A5 behaves strangely

[Stabist]

Hi!

I friend of mine decided yesterday to make pretty detailed measurements of his Aleph ...
We were not able yet to comment the results in "live" - so I'm not too detailed informed yet ...

As far as I understood this are his finding outs:
- it doesn't develop 60W but only around 15-20!?!?!
- that's because the bias is only around 1,3A!! - altough all values should be pretty close to the ones in original schemtaic - but it is true, that PSU gives out 32 instead of 34V - can that cause such difference? Or are the tolerances of those 1R resistors the guilty ones?
- then it also cuts one half of the signal much much before than the other - but when he put -1V DC offset on the input - everything became much better - is that the problem with input differential pair (their matching) or something else?

Thanks for tips and help!

[wuffwaff]

Hi Stabist,

power depends on bias and ac-current gain. With 1.3A bias and 50% you can expect aprox. 25 watts into 8 ohms (12.5 into 4 ohms). The voltage is more than high enough for the little bias your friend is using.

Bias should be set with R19 (use a pot then replace with a resistor). After that you should measure ac-current-gain and set as you like (between 50% and 65%.........)

The Aleph will clip unsymmetrically by design. Positive swing is limited by bias, voltage, ac-current-gain, negative is limited by voltage and the current limiter. This is only there to protect the amp in case of a shortcut at the output.

If the DC offset is allright (<50-100mV) then the input pair is matched allright.

William

[Stabist]

Thanks wuffwaff!

The thing is - he thought he has "full" (2+A) bias - considering his A5 is built with same resistor values as in original schematic ... The differences are just in transistors (his version uses irfp240, irf9620 and ZTX450) and in fact that PSU voltage is 32V instead of 34V ....
We just wonder how come that it is such big difference betwen original A5 and this one ...
OK now when we know that we must lift bias - no problem (I hope) ...

[Stabist]

Oh another thing - I just don't know anymore which method to believe when measuring bias - in my Zen v4 - If I measure voltage drop on resistor - and than calculate it - it has 2,1A - BUT if directly measuring current that flows from PSU - it is only 1,3A!!??!?!

[wuffwaff]

HI Stabist,

both methods should give the same result. You should know the exact value of the resistor though.

My Aleph 5 had 2.4A bias with the same values........The value of the bias depends a bit on Vgs of the fet you are looking at to set the bias (Q6) and Vbe of Q5.

William

[Stabist]

Considering resistors - I can just believe that they are really pretty much same values as noted (and they are supposed to be 2% ones)- because my multimeter is very inaccurate for resistors below cuople ten Ohms ...
So - a combination of resistors in Zen gives me 0,33R - and voltage drop is around 0,75V
But when measuring current with meter in series in PSU line - it shows it draws only 1,3A ... ?!? OK I said to myself - also the mutlimeter ain't pretty precise - but still - to get such difference - there must be something wrong ...

[wuffwaff]

Hi Stabist,

I think the only thing that is wrong is the measurement Current can´t disappear or appear so if it flows through 0,33R it must come out of the power supply.
Did you look at the signals with a scope?

What you could do is to put another 0,33R in the power supply line and measure the voltage there and calculate the current.

william

[kirc]

Hi!

Right values of bias are:

Left channel: +1.57A, -1.57A
Right channel: +1,59A, -1.59A

Power supply voltage on power capacitor:
+32.08V, -32.08V

Trannie power:
P=218,8W
S=339,9VA

Maksimum power is 30W on 8ohm and 12.5W on 4ohm.

All measurment was taken by professional equipment (curruent probe - Omicron, and power analyser,...)

Picture of maksimum sinus wave without a disstorsion on 8ohm load.

[kirc]

Max sinus wave without disstorsion.

[kirc]

And maximal sinus wave with disstorsion.

What can I do ?
More bias ? Thats for sure!