One Aleph-X working, One to go

[wuffwaff]

Hi Uli,

no. These are called C7 and C8 in Chad´s schematic. What would be the advantage using these?

William

[uli]

They reduce open loop gain. This is often necessary to
avoid phase problems as at a high frequency often the
amp turns phase in such a way that NFB might become
positive FB resulting in oszillation.

Uli



PS: I do not know Chads schematic, therefore I referenced
to a public known circuit. I am absolutely sure that NP put
those caps in his AX series too!

[wuffwaff]

Hi Uli,

I´ve tested it up to 1MHz. I can´t see any oscillations and frequency response is down 3dB at 140kHz 6dB at 217kHz. If there´s no problem to fix I think I can leave them out can´t I?

William

[uli]

Hi William,

I see from the posts before that you put some caps into the
circuit. Without precisely knowing this particular circuit I am
not able to tell anything about it.
In my amps I put 15p parallel to the feedbackresistors (100k),
1n parallel to the bjt´s ce (in ccs) and another 1n as Cdom.
Very close to the Aleph 2 circuit.
In fact it is possible to stop oszillation at many different points.
I just used the standard textbook Cdom for default VAS stages.

All different possibilities to stabilize the amp do basically the
same:
They reduce open loop gain in such a way that you have some
phase margin left for applying NFB.

Uli

[wuffwaff]

Hi,

just look at the Aleph-X WIKI for Hifizen/Gratakus schematic.
I uses 2x 10pF parallel to the feedback resistor and 4n7 in the active current source (3n3 would probably also do the job)

William

[uli]

I would put the Cdom in. The resistor parallel Rnfb increases
NFB with higher frequency to form a lo pass. With Cdom the
amount of NFB stays constant. This improves stability.

Uli

[wuffwaff]

OK,

I´ll try it when I´m back from holidays.

William

[Netlist]

Originally posted by uli
I would put the Cdom in. The resistor parallel Rnfb increases

And what would be Cdom? Rnfb would be the negative feedback resistor?

/Hugo

[uli]

Cdom is called like this in various pieces of literature.(Self et al.)
It is the cap in the VAS between base (gate) and collector(drain).
Yes Rnfb Is the feedbackresistor from output to -in.

Uli



PS: from what I learned dom is an acronym for dominating pole

[wuffwaff]

Originally posted by wuffwaff


Upon startup rel. DC offset is now 18mV, absolute offset is smaller 0,3V and somehow stays there.

William

Well after putting the lid on the amp absolute offset went to +4V........
Had quite a bit of work coming back under 1V since the holes in the lid are in the wrong place. I can´t adjust VR2 without taking it off making the procedure a bit difficult. Will think of this for the next project!
I now checked absolue offset from (stone) cold:
startup -8V
5 min -4.3V
20 min -2.25V
60 min -0,3V

it sort of creeps up very very slowly after that. I suppose that´s because my little room gets warmer and warmer.....

The transformer has also developed a little bit of mechanical hum (would have been too good to be truth). So I suppose I´ll have to decouple it from the frame with some rubber.
I also have a bit of hum in the speaker (can hear it when my ear´s nearer 10cm to the woofer). I hope I can get rid of this by shielding the chokes with some mu-metal sheet. Or maybe it is gone when connected to my X-BOSOZ.

William