Good point. If you built the amp according to the schematic in post 12, the +IN should be grounded via R2, sort of. However, -IN will be floating, since you disconnected the feedback and C1 does not pass any DC. Sorry, I forgot about this, because I didn't use C1 in my Alephs... Please do the following (for testing purposes with the feedback loop disconnected):
* Connect both +IN and -IN to GND
* Put a jumper wire across C1 to establish a DC reference to GND for -IN
* Repeat the VGS measurements. If you get different numbers as before, report them here. If the numbers are the same, see below.
This looks about right, although the MOSFET pairs are not perfectly balanced. But they should be okay as it is.
I am not sure what you mean by this.
So the amp does not want Q7/Q8 to turn on as much as Q5/Q6.
I have not read all previous posts, so I may be repeating things... The DC idle point of Q7/Q8 is controlled by the voltage across R7//R7', which in turn is determined by the R7//R7' resistance and the current through R7//R7'.
Please check:
(1) What is the resistance of R7//R7' (use your DVM to measure it with the amp turned off)?
(2) What is the DC bias current of the input stage?
(2a) Measure the resistance of R8//R8' (amp turned off)
(2b) Measure the voltage drop across R8//R8' (amp turned on)
(2c) Calculate the current from 2a and 2b
Last edited:
With RED lead to Source and BLACK lead to Gate.
I was just establishing polarity of the readings that I took.
I was just establishing polarity of the readings that I took.
I tested every component before it was installed so I'm pretty confident that R8 will be 1K0, as advised this is a fixed resistor.
R8 is right under the silkscreen lettering “LTP Bias.” It’s that green resistor you have installed.
Best,
Anand.
That's not just R7, but R7//R7', right?
(R7 is the fixed resistor and R7' is the trimpot, and they are connected in parallel.)
That would be R8//R8'. Now that's interesting... R8=1k (fixed) and R8'=0-2k (trimpot) according to the schematic. Therefore R8//R8' cannot be higher than 670 Ohm. However, your 960 Ohm value should give approximately 10 mA current through Q2, which is about right. Let's leave that as it is for the moment.
Anyway, please continue as I wrote before (with the feedback resistor disconnected):
(1) Connect the inputs to GND and put a jumper across C1
(2) Measure the Vgs voltages at the power MOSFETS again
(3) Measure the voltages across R8//R8'
I've reconnected R4 and after about a minute I can get the output to settle to about 600mV. There is a horrendous period at turn on when the output is almost full rail.
It was stable, I've not touched anything and now its wandering again slowly increasing through -3.5V