find scientific finger probing to be a excellent way of establish if the amps is having a fever.) Why not a fan? Mount in so that it's blowing at the power tubes pointing down the line of tubes. I've seen this kind of fan mod in fender amps with nice results.

[melvin]

Bob M., I am curious as to what you've got in that home-brew of yours. I'd like to know what is working that power transformer so hard. Regardless, why not upgrade you PT? If that's what's generating all the heat, I would find one that can handle a little more current than necessary on the HT and just right on the filament winding. It will run cooler and give you more options. An "upgrade" Vibrolux PT usually includes increased current handling...

[kg]

bob, generally the temperature of interest is 140*f. above that the commonly used enamel/varnish on magnet wire starts to break down.

ALL heat, however, will lead to premature death. it's just that above that point the rate increases rapidly.

hth
ken

[Bob M.]

Thanks for all your responses. The one fact that seems true is that the insulation rating temp seems first and foremost. I'm not saying the amp is in the danger zone, it's does get somewhat hot. As it's not a manufactured amp, my experience is that these thing vary alot. Of course, it's bound to run a bit hotter after I removed the fan which keeps things forced-cool. Also, we've all seen alot of amps and not all amps run at exactly the same temperature. This amp runs no hotter than most Twin Reverbs I've seen.

Just to answer a few questions: I'm running the exact tube compliment of a Vibrolux Reverb but I'm using a single 8 ohm speaker and have mounted the appropriate output transformer. The power transformer is an early/mid 70s Fender model: L022723. The power transformer draws no current what-so-ever without load. The heater circuit appears to be working normally and is drawing the expected amount of current with a typical AC volts reading. I'm not using the 5 volt winding as I have a solid-state rectifier fitted. In standby mode, the amp remains cool. It usually take about 40 minutes for the power transformer to reach its maximum temperature in play mode. There is zero chassis/transformer vibration. The reverb driver tube, a 12AT7, and the PI driver, another 12AT7, run hotter than the other dual-triodes, as per normal for the circuit I'm using. Both of these tubes exhibit typical voltages. The bias on my power tubes is in the middle range, neither too hot or too cold. The amp is working as expected in all areas. It's just gets a little hot and I'm asking, not so much from a repair standpoint but from a more theoretical one: How hot is too hot? Of course, I can always add a fan or fit a different transformer at a later date. Although the transformer gets hot, the heat transference to the chassis is minimal. Thanks for your replies.

[überfuzz]

So there are no suspicious looking voltages. If yes, then I guess it's a matter of gut feeling.

I just wanted to comment on one thing. A lot of things are kept "forced-cool". In your amp things drawing much or to much current will get hot. Take your computer for instance. Without cooling it would go south in a mater of minutes. If the cooling system is beefed up in a comp you can make it run even faster.

Eventually, a word of cation: Never play this song if your amp doesn't have a proper cooling system. YouTube - Buster Poindexter - Hot Hot Hot

[birdyman]

Can you check bias current?
6l6 should dissipate 25w max so. If your b+ is 450v. Then current should be 50ma or less to stay within spec.

[MWJB]

Not in fixed bias, dissipations of 12-17W are more like it, plate current need not typically be over 35mA...could be lower to reduce heat, just note where tone thins out & bump up a few mA.