|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
64.122.72.31
In Reply to: RE: Eliminating the Line Stage Output Cap posted by Triode_Kingdom on August 07, 2017 at 19:44:53
That preamp was prone to sounding different on account of the chip used in the servo, which was not really sorted out.
The goal is to have the circuit immune to the opamp(s) used.
We had to solve the same issue when we developed our MP-1 preamp back in the late 1980s. The MP-1 has a direct-coupled balanced output.
The thing to do is to hang a capacitor off the output of the servo so as to give you additional filtering. Experience has shown that you want to give it a nice long timing constant; IOW its going to be an electolytic.
Another point: you can integrate the response of the opamp by using a capacitance in the feedback loop of the opamp. So you don't really need the low pass filter as a separate circuit, although it won't hurt.
Follow Ups:
"The goal is to have the circuit immune to the opamp(s) used."
I'm aware of that being a potential issue, but no experience with the previous work of others, such as Berning. In the design I'm working on, the output of the servo amp (opamp) will contain virtually no signal energy due to the 2nd order low pass that precedes it. So, I see the issue as being one of determining only the extent to which the grid signal might be modified by the output Z of the servo. The resistor between the servo amp and grid is planned to be approximately 10 meg. This should provide considerable isolation relative to the grid circuit impedance, which is essentially set by the volume pot. It's not clear to me what would be gained under these conditions by bypassing the output of the servo. Do you have a circuit diagram for the output section of the Berning?
--------------------------
Buy Chinese. Bury freedom.
I don't, but FWIW your circuit is nearly identical.
The second order filter is great, and IME you don't need such a large grid bias resistor to make it work. Bypassing the output of the IC is effective though, if you are worried about 'solid state contamination' though :)
You are also better off with really slow timing constants in the servo. You really are only compensating for drift in the tube once the whole thing warms up. So there is no need for the servo to have any speed to it.
I recommend an additional circuit that lights an LED or the like, in case the tube is unresponsive to the servo. Very handy for detecting a bad tube and you don't want DC at the output in any event.
Bypassing the output of the IC is effective though, if you are worried about 'solid state contamination' though :)
Sorry, I dismissed this too soon. I've come full-circle now after researching the possibility of using resistive optocouplers for better isolation. A simple two-section filter (RC-RC) at the output of the last opamp will increase isolation to 65dB at 20 Hz and nearly 100dB at 100 Hz. That will absolutely render the servo inaudible.
You are also better off with really slow timing constants in the servo. You really are only compensating for drift in the tube once the whole thing warms up. So there is no need for the servo to have any speed to it.
The servo loop is planned to consist of two active stages. The first will be a 1 Hz shelving filter with its non-inverting input attached to the cathode. No feedback resistance or capacitance will be attached to that pin. This will ensure the amplifier presents a totally benign, high-Z resistive load to the cathode. The output of the first stage will drive both the servo amplifier - which further filters and inverts - and a uPC1237 speaker protection IC.
I recommend an additional circuit that lights an LED or the like, in case the tube is unresponsive to the servo. Very handy for detecting a bad tube and you don't want DC at the output in any event.
The uPC1237 provides a large number of features in a small package, including startup delay, positive and negative DC offset detection, latching option, power-off disconnect and relay driver. I'll add a gain stage ahead of it so smaller than normal offset voltages will be detected. A LED wired to the bottom of the relay coil can act as a fault alert.
Funny, I didn't realize how many of your ideas I was chasing down until I went back and read your post. Thanks for the good advice!
--------------------------
Buy Chinese. Bury freedom.
You might consider that you can do everything you want here except the protection relay with a single opamp like a TLO82, which is cheap and not really high performance, and yet still be effective and entirely inaudible.
First, I think you will find that you don't need all that much gain to do the job so you won't get into offset or stability issues with the opamp; a gain of 100 will be **plenty**! One opamp can sort out your correction bias voltage, the 2nd can be used to drive the LED.
I don't filter the LED signal as much; just a simple 6 db slope. In this way, at higher volumes the LED flashes with the music; it gives me an indication that I'm driving the preamp at a higher level. I use bi-directional LEDs.
You should be able to have no more offset then about 0.001 volts or so, which won't harm any amp made. It might even be lower than most cap-coupled preamps can do (unless they don't use feedback).
If you pull this all off, I think you will find this to be one of the most transparent preamp circuits you've ever heard. Its a big deal getting rid of that output coupling cap!
"Its a big deal getting rid of that output coupling cap!"
We're definitely on the same page on that point. I'm expecting this to provide an output Z of a few hundred ohms, and there won't be any phase shift or coloration from a large series cap.
About the opamp, I already have a few OPA445s on hand for this project. This device is rated to +/-45V, and that's what I need to handle the bias voltage directly. The followers will be 5686s in pseudo-triode. With this tube, the DCR of the cathode choke (~2,200 ohms) dictates a negative supply of roughly 24-28V at the bottom of the choke. With the cathode at zero, that will produce about ~12mA current into the cathode. To keep the cathode at zero, I'll need about -18V to -20V on the grid. So, I believe the simplest plan will be a +/-28V supply using a LM317/LM337 pair. That will provide the negative voltage for the tube, as well as powering all the active components and the output relays. Now, if I can just find time to start on the chassis...
--------------------------
Buy Chinese. Bury freedom.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: