Not a different topic. I am asking if I have misunderstood the phase of the correction that gets applied to the _IN node.
i.e. another misunderstanding of the way AC electronics works.
My list of misunderstood topics keeps growing !
Likely if you were to see the whole schematic, OS would have a voltage applied to the NFB signal in the simulation to test the servo. The initial offset needs to be coming from somewhere.
If you look at the design, the input transistors would balance out if a pot was installed between R8, R9 and R10 to trim offset.
Not a different topic. I am asking if I have misunderstood the phase of the correction that gets applied to the -IN node.
i.e. another misunderstanding of the way AC electronics works.
My list of misunderstood topics keeps growing !
This already done.(servo disconnected).
Offset without servo is 2.5mV.
Attachments
The original sch in post1523 is correct, except for that high voltage and the need for 1:1 resistors to apply that big correction.
The revised values shown in post1553 is not good. It appears to have fudged component values without curing the cause that requires the high correction voltage.
what are the voltages at those nodes and what are the voltages across those resistors.
Can you check offset voltage through a whole startup cycle? Does it start excessively high, then drop?
That's interesting! Performance is better without the servo. This would tell us to look at the servo instead of the rest of the circuit.
Yes Jeff,i will post all the distortion measurements, without servo version,soon.Big difference!😱
But that means - the servo output must be somewhere in the 10-th millivolts order - correction required is low.
Let's go!
No servo Symasui Survivor IPS+NS OPS
No servo Symasui Survivor IPS+NS OPS
Attachments
Last edited:
AnderewT that is.
Jeff, without servo offset goes to 2.5mV instantly!
Attachments
Last edited:
I Think so😉
In these two measurements a real 3 way speaker was used as load.
The last one without any load.
Attachments
Last edited:
Testing Symasui Survivor+NS using RMAA
RMAA test Report.
RMAA test Report.
Attachments
Last edited:
-160mVdc at the input node cannot be right.
It should be a +ve voltage, because the base current comes from the +ve supply through R2 to Signal Ground and thence through R3 to Power Ground.
Check the voltage @ R3, looks like it is ~+3mVdc reference the Power Ground and 0.0mVdc reference the Signal Ground.
3mV indicates 300uA flowing from Signal Ground to Power Ground. This could be adjusted to be quite a bit less, but I don't know how it varies with temperature. Worth checking when cold, warm, hot.
Same error for the -147mVdc voltage.
It should be a +ve voltage, because the base current comes from the +ve supply through R2 to Signal Ground and thence through R3 to Power Ground.
Check the voltage @ R3, looks like it is ~+3mVdc reference the Power Ground and 0.0mVdc reference the Signal Ground.
3mV indicates 300uA flowing from Signal Ground to Power Ground. This could be adjusted to be quite a bit less, but I don't know how it varies with temperature. Worth checking when cold, warm, hot.
Same error for the -147mVdc voltage.
Last edited: