Chassis ventilation - Worth the effort?

[pdf64]

'The screen stoppers don't get hot until the tube goes bad.'
Try holding your fingertip on one with the amp cranked!

[soundguruman]

'The screen stoppers don't get hot until the tube goes bad.'
Try holding your fingertip on one with the amp cranked!

That's not hot, that's warm.

[Chuck H]

Just to add... As mentioned, you dont need (or want) to blow the fan right at the tubes. It's usually impractical to locate a fan in such a way that it would blow on all the tubes evenly so you end up cooling one tube A LOT more than the other/s. How does this affect the tubes? Not 100% sure. But it doesn't seem like a good idea. Something I've done before is to locate the fan wherever I have to, but have it blowing out rather than in. Then you can locate vents (or modify existing vents) to give more even air flow to each tube since air will be moving in through the vents and out through the fan.

[exclamationmark]

I don't really think the tubes themselves will benefit from any forced cooling, as most all the heat is radiated through the glass envelope. I've done some designs with a septar socket (using 6c33c tubes with several amperes of heater current), where I used a fan to blow under the socket to stop the actual socket from melting, but as far as plate dissipation goes it won't change that (unless you have some sort of metal base or ceramic tubes).

Removing heat from the air around the tubes using a fan is probably a good idea, but I just wonder how well it actually works. The vacuum in a tube essentially has a thermal coeffecient of nil, which means all the heat dissipated is either radiated from the plate, or cooled by convection through conduction to the base of the tube/glass envelope. Air flow will help with removing any heat transferred by convection or conduction, but will do nothing for the radiated heat. I don't actually know what the split is, in terms of dissipation for your average tube though. The point is, just keep all the heat away from your precious filter caps!

[J M Fahey]

Agree.
Tube refrigeration is 100% radiation, specially because the hot plates live in vacuum.
No other mechanism possible !!!
The forced
cooling benefits *the rest*of the parts, including the cabinet.

[jrfrond]

If you are building your Hiwatt clone "British-style" (i.e. upright tubes), then the tubes are above the electronic components and the heat rises up from the tubes and AWAY from the chassis. No cooling is really needed, and will do nothing more than suck dust into the cabinet, creating an insulating layer on the chassis and general dirt buildup.

Conversely, amps such as Fender, et al with inverted tubes have survived for years with no added cooling.

Unless it's a closed or cramped rack environment, stay away from any cooling schemes. They are not necessary.

[kg]

Removing heat from the air around the tubes using a fan is probably a good idea, but I just wonder how well it actually works.

this is dead easy to test.

put your finger on the tube envelope. observe level of pain. (Pnofan)

add forced air cooling, blowing over the envelope.

put your finger on the tube envelope. observe level of pain. (Pfan)

if Pnofan > Pfan, then fans = working well.

[soundguruman]

If you are building your Hiwatt clone "British-style" (i.e. upright tubes), then the tubes are above the electronic components and the heat rises up from the tubes and AWAY from the chassis. No cooling is really needed, and will do nothing more than suck dust into the cabinet, creating an insulating layer on the chassis and general dirt buildup.

Conversely, amps such as Fender, et al with inverted tubes have survived for years with no added cooling.

Unless it's a closed or cramped rack environment, stay away from any cooling schemes. They are not necessary.

Actually when you bias the amp HOT the fan can help quite a bit...and the amp sounds a lot better biased on the hot side.
When the amp is biased hot, the tube life is shortened, but you can extend the life somewhat with a fan cooler.
Depends, if you plan on using "consumer" grade bias setting, the amp is cooler and the tubes last longer.
But don't sound nearly as good.
Mesa Boogie is one example of an amp that is biased HOT and has a fan. But their fan is too loud.
The fan sounds more like a trash truck driving up the street when its turning.
So I think compromise, bias hot, slow quiet fan, tubes base down, and get some more life out of the tubes.
But I would not say that a fan is unnecessary, especially if you like to push it hard.

[soundguruman]

If your finger vaporizes, it might be too hot.
But I like the sound of EL34, when it screams for mercy.

[hasserl]

That HIWATT Chassis picture is an excellent example of how to do the venting.
It (as shown) does 2 things:
1st it allows convection air currents around the hot devices (tubes in particular) to keep them cooler.
2nd it restricts the paths (from the local hot devices) by which heat can be conducted to other parts (this is the action which is often forgotten).

Note also that heat shield (vertical plate) which blocks tube heat from the components on the other side (Electrolytic Filter Caps in particular) - another simple but excellent idea.

Reeves built stuff to MIL SPEC or at least his builds were heavily influenced by his previous experince of of MIL SPEC construction and wiring.
Probably one of the best examples to follow when planning layout and construction.

If you need extra cooling:
A rectified 6.3V AC Heater supply will give you around 8 Volts DC.
A 12V DC Fan will run slow and quiet on that supply but give sufficient flow to assist the convection air flow.
Accoustic Fan Noise is minimized AND fan efficiency is maximised by blowing into the chassis, never try to suck air out of the chassis.

Original Question:
Worth the effort? - Absolutely, unless you are a commercial builder who is only interested in the amp lasting 3 weeks longer than the warranty period.

