Fan speed control by OP of amp.

[Craig Sawyers]

Back in 1983, in my first job, I designed a video processor. This was to be mounted in a helicopter, with rather a large temperature range. So I used a thermistor, an op-amp controller and a power stage to run the fan.

It did not operate until a preset temperature, then after a kick in transient to get it going the fan ran linearly with temperature until it reached it maximum operating voltage.

All back in the days of the lm741 and similar stuff.

Craig

[Diabolical Artificer]

" If they're both under 100C the TX should be happy even if your fingers ain't! " That's reassuring , but as you say some figures are needed, so I'll look for a thermo device.

Thanks, that's most kind Rob,however I have shed loads of fans being an inveterate scapper of old kit and hoarder of parts.

I think Craig that the amp needs some thermal tinkering rather than a fan controller.

Andy.

[Craig Sawyers]

At 100C is this a class A amp?

Craig

[vidjoman]

Time we had some details of how big the amps are and also the space around them. Photos would help to advise of what to do. As they are valve amps then you must expect them to get hot but in a suitable space air currents can keep them at a reasonable temperature without the need of cooling fans.

[Diabolical Artificer]

They're nowhere near 100 deg C Craig.

Having had a think I'm thinking about adding another fan, see pics. There are two issues here - 1) the top of the chassis on top of where the valves and OPT sit was getting too hot, uncomfortable to the touch. A 4" fan sits underneath this blowing air up and out through holes drilled in the chassis, this was with the fan at reduced revs. With the holes drilled out to a bigger dia, and fan on full revs, this is a lot better. However, noise is an issue, hence original Q.

2) The mains tfmr inside the PSU section was getting hot too, not overly so, IE nowhere near 100 deg C, but I would rather it ran cooler. So I've made a partition, to separate the mains tfmr and PCB's etc. The right side will blow air up and out through chassis as previously described, the LHS will suck hot air from the mains tfmr out, underneath the amp, that's the plan anyway.

I've also found two 11v SMPSU's to run the fans, as previously the 4" was powered from the 12.6v DC heater supply, this was pushing the supply a bit, thus this will put less "strain" on the heater supply. Space is tight as vcan be seen from pics.

Andy.

PS,sorry, mods, subject of thread wandered a bit, thread title may need amending.

[MotorBikeLes]

Ooh, what a lovely insulating box it lives in. What a nice, shiny panel holding the valves.With that thick walled box, you will need more, bigger vent holes than if it were sheet metal. Dull finished metal MAY help with greater radiation losses, though often not as much as "gut feelings" would suggest.
A fan blowing IN cold air will shift more air mass (hence more heat) than a fan drawing out hot air. (remember your gas laws).
Time for some thinking.
Les.

[vidjoman]

You say that you have 2 monoblocks - the picture shows 1 of them with 6 output valves that look like EL34's. Each takes 1.5 amps heater current at 6.3 volts so that's over 9 amps and a lot of heat to get rid of. If I were to put just one of those in a cabinet with generous ventilation holes in the base and an open back and rely on natural convection I think I would go for a space above the top of the valves of about 15 - 18 inches and a space area of around 3 - 4 cubic feet. That should allow an upward draft to keep things fairly cool. Remembering that valve amps run hot anyway.

[Craig Sawyers]

For the Tx, the allowable temperature depends on the insulation grade on the winding wire. For a decent transformer internal temperature might be 100C with a surface temperature of 60C. Which is too hot to touch for long at all. But is is all down to how well the case containing the transformer is ventilated to allow convection cooling.

I spy HP and Tek test gear....nice!

Craig

Craig

[Diabolical Artificer]

Yes, far from ideal Les and some forethought re heat before building the thing would have been better. I guess I could paint the top chassis, something to bear in mind. "A fan blowing IN cold air will shift more air mass (hence more heat) than a fan drawing out hot air. (remember your gas laws)." Didn't know that, thermal design is something I've never looked into.

Yes, much power dissapation, mucho heat, Vidjoman.

