Hi,
I was wondering if someone could help me out with pointers and
a good place to start for learning how to make a Class A output
buffer for a mic preamp. I am uncertain if this buffer should be
inside the feedback loop or outside of it. The input uses a mic
input transformer and the main gain element is an OPA627 with gain variable between 3 and 160. It sounds great with just little
load on the output, but a test with 600 ohm loads introduced
enough high frequency distortion to make hi-hats sound rather crappy, especially at higher gains.
Thank You,
George
I was wondering if someone could help me out with pointers and
a good place to start for learning how to make a Class A output
buffer for a mic preamp. I am uncertain if this buffer should be
inside the feedback loop or outside of it. The input uses a mic
input transformer and the main gain element is an OPA627 with gain variable between 3 and 160. It sounds great with just little
load on the output, but a test with 600 ohm loads introduced
enough high frequency distortion to make hi-hats sound rather crappy, especially at higher gains.
Thank You,
George
> 600 ohm loads introduced enough high frequency distortion to make hi-hats sound rather crappy, especially at higher gains.
The OPA625 can drive 600Ω loads well. Wonder what your real problem is?
Can you post a schematic?
Are you trying "vintage" 600Ω gear? Can you try just a 620Ω resistor in parallel with some good modern load that works fine?
WHat is the DC voltage at the output? "600Ω" inputs some times have very-low DC resistances. Try a 100Ω resistor in series with the OPA625 output. In fact you almost always want some resistance between a hot amplifier and strange real-world loads.
The OPA625 can drive 600Ω loads well. Wonder what your real problem is?
Can you post a schematic?
Are you trying "vintage" 600Ω gear? Can you try just a 620Ω resistor in parallel with some good modern load that works fine?
WHat is the DC voltage at the output? "600Ω" inputs some times have very-low DC resistances. Try a 100Ω resistor in series with the OPA625 output. In fact you almost always want some resistance between a hot amplifier and strange real-world loads.
An op-amp compensated to be stable at a gain of 3 doesn't
have a hope in hell of working at x160 gain into low loads -
its simply running out of closed loop gain at high frequencies
for the feedback to work.
You need an uncompensated op-amp that can drive 600 ohm,
and compensate it for the gain range you typically use. Note
that if you try reducing gain below this range it will oscillate.
🙂 /sreten.
have a hope in hell of working at x160 gain into low loads -
its simply running out of closed loop gain at high frequencies
for the feedback to work.
You need an uncompensated op-amp that can drive 600 ohm,
and compensate it for the gain range you typically use. Note
that if you try reducing gain below this range it will oscillate.
🙂 /sreten.
I really appreaciate everyone's reply. I will upload a schematic as
soon as I get a screen capture of it. Most likely I will try the BUF634T included in the feedback loop as Per recommended.
Mic preamps are tough because you need a lot of gain especially
for dynamic mics and you need a very wide range as well for
different sources. While a snare drum may need only a 30dB
boost an acoustic guitar may need 50dB.
Thank You,
george
soon as I get a screen capture of it. Most likely I will try the BUF634T included in the feedback loop as Per recommended.
Mic preamps are tough because you need a lot of gain especially
for dynamic mics and you need a very wide range as well for
different sources. While a snare drum may need only a 30dB
boost an acoustic guitar may need 50dB.
Thank You,
george
If you want to go with adding a buffer fair enough.
But i've checked the OPA-627 specs :
Output current is good for +/-10V into 300R.
Its compensated for unity gain stability.
Maximum (low frequency) open loop gain is 120dB.
Open loop gain at 20kHz is 60dB.
So at 50dB gain you'll only have 10dB's worth of feedback at 20k,
16dB at 10K, you get 6dB more feedback for every octave you drop.
(And no feedback at 80kHz)
Adding a buffer in the feedback loop will not change this.
So I'm not sure it will fix the problem, but I don't know it won't.
The OPA-637 which is compensated for gains > 5 but otherwise
is the same chip (and the feedback loop should not return the
gain to 1 at high frequencies) basically has the same distortion
as the OPA-627 at five times higher gains.
If you used this at 50dB gain you'd have 24dB's worth of
feedback at 20kHz, 30dB at 10KHz which should be OK ?
It will be better. (No feedback at 320kHz)
🙂 /sreten.
But i've checked the OPA-627 specs :
Output current is good for +/-10V into 300R.
Its compensated for unity gain stability.
Maximum (low frequency) open loop gain is 120dB.
Open loop gain at 20kHz is 60dB.
So at 50dB gain you'll only have 10dB's worth of feedback at 20k,
16dB at 10K, you get 6dB more feedback for every octave you drop.
(And no feedback at 80kHz)
Adding a buffer in the feedback loop will not change this.
So I'm not sure it will fix the problem, but I don't know it won't.
The OPA-637 which is compensated for gains > 5 but otherwise
is the same chip (and the feedback loop should not return the
gain to 1 at high frequencies) basically has the same distortion
as the OPA-627 at five times higher gains.
If you used this at 50dB gain you'd have 24dB's worth of
feedback at 20kHz, 30dB at 10KHz which should be OK ?
It will be better. (No feedback at 320kHz)
🙂 /sreten.
- Status
- Not open for further replies.