By "exchange" you mean switch places, right? (sorry about my not so good English).
A couple of more questions: What's the output impedance of the buffer? Shouldn't there be a small resistor in series with the output?
Can I link out the input caps if the source has no DC? (i'm concerned with phase shift issues).
Thanks!🙂
R7 = 4.7 kohms, R9 = 100 kohms.
The output impedace is in the range of 30 ohms.
A series resistor is often used to prevent oscillation.
Check with capacitive load (your preferred coaxial
cable). It may be easy to get this buffer oscillate.
I think I already suggested that the input cap
can be left away. In this case however a very
small DC correction voltage introduced by the
servo may lead to small noise on the pot wiper
on rotation.
One additional word to C26: it will form a low pass
with undesirably low corner frequency if you use
your potentiometer. C26 must be scaled accordingly
with this kind of souce.
You see that the circuit as is depends on low
output impedances i. e. is suitable in situations
where no buffer is required.
Enough of teaching for now, most of this you
may detect yourself on closer inspection.
Last edited:
Thank you for your teachings and help, as-audio 🙂
So, can I conclude that diamond buffers - like the 2 circuits I posted in the beggining of the thread - are not good for preamp/line level buffer duties because they have by nature always a low input impedance? Should I move to other kind of circuit for this job?
Thank you.
So, can I conclude that diamond buffers - like the 2 circuits I posted in the beggining of the thread - are not good for preamp/line level buffer duties because they have by nature always a low input impedance? Should I move to other kind of circuit for this job?
Thank you.
No, this is not true, they have high input impedance
like all other buffers.
I just wanted to stress that the particular item you
coincidentally picked up, being a poor example of
this type of buffer, should be worked on.
like all other buffers.
I just wanted to stress that the particular item you
coincidentally picked up, being a poor example of
this type of buffer, should be worked on.
Then all is not lost! 🙂
Lets take the servo out. Lets make R11 1.5kOhm and C26 330pF (I believe it will gives me about 321 kHz low pass cut frequency). Lets connect C26 to ground, make R9 100kOhm and link with a wire R7. Then I connect my 100kOhm pot at the input leaving out C16. Would this now work?
Thanks.
Lets take the servo out. Lets make R11 1.5kOhm and C26 330pF (I believe it will gives me about 321 kHz low pass cut frequency). Lets connect C26 to ground, make R9 100kOhm and link with a wire R7. Then I connect my 100kOhm pot at the input leaving out C16. Would this now work?
Thanks.
Cosider that your pot has a "source impedance"
which adds to R11 . It is max. 25 kohms at electrical
center and so C26 must be adjusted accordingly.
Without the servo R9 can be set to zero (short)
or R9 100k and link R7. C16 shorted.
which adds to R11 . It is max. 25 kohms at electrical
center and so C26 must be adjusted accordingly.
Without the servo R9 can be set to zero (short)
or R9 100k and link R7. C16 shorted.
The pot also contributes for the RC filter... I didn't see that! Lets see if now I understand: The pot (wich resistance varies between 0 Ohm and 100 kOhms) is in parallel with R9 wich is now 100 kOhms (R7 is linked out). The resulting resistance is in series with R11. Making R11=100 Ohms and C26=100pF, the equivalent resistance of the net would be between 100 Ohm (pot at ground) and 50.1 kOhms (pot at max signal). The RC time constant is between 0.01 and 5.01 micro-seconds. Multiplying by 2*PI and inverting, we get for the corner frequency: 15.9 GHz and 31.7 kHz. That's a huge variation! With pot at 25 kOhms, Req=20.1 kOhms, RC=2.01 uS and corner freq = 79.2 kHz.
Is this good?
Thanks.
Is this good?
Thanks.
No, the "source impedance" is low at min and max
volume pot settings, in center position there is a
resulting source impedance of about 25k (both half
sections of 50k in parallel), hence the variation of
corner frequency. This is a common effect with
volume pots and non-zero input capacitance.
volume pot settings, in center position there is a
resulting source impedance of about 25k (both half
sections of 50k in parallel), hence the variation of
corner frequency. This is a common effect with
volume pots and non-zero input capacitance.
Last edited:
Yes, it is a common effect with an unbuffered volume control.
It is avoidable.
I will not allow an adjustable HF filter to be an inadvertant part of changing volume.
You need not either.
It is avoidable.
I will not allow an adjustable HF filter to be an inadvertant part of changing volume.
You need not either.
I was seeing the source as an infinite impedance voltage source, you see it more like a short-circuit, right?
Is the RC filter really necessary? Can I bypass it?
Going back to the servo: Isn't it missing a cap between the non-inverting input of the opamp and the ground? I will not use the servo, but I'm curious in understanding it. Should be a 2C servo, right? Why has the designer omited the cap?
Is the RC filter really necessary? Can I bypass it?
Going back to the servo: Isn't it missing a cap between the non-inverting input of the opamp and the ground? I will not use the servo, but I'm curious in understanding it. Should be a 2C servo, right? Why has the designer omited the cap?
I haven't. The thread is a bit advanced for me and requires my carefull study. Right now I'm reading Walt Jung "Realizing High Performance Buffers" and trying to understand diamond buffers. I'll give the due attention to the servo thread next. Is just I had this question in my mind about the servo on this circuit and decided to post it.
- Status
- Not open for further replies.