F5 Turbo Circuit Boards

[sippy]

.......yes Zen, on a pair of horns = 'OH MY!!!!!'

[Daniel(Del)]

Originally Posted by jacco vermeulen

I'd be surprised if they handle more than 85W at 25C temperature rise, much less would not be a shock.

I'm thinking about getting some Constantan wire and placing it across the sink with thermal grease and measure how much the temperature increases for a given dissipation. Do you guys think it would work? This way I can be sure how far I can go with it.

Originally Posted by jacco vermeulen

(really pretty heatsinks for a Class AB amp though)

Even so, a Class AB amp inside this case would be limited to 170W/channel, so that worst case dissipation is 85W/channel, right? Or more than 25C temperature rise is allowed when the amp is retaining half of the output power?

Obrigado

[AndrewT]

One of the attributes of Constantan

Temperature coefficient at 20 °C[4] 8 ppmK-1
Temperature coefficient -55 to 105 °C[1] ±40 ppmK-1

is that it's thermal coefficient of resistance does not change, significantly.

worst case dissipation is 85W/channel, right?

No.
Worst case would be @ ~60% of maximum power output continuously while the mains power is at maximum voltage.
That will probably give a worst case heat dissipation of 120W/channel.
If you want to design for that worst case scenario, then expect the heatsinks to be enormous.

[jacco vermeulen]

Originally Posted by Daniel(Del)

some Constantan wire and placing it across the sink. more than 25C temperature rise ?

Easier would be to just stick a power resistor/transistor to the heatsink with a spring.

Worst case for a Class A amplifier is the idle situation. With a signal on the input, the load takes care of part of the dissipation.
Thermal calculation starts with the steady-state operating temperature of the heatsink.

Heatsink calculation for a Class AB amp starts with worst case dissipation when driving a load.
Temperature will vary and max temperature of the heatsink for the math is much higher.
For the thermal calculation, the load number has to be picked.
Why some commercial manufacturers had a regular version of an amp model, and a Heavy Duty type of the same one, with a bumped-up power supply and increased heat sinking capacity (aka heatsink exchange).

As Class AB is a compromise by definition (imo), monitoring dissipation and temperature is standard practice (V/I limiters, output relays, thermal breakers).
Which is called smart engineering, only a commercial fool would lay out a subwoofer power amp and market it as a domestic amplifier.
So the typical high volume Japanese power amp has whimpy heatsinks and shouldn't drive low impedance loads.
A serial production PA amplifier on the other hand will have a fan on stand-by, for when the going gets rough.

Story is different when the amp is manufactured in/for elevated temperature areas.
Your Nashville amp was made do Brasil, so it has larger than average heatsinks, and very pretty ones imo.

De Naah-da.

[UKToecutter]

Anybody else want to volunteer for prototype builds?
First boards have been shipped.

[a.wayne]

Toe, Remember to get a set to dutchie ........

[Zen Mod]

Dutchie is only faster than moi .

that's pretty slow

[Daniel(Del)]

Originally Posted by jacco vermeulen

Easier would be to just stick a power resistor/transistor to the heatsink with a spring.

Spring? I didn't understand!

Originally Posted by jacco vermeulen

In the other pictures it looked like two heatsinks in series, turned out to be just a scratch, right in the middle of the heatsink.

Why 2 heatsinks in series work better than one?

[AudioSan]

Originally Posted by UKToecutter

Anybody else want to volunteer for prototype builds?
First boards have been shipped.

i've been thinking now. and it will be a V3 over here
with 0.7A bias pr device. i start there. maybe i can pump up the bias a little bit. rails will be +/-32V.

edit: sorry. 0.8A

[AndrewT]

Originally Posted by AudioSan

it will be a V3 over here
with 0.7A ([/i]sorry. 0.8A[i] bias pr device.......... rails will be +/-32V.

a single pair will give a maximum ClassA output power of P(ClA) = [2*0.8]^2 * 8ohms / 2 = 10W into 8r0.

If you fit a 2pr output stage the maximum ClassA output Power increases to 4times (P*2^2) the single pair. i.e. 40W
a 3pair stage will give 90W (*3^2)

This latter 3pr will require an output voltage of 0.8* 2 *3pr * 8ohms = 38.4Vpk
This will require operating supply rails of ~ +-42Vdc

Substitute 1pr or 2pr or 4pr into that formula to find Vpk for the alternative output arrangements.