F5 power amplifier

[bobodioulasso]

Originally Posted by EUVL

> That's one. I can think of a couple others...

Let me make a fool of myself :

Connect the source of Q1 & Q2 on each (left or right) half together, then connect these two junctions across the two halves with a 10R resistor. The feedback resistors are then 25R and goes from output to the source junction. Non-inverting then.

According to John Curl, this "H" diff pair has very good distortion cancellation.


Patrick

Could'nt you show a little drawing, please?

[osscar]

I finally connected together correctly generator and scope.
Here is a picture with a 20kHz square signal on dummy load (i try higher frequency signals also - looks good too), In addition, I connected a scope with 8 ohms speakers- and played Denon audio technical CD (with different signals etc) Even at max volume with passive preamp there is not visible clipping.
After this I conclude that my amplifier works well technically. The rest is a matter of taste.

[a.wayne]

Originally Posted by osscar

I finally connected together correctly generator and scope.
Here is a picture with a 20kHz square signal on dummy load (i try higher frequency signals also - looks good too), In addition, I connected a scope with 8 ohms speakers- and played Denon audio technical CD (with different signals etc) Even at max volume with passive preamp there is not visible clipping.
After this I conclude that my amplifier works well technically. The rest is a matter of taste.

Oscar , could you put up a 20 Hz Square wave , 1 watt , 10 watt etc ...

[gautamz]

I have managed to get 2sj108 but could not get 2sk370... so instead I have picked up 2sk364. The main differences I can see from the datasheets is that 364 has higher power dissipation and higher forward transconductance... Also the 364 datasheet does not contain noise data.

Will 364 be a decent replacement for the 370?

One more thing , I have only managed to get the GR series ... I couldn't find the BL series ... will this be a problem? NP's article said BL is preferable but GR and V should also do.

I guess best way is to try and see if it works

[osscar]

ok, i will try to measure after few days and then post results.

[umut1001]

Originally Posted by cviller

The problem is that you have tried to adjust everything with just one pot. If you have 4.8 over R3 and 3.8 over R4, it means that you have Q3 running at full throttle, only limited by the fact that you have running Q4 only half turned on.

Turn P1 so that you get ~400mV over R11
Turn P2 until you have 0V dc on the output (without burning anything of course)
Check R11 and R12
Make tiny adjustments on BOTH P1 and P2 until you get 0.6V on R11 and 0V dc

I solved the problem.Trimpots were reverse polarized.now i connected reversed and it is playing well now sound is smooth and very detailed My aluminium cooler is 50cmx10cmx5cm buckhorn.It is about 1.5 kilograms.My lab-room is too cold now and i cant guess how it will be in summer.Here in my country it reaches to 39Celcius degree.Do you think my cooler is enough for F5?

[a.wayne]

Originally Posted by umut1001

I solved the problem.Trimpots were reverse polarized.now i connected reversed and it is playing well now sound is smooth and very detailed My aluminium cooler is 50cmx10cmx5cm buckhorn.It is about 1.5 kilograms.My lab-room is too cold now and i cant guess how it will be in summer.Here in my country it reaches to 39Celcius degree.Do you think my cooler is enough for F5?

Post up a pic .......

[kp93300]

I hope someone can help .
I use a promitheus DAC and connected to the F5 via a passive attenuator--Lightspeed.
At 4 V output form the DAC, there is some distortion with some disc.
It sound fine with 2 V output .
What is the maximum tolerable output from the dac ?


It sounds excellent with a total of 160000uF . Is there any advantage in increasing more capacitance in the power supply rails ?

Thanks
kp93300

[udailey]

I dont know the answer here but its true that 4V is higher than optimal for voltage through the resistors of the lightspeed. But you would have to have it at near zero volume to get 4V across that series resistor anyway. This probably is not your problem its just a shot in the dark. The best voltage is around 2.5V for the lowest distortion. I looked for the datasheet that showed that but havent found it.
Uriah

[cviller]

You could also be clipping the I/V. I have a suspicion that some of my problems with pcm1738->D1 I/V stage was due to clipping of the I/V.