Cheers,
Ian

Unless there are other vents to allow cooling air INTO the chassis (which I don't see), those slots around the tubes will not encourage convection. Also, I think the filter caps for that amp are INSIDE the chassis, not on the opposite side of shield, so it does nothing to protect the filters from heat, and if anything it will cause an increase in heat transfer to the chassis (and the components inside it) from the heat off the tubes that would normally be dissipated into the air around them.

[hasserl]

I'll step through what I'm thinking starting at the exhaust since that's where the "pull" is coming from.

1) The heat from the power tubes rises upwards and out the vents at the top and back of the head shell. Most of the heat will rise upwards, this will create a negative pressure on the vents around the base of the tubes. Note the upright metal between the power tubes and the preamp tubes. Many state this was done as a shield, but it will also act as a louver to direct air flow. This is important for later.
2) The negative pressure pulls air in from the inside of the chassis which creates a negative pressure on the holes at the front of the chassis
3) The negative pressure on the holes at the front of the chassis pulls air from the empty space inside the head shell, this creates a negative pressure on the air inside the head shell
4) The negative pressure on the volume of the head shell pulls air in from the vents, mainly the back vent because the top vent is expereincing positive pressure due to step one (air flow, like electricity, has to have a circuit) The metal shield that I said was important in the first step helps direct air around the transformers and into the front of the head shell space

I think the Hiwatt cooling system is more than just the vents, it's the entire layout of the chassis and components. If it isn't copied directly (or modified to be significantly better), just making holes isn't going to do you much good, though understanding how it works may help you with your fan layout to make it more effective.

So about your fan. Here's what I suggest, go to Newegg.com, look at fans there. Find one that has 8 bazillion reviews and great ratings, buy it. Yes, it will probably be plastic, so position it accordingly. I would advise getting the largest in diameter fan you can because it will usually have lower noise and flux (basically air pressure, you don't want a big thermal gradient from one side of the tube to the next assuming you'll be blowing these at the tubes sticking out of the chassis) for a given CFM rating. I seem to remember ball bearing being the prefered type, but a quick skim of a review site that has a "fan shootout" or similar will usually contain those sorts of generalities.

Maybe it's just my old eyes, but I'm not seeing any holes in the chassis to allow air in.

[hasserl]

If you are building your Hiwatt clone "British-style" (i.e. upright tubes), then the tubes are above the electronic components and the heat rises up from the tubes and AWAY from the chassis. No cooling is really needed, and will do nothing more than suck dust into the cabinet, creating an insulating layer on the chassis and general dirt buildup.

Conversely, amps such as Fender, et al with inverted tubes have survived for years with no added cooling.

Unless it's a closed or cramped rack environment, stay away from any cooling schemes. They are not necessary.

^^ This

[defaced]

Maybe it's just my old eyes, but I'm not seeing any holes in the chassis to allow air in.

[ATTACH=CONFIG]17579[/ATTACH]
[ATTACH=CONFIG]17580[/ATTACH]

Circled are the holes I was questioning if there is any obvious use for thinking these would be the rest of the circuit. From these pics, they don't look to have a use to me, but I've never had one of these amps in my hand:
http://hiwatt.org/pix.php?p=STA200gutsL
http://hiwatt.org/pix.php?p=STA200gutsR

Air comes in the back vent, around the transformers, through these holes, through the chassis, up the power tubes, out the top vent.
http://img37.imageshack.us/img37/8793/dr201front.jpg

[trobbins]

I don't really think the tubes themselves will benefit from any forced cooling, as most all the heat is radiated through the glass envelope

As kg has alluded to, a substantial % of the heat radiated from the anode is absorbed on its way through the glass. The glass temperature then becomes a part of the thermal 'mix', and big bottles definetely need to manage how well the glass itself is cooled. One observation to make to understand this is to touch the very top of a KT88, versus on the sidewall (which gets the main blast of heat). Normally, the glass cools mainly from convection in to the air, but a little also radiates away - and effectively none would conduct (because it is very thin). ANother observation is to look at those maga amps with 8-10 output valves all sandwiched together with a fan blowing directly over them.

[Chuck H]

+1
Taken in the other direction for observational purposes... Imagine operating your tubes in an oven. I imagine they would suffer. Cramped conditions and not enough venting has been a well known problem in amps like early AC100's. Point is, if the temperature outside of the tubes glass envelope is moot then why do we bother venting at all? More heat can move into a cooler environment than into a hot environment. If NO heat escapes a tube then there would be no need to power the filaments more than once!?! So obviously heat is evacuating the tube into it's surrounding environment. Therefor evacuating heat from that environment allows the tubes to run cooler since all things are trying to equalize. I don't know much about physics but I do know that if you put a hot pan on a cold surface the two objects will try to reach a state of equalibrium. Tubes sit in air. So we cool the air and the tubes try to reach equalibrium to best of airs heat conductive limitations.

As for whether the tubes themselves are in any danger running in a hot environment I guess I can't say. But there are other things in an amp that certainly shouldn't get too hot. Resistors derate, capacitors derate, wood catches fire, etc. I'm sure there is a temperature at which tubes also derate in one way or another. Do tubes running at hot amplifier temperatures with poor venting ever reach that temperature? Not at all sure. But I plan to keep venting my amps and if I ever design one that has cramped quarters and gets really hot I'll probably add a fan to the design.