F5 power amplifier

[woodturner-fran]

Well Peter, you are bang on correct and I am disgusted. I installed Q2 backwards. Never saw it despte checking.


So I turned them around but neither channel works. Could some kind soul let me know whats likely to be dead and what I need to replace?


Fran

[dcbingaman]

Bridging two F5's across the same load (say nominally 8 ohms) as a single F5 both doubles the voltage swing and the current, all else being equal. P=IV, so 2 x 2 = 4. A bridged F5 pair should produce at least 100W, but it will come out of Class A operation at a percentage of it's output rating....probably ~ 50-75W. The big IF is the current limit of the amplifier, which in NP's circuit is nominally 10A. Driving an 8 ohm load, each side will conduct about 2.5A and together put about 30V across the load.....at which point they will go into AB mode.

The ultimate limit of AB mode is probably set by something else in this circuit, but when I get mine built I will have my buddy Zaw test it on his HP bench and I'll tell you when it clips.....knowing NP's bullet-proof design approach I'm betting about 120 watts into 8 ohms. Perfect match for a set of Vandersteen Quatros or 5A's, and the first watt will still be sweeeeeeett Class A "quantum triode."

[EUVL]

I still have some matched Fairchilds left that I can let go, but only as sets of 4 (2 matched FQA19N20C and 2 matched FQA12P20).

Matching is single point. Matching details pls see European GB thread.

http://www.diyaudio.com/forums/showt...44#post1495444


Patrick

[EUVL]

> You could also use the Toshiba 2SK1530/2SJ201.

But they are much more expensive than IRF and Fairchild.
If you want true complementary, I have not seen better.
Measurements posted in this thread before.

http://www.diyaudio.com/forums/showt...t=#post1507044

> Am I right in thinking that we'd need to see ~2.7V across R3/R4 (instead of the 4.2V for the IR or Fairchild parts) in order to bias the Toshibas to 1.3A?

Yes, so you need to adapt the circuit (e.g. lower bias for the JFETs). I used 6mA Idss, which are unfortunately not easy to get. You can of course also lower the resistor value at the drain of the JFETs.


Patrick

[umut1001]

hi EUVL..i 'm trying with all of these output mosfets..fairchild sounds more clear..treble is better..and 2sk1530 is stable and good..
did you try irf540/9540? has anyone tried them? i asked twice

[EUVL]

> i 'm trying with all of these output mosfets..fairchild sounds more clear..treble is better..and 2sk1530 is stable and good..

That depends on how you change the circuit for the Toshiba's to make the most of their complmentary characteristics, and to compensate for their lower Vgs.

As mentioned, those who need Fairchilds but could not wait for jackinnj can contact me by PM.


Patrick

[umut1001]

i only reduced the trimmer value and got 2,7 volts for 2sk1530/j201..and everything is allright..i didnt change anything other..
but these mosfets are expensive..can we try irf540 /9540 ?they are much cheaper...but maybe the circuit must be redesigned

[EUVL]

> but these mosfets are expensive..can we try irf540 /9540 ?they are much cheaper...but maybe the circuit must be redesigned

Maybe you should just stick to the Fairchilds and the original Pass circuit.

You would not get the best from the Toshiba's unless you change the circuit to balance the open loop gain of the top half to the bottom half.

As to FETs being expensive, I wonder what % they cost compared to your finished amp with proper housings, heatsinks, power supplies, connectors, .......


Patrick

[umut1001]

you say you should change other things..can you show us redesigned f5 amp for 2sk1530/j201 ? thx

[woodturner-fran]

I installed Q2 backwards. Never saw it despte checking.


So I turned them around but neither channel works. Could some kind soul let me know whats likely to be dead and what I need to replace?

Anyone any ideas?

If I just order more of Q1 and Q2... would Q5 and Q6 be affected by my mistake?


Fran