Does a tube amp cooling fan make a difference? Yes! (quick study results)

[agedhorse]

it is unavoidable.

 

[byacey]

I don't think the output tubes care much about the ambient temperature. What does benfit from a.cooling fan is the rest of the components inside the chassis, which becomes.an easy bake oven, especially when the tubes are mounted on the underside.

 

[Wasnex]

You should look up the various threads on the SVT for its fan direction, should it blow frontwards out of the amp, or intake and blow backwards towards the tube. It's highly debated.

I have one of each. I actually took the time to reverse the fan on the 87. Had to put it back because it was a lot noisier. My theory: The 69 has a cut out in the grill for the fan, the 87 does not. The air passing through the little holes in the grill form into high speed jets. When the jets hit the fan blades, they produce a rather obnoxious whirling sound.

 

[Oldfartguy]

I have several mass produced tube bass amps that disregard this advice.

There are a lot of reputable enginering information and data can prove the hot spot point on valves. There are several reasons why fan cooling on valves should be carefull design to not short life valves. It's an old trade now...

 

[agedhorse]

There are a lot of reputable enginering information and data can prove the hot spot point on valves. There are several reasons why fan cooling on valves should be carefull design to not short life valves. It's an old trade now...

IMO and IME, this is a perpetuation of urban myths (within reason) for most audio power tuner that I can think of.

 

[Stylee Mousse]

The PO of my vintage V-4B built a new back panel with an ac fan built in. It plugs into the ac accessory plug conveniently located.

I have a wall wart supply powering a cpu fan Velcro-ed to the back of my Traynor yba1a mkII and it also has worked well for my 20 year guitar playing (gigging) life.

My other stage guitar amp is an Ampeg SB12 and it seems to not need a fan. Poss because the tubes are in free (or caged) air.

 

[Miles_ONeal]

So why did so many amps put the tubes upside down and allow the heat to gently broil delicate electronic parts?

Early tube gear all had the chassis on the bottom of the cab with tubes on top of tge chassis. This worked great for tv, radios, etc. that were essentially furniture and had closed backs. But instrument amps mostly had open backs. Musicians wanted to toss stuff inside — even if it was only the amp's power cord — and tubes got knocked askew or broken. They moved the electronics to the top (upside down) for practicality. Most designs were engineered conservatively for clean power without stressing the tubes much, so they didn't get all that hot. I say most because there were exceptions. Look up VOX AC30 fires.
This practice continued when heads were separated out of combos because it let manufacturers just have one chassis build. They could have flipped things but then controls would have been backwards on one or the other.
Ampeg's Portaflex combos were probably the most obvious exceptions.

 

[runmikeyrun]

Except that the cool air enters on the other side of the tube, either direction causes one side to be warmer or cooler than the other side.

I think there's an (albeit slight) argument to be made that blowing a fan directly at one side of the tubes has a more dramatic cooling effect than air being pulled through a the wide opening in the back of the amp.

Bernoulli's principle states that the same volume of air moving through a small opening has a higher velocity than a large one. So inherently a wide opening will have a slower stream of air going past the tubes, which should cool one side of the tube less than a fan directed towards it.

BUT… a slower stream of air has more time to pick up heat. If you run your car without a thermostat it can actually overheat because the coolant is rushing past the heads too quickly to absorb a lot of heat, and too fast through the radiator to dissipate it. (I'm also a gear head who needs to think about how to cool engines and get air properly to the carburetors). So if you're blasting the tubes with a fast stream of air, is it losing efficiency because the air is moving so fast?

Variables affecting both of my points would be the velocity and volume of air along with the distance a fan would be from the tubes.

I have a healthcare degree that is nowhere close to an engineering degree (save for the "science" designation in both) so at this point I'll defer any further explanation from where I left off to someone who has a degree/advanced knowledge involving fluid dynamics and thermodynamics.

Didn't realize this post would become so complicated LOL

 

[Wasnex]

I think there's an (albeit slight) argument to be made that blowing a fan directly at one side of the tubes has a more dramatic cooling effect than air being pulled through a the wide opening in the back of the amp.

True however, in some amps the amount of cooling on a given tube is going to vary considerably more than you describe

This is a Mesa 400+

Fan on the right blows air across 12 6L6GC tubes.

The fan in the 400+ has three settings, off, low-speed, high-speed.

The back, side, and top are vented. If you run your hand across the top panel, it get's progressively hotter as your hand gets further from the fan.

