OCD about amp temperatures

OK this is something I never thought to much about. Recently I decided to get a tube power amp. I posted about it and ordered fans and when I tested the Rogue ST100 I confirmed in my setup it needs a fan. For years I have had Bryston 4b &3b amps and they run hot. So I ordered 3 fans and thought I would use dimmers because I was thinking the 110 CFM fans cooled the amps down to much.

So my question to you guys is, being that all my equipment is in a separate room and fan noise is not an issue, is cooler better? The surface temperature of the ST100 was 122 deg in 20-30 minutes but with the fan it dropped to to 84 deg. Is that too cool? Or is optimal amp temperatures a bit higher?

 

Power tubes are supposed to get hot. Why are you trying to cool them down?

 

His setup, mentioned, is in a confined rack. It needs some sort of cooling because convection alone won't keep it from overheating.

Partly to do less cooking of caps and transformer windings, I use small fans to improve airflow myself. There is also less heat affecting other gear above my amp.

 

Most people like myself for years never think or have concerns about how hot their equipment gets.

The Bryston stuff always gets hot. I remember when I bought the 4B some 25 years ago the salesman said it's supposed to get hot. He believed that the cooling down and heating up daily caused more harm. So he left his Bryston amps all the time. I never bought into all that plus it's not the green thing to do.

Now the rep at Rogue got me thinking. He said 84 is good and heat is the enemy of electronics. I thought a little warmer might sound better.

 

I think folks tend to associate "warm up" to temperature. In reality a warm up is to allow the circuit to stabilize. Capacitors and inductors need to saturate and tubes, in particular need to reach their operating range for their magic bits to operate efficiently. Proper equipment design should allow tubes to operate at the required level without overheating the rest of the electronics. Heat is absolutely the enemy to electronics. Generating heat is fine, but it is imperative that it be removed effectively.

 

Just a data point, the data sheet for KT120 shows an envelope temperature of 250C. Just under 500F.

http://www.tungsol.com/tungsol/specs/kt120-tung-sol-specs-curves.pdf

You might use a non-contact laser thermometer and check the tubes themselves. I got one, $<20, just to investigate. I'm not that OCD, but you could use a thermometer to ensure even cooling in case it might affect the sound. Might be even more fun than arguing about cables..."what's the best temp for listening to Pink Floyd", maybe?

 

That's a design maximum rating, the typical operation point is half of that. Further, the rating is for the tube envelope only, which is the actual top of the glass bottle, not for any surrounding area. I am sure that you know this, but for the benefit of others reading, they should know that components near the tubes are not designed to handle temperatures near that range. Even at 125 degrees C, the actual tube envelope is over twice the temp rating of most associated capacitors (105 degrees C). Many capacitors are only rated at 85 degree C, so care needs to be taken to allow for proper ventilation and thus cooling. Fan forced cooling is not a bad thing.

I doubt that anyone could tell the difference sonically when listening to an amp that was statically dissipating at 100 degrees above its cage and one that was fan cooled and only 80 degrees above its cage. If I were a capacitor, a transistor, a diode, or even a resistor sitting beside one of those tubes, I know which amp I'd rather be in though!

-Bill

 

Yes, many of my tubes run ~300F or so, depends on the application. I was being facetious about sound quality vs temp. I cool to keep the caps happy mostly.

Good sockets and fully-seated tubes help keep things cooler in my experience and good convection cooling should be more than an afterthought for any designer.

 

This unrestored deuce has been running for over six years.
I'm pretty sure keeping the caps and other passives cooler has prolonged its life.

 

They are not supposed to get hot, the heat is a by-product of their use. A perfect tube would utilize all energy coming in and not release any heat but that would be physically impossible.

I'd speculate that the closer a tube component is in temperature to equilibrium the longer it will last and quite possibly sound 'better'.

 

So the consensus seems to be cooler is better as far as component life span. And the warm up period is more about time than temperature.

The other component I am concerned about is my Marantz HT preamp I think is the least reliable component I have. I had to move it to make room for the ST100 and now the shelf it sits on I only have 5/8 of an inch of air space. So now I have a fan between it and a Adcom 2535 a 4ch amp for my high channels. Now both stay very cool just a little warm to the touch.

When I get my ST100 I will listen for myself to see if it sounds any better cool or warm.

