JFET input, MOSFET VAS, LATERAL output = Perfect!!

[mikelm]

mmm - ok, I'll check it out sometime soon - thx

but I don't think that <5pF miller cap will be audible

[mikelm]

Originally Posted by MiiB

I dont think that the sound suffers by a lower OLG...

Well, this is dependant upon other factors.

I would agree with you if you have an extremely clean power supply, but in my experience whenever measures are taken to reduce PSU ( and other ) noise, the sound of the amp improves.

More FB, through Higher PSRR reduces PSU noise being expressed at the amp output so personally I would want to see extensive power supply filtering before I would be prepared to say that lower OLG does not decrease sound quality in some ways - even though it might improve it in other ways.

I will experiment with this when I have built my CLC PSUs.

mike

[AndrewT]

Hi Mikelm,
I have only built one channel.

Have you noticed that output offset varies with speaker load?
Open circuit is different from 8ohm and that is different from 4ohm.
This will also apply to headphones, particularly with the very wide range of HP impedances.
I cannot hear the 0.8mVac of hum with my speakers, not even a pair of 8ohms in parallel for an effective 4ohm loading and +6dB more hum.
I can easily hear that same 0.8mVac of hum through headphones. I will have to try rCRC instead of my rC supply.

The next channel will be built with output source resistors to see what effect they have that I am able to detect.
My small board layout does not suit modifying, but I knew that before I started.

[keantoken]

That offset changes with load probably only indicates that the amp has high DC output impedance, or that the supplies have high DC impedance.

- keantoken

[AndrewT]

Originally Posted by keantoken

That offset changes with load probably only indicates that the amp has high DC output impedance, or that the supplies have high DC impedance.

- keantoken

the offset current flows through the NFB upper resistor and then through the speaker to ground. The output impedance of the output stage tries to bypass the speaker impedance. So yes high DC output impedance will force most of the offset current through the speakers or headphones.
It can still be trimmed out using the output offset adjuster, but that only holds low output offset, if one does not change the speaker/HP impedance.

[mikelm]

Offset spot check

8ohm impedance speaker connected = -30mV offset

with speaker disconnected = -32mV offset

[mikelm]

Originally Posted by AndrewT

the offset current flows through the NFB upper resistor and then through the speaker to ground. The output impedance of the output stage tries to bypass the speaker impedance.

Hi Andrew,

I do not understand your description above.

The NFB path and the speaker are normally in parallel.

my o/p impedance when I measured it 2 weeks ago was about 0.2 ohms

probably a bit less now

mike

[keantoken]

If you adjust offset to zero with one load and it changes for another load, it could be oscillation. An offset set to zero should stay there no matter what the load.

- keantoken

[GregH2]

Hybrid CCS courtesy of Paul Bysouth

Sounds very very nice.

[keantoken]

Hey, just a thought. Now that the CCS is bootstrapped, it should have pretty even dissipation. Maybe it could be changed to a Jfet, for better performance?

Perhaps a GR grade 2SK170, with R18 being 1.8k, would give about 6-7V across it, for about 56mW of dissipation. There is plenty of room here to experiment with bootstrap values and voltages, to compromise between bootstrap and Jfet.

EDIT: I calculated this based on 8mA VAS current, but now I see in the schematic VAS current is 10mA.

I wish I had thought of this earlier.

- keantoken