Trying to build...an...OMG

[PhaseLockLoopy]

... a MONSTER of an amplifier
...fuelled by drinking too much Scotch.

Either the final version will explode in my face, leaving me with permanent scars, full removal of all facial hair...or it'll provide the quality and current I need to run ribbon speakers properly.

So development has started in minature scale first.

[PhaseLockLoopy]

Based on thinking behind an earlier Levinson design I saw back in the late 1990's I'm mucking about with a symmetrical current mirror input stage - get slew rate up - noise down and aiming for low distortion with devices available on the market today. Last time I designed an amp was 1996 - technology has moved on since then, I'm pleased to say.

[PhaseLockLoopy]

....Intended final scale

A Pair of 300Watt Mono-Blocs ------ heatsinks already purchased

[Bigun]

This looks quite an exciting project !

You clearly have a properly equiped working area

[PhaseLockLoopy]

I was working on this until 3.00am - dogged by a couple of problems
Managed to get positive and neg drive near perfect
but have xover distortion at 160Khz when testing extreme HF performance
Not quite got the gain I was hoping for yet.

What I may do is swap BC550 and 560s for the 2SC2240 and 2SA970 pairs for the input

[PhaseLockLoopy]

Hi Bigun

I'd like to change the bias stage on this thing - I'm not entirely happy with it

what's the best simulation package these days?

[jleaman]

Where did you get the heat sink stuff ?

[PhaseLockLoopy]

Hi jleaman

A heatsink designer/manufacturer near me they sorted me the copper bars also

I have some more on the way for the 120W version of this -- the big monster sinks are for the 300Watt variant I'll build if I get the 120Watt correctly adjusted.

[jleaman]

Originally Posted by PhaseLockLoopy

Hi jleaman

A heatsink designer/manufacturer near me they sorted me the copper bars also

I have some more on the way for the 120W version of this -- the big monster sinks are for the 300Watt variant I'll build if I get the 120Watt correctly adjusted.

Very sexy heat sink's, would be good for a nice Class A Beast!

[PhaseLockLoopy]

Been having some instability issues - so been a while since posting - not getting the neg feedback chain and zobel filter problems totally ironed out, it's put a bit of delay into the project -- it's also made me change the bias chain to look at current shunt. - I've taken out the pre driver stage as it was offering nothing more than overcomplicating the design. I'm now looking over the zobel trap


Design beginning to come together - through some trial and error - the control relating to forward feed into the rear of the two current mirrors has been causing some induction of RF oscillation and I'm not entirely sure why yet.

My initial trials have made me change post bias drivers - but has set me to use the BC550c and BC560c for the current mirror front end. Managed to get down to...

...spectral density of 3.9 nV/√Hz using my Racal spectrum analyser
with input offset current only 2.7nV -
but this was working in miniature at ±20V - not yet the full ±55V the finished design will be loaded with.

as for the rest of the circuit -- I've yet to devlop it fully

with the issues I am having with feedback loops - I managed to turn it into a 2Mhz oscillator -- not very funny because I'm not trying to make a radio transmitter.