 

I run a Bryston and yes they get hot. After talking to Bryston I confirmed that the 12v trigger is robust enough to power a pair of 12v "computer case fans". I selected a pair of lower amp. ANTEC 6" fans and now have then rigged to come on when the system powers up. They blow across the heat sinks on either side. So my suggestion is not to use hi-voltage fans with dimmers or fan controls. You can power low voltage vans from a wall wart as well (done this too). You do not need large air volumes. The low voltage fans are quiet and low vibration. Their duty cycles are very long.

 

Nope. They are supposed to be quite hot, with few exceptions. The cathode gives up Barium and it is redeposited, in many tube designs, all requiring heat cycling. The anode of many designs outgasses significantly, the KT88 is an example. The getter absorbs when cold but it will also collect many of the outgas byproducts best when hot. Tubes have to be hot. The question for us is how to allow that, mostly, without cooking the other components in the amp or even on the shelves above it.

 

I inquired about using my my McIntosh 12v trigger output to power fans and it was shot down I can't remember if it was a McIntosh rep or just forum members but they said it was not designed for constant power draws. The Bryston may be different.

 

Thanks Bob, learn something new every day!

 

So my question is what is a good surface temperature range? I'm thinking and this purely a guess because I think it should be warm but not uncomfortably hot when you lay your hand on it. For a tube amp 95 deg f to 105 deg f for SS preamps less.

 

My opinion is that you should be able to touch any exposed part of the device, other than the tubes themselves, and keep your hand there without being burned. Some transformers can get VERY warm, for instance. With heat transfer, that might translate to as much as ~135F or so measured on chassis, panel, or cage areas. Just from a reliability standpoint, anything hotter could generate a McDonald's coffee-style lawsuit, even if no one cared about component longevity. So the ranges you mention are probably fairly typical and common. My tube amp runs about 125F on top when all closed up as it was delivered, but there is a top panel I leave off of it since it is in shelving, with small fans.

I mentioned a non-contact thermometer I use, it's really handy. It was only $14 on a special manager sale at HarborFreight. I imagine they are cheap on ePray and elsewhere, if you shop around.

 

Sometimes heat can be a diode in the PS board, installed too close to the board with not enough clearance to allow some cooling, a fan will help or the auto shut off could engage. Replacing the diode and leaving the legs a bit longer helps too...

 

I should mention here that there are articles on the web, if you look around, that insist on little or no cooling being best. This is due to the cathode and anode heat cycling I mentioned above and maybe the source of F1nut's comment after the OP. I can't dispute that I may be doing the wrong thing by adding extra fan cooling, but I only add a little with my 1/2-speed DC fans--wired in series. If I could leave the chassis open top and not in a confined area, I would rely on convection only. Too many little, fast hands around, though.

 

I am using the ps that came with the BP-25. From my conversation with Bryston they told me first that the trigger is not a trigger voltage but a constant 12v applied to the circuit control of the line voltage. Some triggers are not that way. When activated they apply their voltage to perhaps close a circuit and then when activated again to open the circuit. In the Bryston design it seems to be a constant voltage that acts to keep a "relay" closed and when remove the relay opens breaking the circuit.My concern was that the ps is not sufficient to support the additional load of the fans or that the added draw would cause overheating of the ps. When I supplied Bryston with the specs they said no problem. I have had no issues to date. The amp sits in a space open front to back but closed sides and top. The gentle airflow is sufficient to eliminate heat buildup.

 

Thanks for the Harbor Freight tip. I drive by one when I pick up my ST100 so I'll be getting one.

 

Be sure to look for a coupon. The usual coupon is $24.99 I think, still not a bad deal.

 

I've got a brand new Bryston 4B3 and it barely gets warm to the touch, my old solid state Conrad Johnson got much hotter in comparison running the same speakers.

 

While a whole lot of cooling may not be necessary, cooling will help to preserve the life of passive components, if not the tubes themselves. If you look at the specification sheets for resistors and capacitors, for example, they have different expected useable life ratings that vary based on the operating temperature--the higher the temperature, the lower the life. These ratings cover a pretty broad spectrum of temperatures, so it is not just extreme heat that shortens life. For your own peace of mind, forced air cooling wouldn't be a bad idea. As to tube operating points, I have no idea if there is such a thing as excessive cooling from just moving a lot of room temperature air over the tubes, but I suspect that that would never be the case. With tubes operating completely in the open, air convection would be constantly moving cool air over the tubes and I don't know of anyone saying that such cooling hurts performance.

 

What does Rogue saw?