F5 power amplifier

[Zen Mod]

Iznogood

A Class amps are outa fashion

go and buy Devialet

 

[adason]

Winter comming, classA good, very good.

 

[adason]

Isn't Devialet nested classA amp inside classD?
Got schematics ZM?

 

[ACnotDC]

Devialet is out of the question, has no KISS

 

[adason]

But it sounds better than F5.

 

[6L6]

ACnotDC - that looks very, very nice! 2h dominant, consistent stair step down 2h 3h 4h, no high-order artifacts poking their head up through the grass, there is nothing at all to complain or comment on about that measurement.

Sit back and enjoy the music, the amp is beautiful.

 

[leebert]

6L6 - I know it's in here (and other places) why there's a spike @ 1k - can you explain why, or do you have a link discussing the topic? Thanks - about to build an F5 and would be alarmed by the "1K spike issue" (or non issue).

Lee

 

[Nelson Pass]

Are you referring to the 1k test signal?

 

[leebert]

Mr. Pass - yes, that's it (from ACnotDC's pictures a few posts back) - again, sorry for bringing up something that's raised/answered elsewhere - a link would be fine, or the right search terms.

Merci!

Lee

 

[A Jedi]

Mr. Pass - yes, that's it (from ACnotDC's pictures a few posts back) - again, sorry for bringing up something that's raised/answered elsewhere - a link would be fine, or the right search terms.

Merci!

Lee

A link is not needed as Nelson just answered the question - the 1K is the test signal.

 

[leebert]

Ha! Indeed! Quite embarrassing. Didn't see that anywhere, but should have deduced (what else could it have been?).

Cheers!

Lee

 

[mikelm]

Hi All,

I'm in the process of building an amp loosely based on the F5 and have a few questions.

Have been reading through this thread but not yet finished so please excuse me if these questions have been asked & answered already.

1) Has anyone noticed a subjective improvement by removing the thermistors ?

2) I have seen the damping factor quoted as 40, 60 & 80 in different places - has anyone measured the damping factor their F5 is giving ?

I mention this because in Spice I am finding that I need about 30db of FB to get a DF of 32 and have read that the F5 FB is not this high. So this situation is as yet not clear to me.

thanks

 

[rayma]

I mention this because in Spice I am finding that I need about 30db of FB to get a DF of 32

Don't worry much about DF, since anything over ten or so is fine.
Of course, the amount of feedback will also affect the measured distortion.
Beware of changing the circuit, the drift will increase.

 

[Nelson Pass]

My thumbnail calculation says the DF ought to be about 35, but I will have to travel to
storage to confirm what the real number is.

 

[Zen Mod]

My thumbnail calculation says the DF ought to be about 35, but I will have to travel to
storage to confirm what the real number is.

do not

it's snow outside

 

[mikelm]

Don't worry much about DF, since anything over ten or so is fine.
Of course, the amount of feedback will also affect the measured distortion.
Beware of changing the circuit, the drift will increase.

Thanks ramya, Good to hear. For the initial build I am planning 27db of FB which should give DF of about 22. Yes, without thermistors the current will drift but that does not bother me so much. I will try both ways and see how they sound.

 

[mikelm]

My thumbnail calculation says the DF ought to be about 35, but I will have to travel to
storage to confirm what the real number is.

Thanks Nelson, The thumbnail calculation is just fine.

Will try few different CLGs and see how they sound.

 

[Sangram]

1) Has anyone noticed a subjective improvement by removing the thermistors ?

Marginal and not worth it for loss of thermal stability. The offset also tends to wander around a lot, so it's not just the bias.

Can get good result bypassing the thermistor with a small cap, about 1nF. This will affect the bandwidth but improves stability, again only marginally.

I simulated the thermistor with a small amount of series and parallel inductance, the frequency response really didn't change as the value was moved up and down. Unfortunately I no longer have the sim files, thanks to Microsoft.

 

[mikelm]

Interesting - I tried it in spice and the effect seems harmless.

If the thermistors are generating some noise this should help.

thanks

 

[rayma]

Thanks ramya, Good to hear. For the initial build I am planning 27db of FB which should give DF of about 22. Yes, without thermistors the current will drift but that does not bother me so much. I will try both ways and see how they sound.

Also the output DC offset voltage will drift more.