The Mesa 400 and D180 use the same air path, but there are six tubes in one line.

The fan in the Aguilar DB359 uses the same air path as well, but there are only four tubes in one line.

My impression is the variance across the tubes would be far more significant if the fan sucked air out. The output air flow of a fan tends to be more directional than the input air flow.

Additionally, pulling hot air through a plastic fan doesn't sound like a good idea.

 

[agedhorse]

True however, in some amps the amount of cooling on a given tube is going to vary considerably more than you describe

This is a Mesa 400+
View attachment 7299928
Fan on the right blows air across 12 6L6GC tubes.

The fan in the 400+ has three settings, off, low-speed, high-speed.
View attachment 7299930

The back, side, and top are vented. If you run your hand across the top panel, it get's progressively hotter as your hand gets further from the fan.

The Mesa 400 and D180 use the same air path, but there are six tubes in one line.

The fan in the Aguilar DB359 uses the same air path as well, but there are only four tubes in one line.

My impression is the variance across the tubes would be far more significant if the fan sucked air out. The output air flow of a fan tends to be more directional than the input air flow.

Additionally, pulling hot air through a plastic fan doesn't sound like a good idea.

As a practical matter, there will always be a temperature difference between tubes, it's not a huge deal. The more important thing that a fan brings is a significantly LOWER overall temperature across all tubes. It also lowers the chassis temperature which lowers the internal temperature of all components.

The Bass 400+ has been very reliable over a long service life, and not particularly hard on tubes either. Most players on tour use the higher fan speed, off is really for studio applications.

 

[runmikeyrun]

Decided to do the same test on my ampeg V4, which runs considerably hotter.

10min warm up time with fan on, amp off standby, at idle.

Shut fan the fan off, let it cook five mins. Tube base temp: 229 F.

Turned fan back on, let it cook another 5 mins. Tube base temp: 201 F.

This is a much smaller difference in cooling vs the Mace, however there are a number of factors at play.

The Ampeg has 7 additional preamp tubes heating things up. The Ampeg has new Tung Sol 7581a which presumably run a little hotter than the 50 year old GE 6L6GC in the Mace. The front panel of the Ampeg is an open slot so the exhaust fan inevitably draws a small amount of air in from the sides of the front panel and not the back panel.

I can try blocking the front panel slots on the Ampeg to see if there is any cooling benefit.

 

[agedhorse]

Do the same test for several hours, 10-15 min is absurdly short.

 

[agedhorse]

Fir some real data based on a 4 hour test from safety certification testing (guitar amp)...

Power transformer core: 129F, 167F
Output transformer core: 127F, 158F
Back panel touch temp: 123F, 147F
tube guard touch temp: 116F, 131F

The first number is with the fan, the second number is without a fan.

 

[6thplanet]

Is that with the amp at idle or under full rated output?

 

[agedhorse]

Is that with the amp at idle or under full rated output?

That's at 1/8-rated power, the standard value used for normal operation regulatory testing.

 

[6thplanet]

And when you state "core" temp, how is this measured?

 

[Jeff Scott]

And when you state "core" temp, how is this measured?

 

[Wasnex]

That's at 1/8-rated power, the standard value used for normal operation regulatory testing.

My understanding is live bass has about a 12dB crest factor. If I understand, 1/8th power with a 12dB crest factor puts transients of the signal right at the amp's power rating.

AFAIK 1/8 of rated power is -9dB down. Convert the RMS power to peak power and you get +3dB. Put these two together and you get a 12dB crest factor.

For a 100W amp:

• RMS Power = 12.5W RMS long term (-9dB)
• Program Power = 100W RMS short term (0dB)
• Peak Power = 200Wpeak(+3dB)

 

[agedhorse]

My understanding is live bass has about a 12dB crest factor. If I understand, 1/8th power with a 12dB crest factor puts transients of the signal right at the amp's power rating.

AFAIK 1/8 of rated power is -9dB down. Convert the RMS power to peak power and you get +3dB. Put these two together and you get a 12dB crest factor.

For a 100W amp:

• RMS Power = 12.5W RMS long term (-9dB)
• Program Power = 100W RMS short term (0dB)
• Peak Power = 200Wpeak(+3dB)

Yes and this is the value specified under IEC 60065 and 62368.

 

[agedhorse]

And when you state "core" temp, how is this measured?

Thermocouple. When an amp come back from UL, it's covered with superglued thermocouples!