Pictures of your diy Pass amplifier

Most of the better quality caps have text printed on them detailing the temperature allowance. Usually it is ~85 degrees C if you didn't specify. 105 degree caps are what you should be getting, but they usually cost considerably more. Still, this rating refers to the caps ability to stay in tolerance for a specific amount of time at that temperature. I have been repairing a lot of LCD panels (monitors, TVs etc) that simply have 'bad' caps in them. It is not the value of the temperature that killed them, but the extended time at any elevated temperature that kills them. You guys should always consider drilling vent holes in the base plate and making sure the heat can escape somehow from within the inner sanctum of the amp (IMHO).
 
Thanks Mixon1, evanc - and Thanks EUVL for pointing the temp of the 'gnd plate' - I have checked it this morning, run the amp with lid closed and after 2hrs checked the temp of the ' power resistor' and gnd plate - and yes, its quite warm; close to the temp of the heat sink!! - I'm sure it's not the ambient temp. inside the chassis that's causing it to get warm; very bad for those caps. Because of the limited space I decided to put it there thinking the power resistor ( 50w printed on the body ) would handle the heat pretty easy, and the gnd plate would not be affected by heat coming from the resistors. The resistor in parallel ( 0.5 ohms 100w ) has a voltage drop of 1v and would pass, say 1.5 amps. P=IV - 1.5 W ?? Am I missing something ? I might isolate the power resistors from the gnd plate - will post pix once completed. - By the way the Amp sound really good!! one word to describe it is " SPECIAL "

some tech details:
600 VA x'former
measured supply rail: + -21.68v
Toshiba Output
Bias: Stabilized to 0.6v after 30 minutes
DC offset: 0-2mV
Cap bank: 68000uf x 4
 
Reytnz,
I added a small fan salvaged from a dead computer blowing out the back of the amp case. The 12 volt fan running at 5 volts is silent and now the inside of the case is the same temp as the outside. The heatsinks for the rectifiers and power resistors still get hot, but now I'm not worried about cooking the caps.
 
Member
Joined 2004
Paid Member
reytnz: That's a tastefully neat and compact F5 you've made!

I don't think you're missing anything in your power calculation of the resistors. "Quite warm" is not "hot," so you should be OK as far as the caps go (unless you're drawing power much beyond the 1.3A class A 'limit'). And it appears you have good ventilation in the top cover. But perhaps more or larger vent holes in the bottom might be a future option if you're still concerned?
 
Last edited:
I might consider putting a fan; if I find a space in the chassis or drilling a ventilation holes on the chassis just above the capacitor banks, what puzzles me is the whole power resistors 4 in total , 1 ohm 50w each , 2 pcs. in parallel for each +- supply rail. drawing 1.5 watts each approx, 6w total for the 4 resistors can generate so much heat. or I may be wrong and the heat is probably coming from the convection effect inside the chassis. I can touch the resistor for 5-10 sec after 2 hour of operation. I am concern that the capacitor connected to the gnd plate would cook. This is really a "HOT amplifier"
 
Rey,
to get an idea of what 6W feels like, set up a plate with one power transistor attached.
Wire it up to dissipate 6W (500mA @ 12Vce).
monitor temperature over a couple of hours. Now imagine that plate inside a chassis where Ta is considerably higher and protected from the draught, "close that door".
 
Add up all the dissipations of all the components making up your amplifier.
That should exactly equal the Power Input from the Mains.

Subtract the heat load dissipated by external heatsinks.

Everything else is dissipated inside the chassis and has to escape via ventilation or conduction through the chassis.
It does not leak out via the connectors (too much thermal resistance).
 
Let me perhaps explain why I spotted your problem straight away.
And what Andrew said is true is any case.

The power dissipated by the MOSFETs are taken care of by the large heatsinks. That's fine.
For the CRC, you calculated 1.5W per resistor, or 6W in total. That is also quite OK.
Had you used normal metal film power resistors, or ceramic resistors as I suggested, you would have had NO problems.
These are not thermally coupled to anything, and they give off their heat to the ambient.

The problem is that you used metal cased power resistors which are normally thermally coupled to heatsinks.
By mounting them onto a metal plate, you now coupled them thermally to that plate, and thus the plate becomes the heatsink.
Which it can for 6W, but it will still get quite warm.
Still OK if you do not use that at the same time as your Gnd bus bar and connect it to the capacitors.

I presume you don't want your caps too warm.
So the easiest thing you can do is to thermally decouple those resistors from the Gnd bus bar.
And the easiest way to do so is , as I suggested, use free-standing 5~10W ceramic resistors.

Now why ceramic resistors ?
You want to filter off high frequencies from your power supply.
So in addition to resistance, a bit of inductance is actually beneficial.
On top of that, they are cheap and easy to get.
So why not ?

;)


QED,
Patrick
 
Thanks for all the inputs, much appreciated. I kind of like the look and feel of these power resistors it matches the " heavy duty" feel of the whole amplifier. What i'll do is thermally decouple the resistors from the gnd plate - but before that i want to listen to the F5 some more - to lazy to go back to the bench.
 

Attachments

  • Power Resistors.JPG
    Power Resistors.JPG
    57.2 KB · Views: 2,395
Last edited:
Yep. Those "metal jacket" power resistors are intended for heat sink mounting, and should be significantly derated if not used that way. In my F5, I added another RC stage to the RCR PSU, and used power resistors (photo). What's obvious, is that I have both the bridge rectifiers and power resistors firmly mounted to the 3/16" chassis base plate. What's less ovbious is that I had masked off pads for the rectifiers and power resistors on the base plate, prior to powdwercoating and used thermal grease between the rectifiers and the resistors and the base plate when I mounted them. (And the entire internal chassis is actually a forced-air plenum that gets circulation from two low-speed, low-noise fans.)

Bottom line..... Don't go for the "industrial look", if you circumvent the proper engineering to make practical use of it.

Ceramic resistors are a better choice, unless you want to "go the distance" with the finned metal jackets.
 

Attachments

  • 42.JPG
    42.JPG
    46.6 KB · Views: 2,390
Member
Joined 2004
Paid Member
Yep. Those "metal jacket" power resistors are intended for heat sink mounting, and should be significantly derated if not used that way.
CanAm Man, that's not how I remember it on several metal Dale's like this. The 50W (on mine) rating is in free air I believe, and at 50W they're hella hot, but I have added additional hs and gone way above 50W.
 
Last edited:
Andersonix....... Yeah..... You've got me wondering about that, and I think you're right. I know in my case, after my PSU calculations, I decided that using the chassis plate as a heatsink was a very good "safety margin" for my resistors. Let me dig out some Dale spec sheets and convince myself..... In any case, my resistors run a little warm, as is the chassis plate surrounding them......
 
I believe

A 50W Dale RH series in free air does 20W at 25C ambient.
The 50W rating at 25C requires a heatsink surface that shouldn't be underestimated.

(btw, the Dale RH-50 and RH-100 have been around for many decades. When i first used them in the mid '80s, Dale was the single ready available industrial power resistor brand available overhere)