See Extrusion Profiles - Heat Dissipating, pick an extrusion profile similar to one you can find locally and change the length in the calculator to suit your desired height. Note that each extrusion side by side will give you half of the thermal resistance, but doubling the height will not.
I have no interest in Aavid Thermalloy, I just find the calculator useful. Be sure to check B and C stock for a match to a profile you find surplus. You'll often find an exact match.
You're going to need a lot of heat sink or a fan to dissipate 540W
I have no interest in Aavid Thermalloy, I just find the calculator useful. Be sure to check B and C stock for a match to a profile you find surplus. You'll often find an exact match.
You're going to need a lot of heat sink or a fan to dissipate 540W
A very large heat sink. Force cooled would be advisable. Firstly you must decide what temperature rise above ambient would be acceptable, then you look for a heat sink that tells you how much K/watt per length it dissipates and you calculate how many meters of it you need to stay withing the temperature rise you anticipate to be safe.
Lets say you find a heat sink that states 0.5K/watt (per some length) you dissipate 560 watt, thus the temperature rise would be 56 degrees centigrade above ambient. If your worst case summer ambient is say 40 degrees centigrade, the final heat sink temperature would be 40 + 56 = 96 degrees centigrade. Besides you will have to consider what would the heat sinks heat up. If you have capacitors rated at 86 degrees centigrade, obviously they will not last. Plastic parts may all go floppy.
Even a small fan circulating the air would make a significant difference.
That is a serious amount of heat just to listen at moderate levels.
Have you looked at a Krell KSA100? It is a ClassA amp with similar 8ohm output but running on 2.6A output bias spread over 4pairs of 250W 200degC devices.
Hi Andrew you a right.
bias 11A
PO 120WRMS into 8ohm
P=V*V/R (Take Drop 3V)
V needed= sqr(120*8)*1.4=31*1.4=44 plus drop = +/-47V =94V
Bias current needed is only 5.5A for SE.
Power wasted: 5.5*94=517W
This SE current total is 11A
Power wasted: 11*94=1034W
Its a dangerous Heatsink calculation. 500W heatsink won't work, you need a Water cooled Radiator!
It is probably bridge mode. two of 47V 5.5A SEclassA, then the total is 517W, then drop is only 1.5V each amp.
bias 11A
PO 120WRMS into 8ohm
P=V*V/R (Take Drop 3V)
V needed= sqr(120*8)*1.4=31*1.4=44 plus drop = +/-47V =94V
Bias current needed is only 5.5A for SE.
Power wasted: 5.5*94=517W
This SE current total is 11A
Power wasted: 11*94=1034W
Its a dangerous Heatsink calculation. 500W heatsink won't work, you need a Water cooled Radiator!
It is probably bridge mode. two of 47V 5.5A SEclassA, then the total is 517W, then drop is only 1.5V each amp.
Last edited:
This is the schematic
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
That looks like a guess at making a sort of bridged Zen with an Aleph afterburner. I don't see any "X" feedback. I don't see any bias on the input mosfets. He's missing input caps, required because if this beast is to work the gate of the IF9610s need to be at ~4v. There should be a 100R or so resistor between the gates of the IRF9610s and the zener strings. I wonder if anyone has built this successfully.
IMHO, if you want an AX type amp a much better proven solution would be to build an AX or AJ-X on Peter Daniel's boards. You can fairly easily get 120W+ out if you build it with three outputs per quadrant. Dissipation will be a bit less, but it will still take a lot of aluminum and a lot of power transformer.
If you don't want to or can't find the JFETs needed, you can use ZVP3310s, being careful to reverse it from the orientation on the boards to correct the lead order.
IMHO, if you want an AX type amp a much better proven solution would be to build an AX or AJ-X on Peter Daniel's boards. You can fairly easily get 120W+ out if you build it with three outputs per quadrant. Dissipation will be a bit less, but it will still take a lot of aluminum and a lot of power transformer.
If you don't want to or can't find the JFETs needed, you can use ZVP3310s, being careful to reverse it from the orientation on the boards to correct the lead order.
Assuming that you use enough output mosfets, heat sinking, bias and rail voltage you can get virtually any power you want. The caveat with JFETs is that the rail voltage probably shouldn't go much higher than 24V.
I've reposted the AXE1-2 spreadsheet by Wuffwaff (IIRC). Peter's boards can be configured with 4-12 FETs. At 24V rails, 60% AC current gain and at least 4.5A total bias you can get 121W RMS. (You can also get there at 50% and 6A min) At 240+W dissipation you'll need a lot of heat sink, but still less than your other design.
I've reposted the AXE1-2 spreadsheet by Wuffwaff (IIRC). Peter's boards can be configured with 4-12 FETs. At 24V rails, 60% AC current gain and at least 4.5A total bias you can get 121W RMS. (You can also get there at 50% and 6A min) At 240+W dissipation you'll need a lot of heat sink, but still less than your other design.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.