Wire used has the following specs - 200 Deg Celsius - Sizes: 0.112mm To 3.00mm - IEC 60317-13 / NEMA MW 35/36 180 Deg Celsius - Sizes: 0.030mm To 0.1mm - IEC 60317-51 / NEMA MW82 so should be ok there. Insulation between windings is done with Kapton tape, so another plus. Yep the yanks did know how to build test gear Craig.

I will mount the 4" fan on grommets (ordered some) that'll cut down noise and have a play, see if I can get rid of some heat off the tfmr which is being blown onto the chassis, we might be ok. As I said the chassis isn't a burn hazard, I just need to reduce noise and heat a tad.

Thanks all, Andy.

[mhennessy]

I've said it before, and I'll say it again: they are lovely looking amplifiers

At a glance, it does look like there's not enough places for the air to escape from underneath the amplifier top-plate. As an experiment, I wonder if one end could be lifted, just to create a temporary outlet? The opposite end to the connector, of course. That might help to prove that point. Also, more outlets should result in lower noise (as the air velocity through each outlet is lower). Obviously, a grille is recommended for safety if the holes get above a certain size.

Although it's a worthwhile experiment, I personally wouldn't use the switched-mode power supplies to run the fans - they might cause interference and other problems. A fan takes hardly any current really. Certainly not in the context of all those heaters.

[julie_m]

When a gas is hot, the molecules are spaced further apart and moving about more than when it is cold. This means that one litre of air at 20°C will have more molecules in it than a litre of air at 60°C. Although, not by a factor of 3, since gas temperatures have to be considered from Absolute Zero, which is -273°C; rather, there will be 333 molecules at 20°C for every 293 molecules at 60°C. That's still 15% more effective, though. And the number of molecules is important, since molecular collisions are the mechanism how heat is transported by a gas. Each turn of the fan pushes the same volume of air through; but having the fan acting on the intake with cold air will mean more molecules of air are present in each litre, and therefore the cooling effect will be greater.

The stream of cooling air will expand once it heats up, so it is important not to limit the rate of escape of the heated air -- there will be more of it coming out than went in! But you don't necessarily want to start moving it away too fast, or it might never get the chance to pick up all the excess heat available. Warm air will experience a buoyant effect anyway, tending to rise up above the denser, cooler air below it. Sometimes this can actually be seen with the naked eye: the refractive index of air depends on temperature, and sometimes images sen through the space around a hot object can be distorted. (In really extreme cases, the refractivity can be such as to produce an image of the sky at ground level in the distance; which is scarily easy to mistake for a reflection of the sky in a distant body of water. A mirage. Usually depicted rather better-appointed in cartoons than in real life .....)

My gut feeling says blow plenty of cold (perhaps even pre-chilled with Peltier devices?) air in at the bottom; then leave it to waft away at its own pace, with perhaps a gentle cross-draught above the set-up. You're not trying to go for maximum efficiency, which would call for setting up an exact opposite temperature gradient across the heat source; but rather to give Nature the gentlest possible nudge that will make a difference; to preserve your nudging finger, as it were.

[MotorBikeLes]

Andy and Julie, I also forgot to mention that hot air has higher viscosity than cold, so the gain is likely to be greater than simple gas law prediction. Now of course if it were a positive displacement pump, not a simple fan, the only gas laws would apply.
The resistance of a simple "hole in wood" is much greater than a simple orifice in sheet.
For practical purposes, the resistance is inversely proportional to the area for a simple orifice, but gets more complicated with a "length", where the surface roughness has an effect, and strange "rules" come into play. A full understanding is way beyond me, I just know enough to know I don't know.
Les.

[Diabolical Artificer]

"I've said it before, and I'll say it again: they are lovely looking amplifiers" Bless you, thanks. "I wonder if one end could be lifted" that's an idea Mark. Got you re SMPSU, I fitted a Micronell fan to suck air from around the tfmr yesterday, I'm sure I heard it ringing a bit, unfortunately space is tight and even though both fans need 200mA ish, don't want to tax the 12v DC supply too much.

Thanks for the explanation Julie and Les.

As mentioned I fitted a fan yesterday and put some rubber strips in to dampen noise, will experiment today and report back with the results.

Andy.