F5 power amplifier

Hi Anil, look at Peter Daniel's implementation with two Conrad heatsinks on this thread. I had the same question and was working on a two-sink layout, but eventually I managed to get the sinks locally (one of which Ali is talking to you about - I've been advising him on the heatsink selection).

If you keep them close together and have a metal bar joining the sinks you can ensure sufficient thermal coupling to keep the offset under control. You can also use an 8" length of 8mm thick copper bar to mount the devices, and secure the bar to the sinks. The temperature of the power transistor has some correlation to the operating point, and relative thermal and electrical balance between the channels determines the output offset.

In the case of the profile in the attachment, a single 10" profile is what I use for one F5 channel and it works pretty well, in winters I get a maximum temperature of 45 degrees, and in summer it goes up to maybe 50-55. I believe Ali has recommended this as the 7" profile you are talking about.

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.


The amps don't look like this anymore, this was a testing setup only but you get the idea.

Horizontal fins is not a great idea, unless of course you can get something like the First Watt originals (which are patented AFAIK). Long heatsinks are not necessarily much better, the further away the edges go from the heat source, the less effective it becomes. You need a lot of mass and sufficient surface area within close distances from the source of the heat.

Good Luck. Finding heatsinks in this country is very difficult - almost impossible to get anything more than 8" wide. Ali is working on some chipamp cases for me using some 8" width sinks, even cutting them to length was an issue, let alone extruding them.

Sangram,

Thanks. I agree finding heatsinks is tough. I am planning to do with what I get and make best out of it. Having a copper strip connecting two pieces together seems a good idea.

I am still figuring out things on this matter. Will keep you folks posted.

Cheers.
Anil
 
Last edited:
Many of these "unfriendly" heatsinks will work extremely well with a slow draught (draft = slow breeze) across them and as the new breed of computer fans are now actually really quiet, it is a very viable alternative to big mass heatsinks and more flexible amp cases - just a bit of that lateral thinking works wonders.
 
Moderator
Joined 2002
Paid Member
Yes, a fan will make even smaller heatsinks work well. But a good PC fan costs INR600-700 (~$12-14), two of them will cost as much as the difference between a good sink and an average one :) I crunched the numbers for a long time before heeding advice on the thread and buying decent sinks.
 
Many of these "unfriendly" heatsinks will work extremely well with a slow draught (draft = slow breeze) across them and as the new breed of computer fans are now actually really quiet, it is a very viable alternative to big mass heatsinks and more flexible amp cases - just a bit of that lateral thinking works wonders.

Using quiter fans opens huge number of alternatives in terms of heatsink choices and cabinet designs. Should the draft of air flow through the fins of the heatsinks or can it be on the smoother side of the heatsink where the MOSFETs are mounted?
 
The heatsink still works as a heatsink, air flow or not, so heat "resistance" equations are generally applied thru the median (average heat transfer) of the fin area with the airflow taken as constant flow thru the fins
- as none of this is strictly true in practical applications, the equations have built in "fudge factor" (or variations) that is suprisingly pretty accurate - the tricky bit is finding the technical specs for the particular heatsink profile and it's usually simpler to just apply some heat to the sink (power resistors with known VI) and measure temp rise reduction via different airflows - this is all pretty "old hat" these days.

There is a whole world of information on the net about this sort of thing - industry has been doing this type of thermo control for many years, so plenty of information is available.
 
Hello gents first post on this thread, I have a quick question in relation to heatsinking the devices. Can I use an L bracket that mounts directly to a heatsink? The join would be thermally treated etc.
I have the original thin longer pcb clones, and can see no other way of mounting these suckers without customization. An L bracket would save a lot of pain.
 
I keep getting PMs from people asking me about the Xed F5.
You can see the details in post #1188 onwards.

I am sorry to disappoint you, but I don't sell PCBs because the original circuit is Nelson's commercial design.
Nor am I publishing the layout, because I have no time to support all technical questions.
And if I were to make the same again today, I might also do things slightly differently.

But the Xed version as published has been successfully built by at least one other person (Danke, Uwe) and he seems to like it. So it cannot be too difficult and it cannot be too bad, I guess.

Sorry I am not of more help.
This is only hobby for me, and I also need some time to finish unfinished projects.


Patrick
 
I've been looking around for heatsink for the F5. Not many choices. I'm probably going to order heatsink from HeatsinkUSA . I assume their 10 inch profile will be good for the F5. Could someone tell me how much of the 10" profile would be required for each channel? I'm guessing 10 inches would be the minimum. However I don't want to get it all together and have it running too hot.
 
I've been looking around for heatsink for the F5. Not many choices. I'm probably going to order heatsink from HeatsinkUSA . I assume their 10 inch profile will be good for the F5. Could someone tell me how much of the 10" profile would be required for each channel? I'm guessing 10 inches would be the minimum. However I don't want to get it all together and have it running too hot.

I am using a 7" long piece of the 10" per channel and I measure about 45 deg C after about a couple of hours running. The fins are vertical for airflow and that makes a tremendous difference.
 
Experience with F5 and Quad ESL57?

I am looking forward for saturday:

I will get two Quad ESL57's and two Quad Monoblocks, a FM tuner and a line amp (for a little bit more than an apple and an egg).

The goal is to refurbish the monoblocks, the tuner and the lineamp, sell it and get the needed money to (let) refurbish the ESL panels.

My first highend speakers where ESL57, 35 years ago. This was the time, when my audio fever started :D

Does anyone have experience, driving ESL57 by a F5?

It should not be a problem for the impedance, going down to about 1 Ohm for some frequencies with the ESLs.

The power should also be O.K.

And the gain? Enough?

Franz
 
Last edited: