diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   Negative Feedback? (http://www.diyaudio.com/forums/tubes-valves/140076-negative-feedback.html)

Binaural 9th March 2009 04:22 AM

Negative Feedback?
 
2 Attachment(s)
Been fooling around with a Bogen AP 30, after I was able to fix the power supply. It actually doesn't sound bad. I bypassed the tone section, balance control, and volume pot (since I have a monitor controller, it's a bit redundant) and sounds even better now.

Another thing I wanted to fool around with was the negative feedback loop, experimenting with the resistor value.

I don't know much about it, but I've heard that it's used to stabilize, lower distortion, and increase gain of circuits. I've also heard that too much of it causes a very sterile "solid state" sound.

Anyways the schematic shows the NFB resistor as being 27k. The value in the unit was actually 19k. Now it would make sense to me that lower resistor values would allow for more negative feedback, correct? If this is true, i have had some very strange happenings. I started with a 33k resistor and moved all the way up to a 120k, and each time the amp grew in volume and seemed to really come alive. I figured the opposite would happen. Could anyone explain to me what's going on here? I'm lost as to why the circuit is behaving the way it is. The schematic is included for reference.

Binaural 9th March 2009 04:28 AM

2 Attachment(s)
Sorry, the resistors I'm referring to are circled in the schematic here.

Boris_The_Blade 9th March 2009 04:32 AM

Got your negative feedback ideas mixed up a bit. Adding more negative feedback reduces gain, reduces distortion, and if not done properly, causes instability. So when you increase the value of that feedback resistor, the feedback ratio decreases and the gain goes up - exactly what you are experiencing.

SY 9th March 2009 09:16 AM

What Boris said: you're increasing gain. "Coming alive" is the same thing as "turning up the volume."

AndreasS 9th March 2009 10:00 AM

Re: Negative Feedback?
 
Quote:

Originally posted by Binaural
Anyways the schematic shows the NFB resistor as being 27k. The value in the unit was actually 19k. Now it would make sense to me that lower resistor values would allow for more negative feedback, correct?
In this schematic you have negative voltage feedback from the secondary of the output transformer (R27) and positive current feedback from the cathodyne phase inverter (R26). Both reduces the output impedance. Changing (or removing) voltage feedback may affect stability of the circuit.

Regards Andreas

richwalters 9th March 2009 10:42 AM

Re: Negative Feedback?
 
Quote:

Originally posted by Binaural
quote:- < time the amp grew in volume and seemed to really come alive>.

Some years ago I did a fault find on a sim circuit; which had no fault on it. This type of circuit uses the resistance of 2 input tube heaters to set class A o/p stage cathode bias. Even the other channel is paralled. Replacing the two input tubes with other vendors types with slightly different heater current ratings can drastically effect emission. It's not my type of circuit arrangement.
As to nfb, I notice there is no phase correction cap on 1st amp stage anode (where nfb is connected to cathode). Reducing 27K to a lower value will increase suspectibility to HF instability, a square wave will easily show this up. By all means tweak it, if the amp makes a squealing noise when turned off or turned on with LS in circuit, instability is close.

richy

Binaural 9th March 2009 05:23 PM

Richy,
What is this phase correct cap you speak of? Would it be a useful addition to the circuit? Also about the 12ax7's heaters in the bias circuit, the original bogen tubes are in those positions, so this should hopefully not affect the bias.

You also say that the OP stage is biased for class A? Also, what would happen if i completely removed the negative feedback? Any ideas?

46davis 11th March 2009 01:41 PM

I think it's a dangerous idea to remove the negative feedback look entirely. It's there for a reason.
Without negative feedback, the drivers will have way too much gain for the finals and will push them into distortion and probably into grid current at high signal levels. Could damage your amp.
Negative feedback loops in the kind of amp you have usually have a capacitor as well as a resistor in them to tailor the feedback at different frequencies to compensate for the fact that no device has a perfectly linear frequency response. The gain of the driving stages is carefully matched to the finals for a certain level of NFB to give the best output level, lowest distortion and best stability at the most common drive levels. Good amps have very carefully designed NFB loops and it's hard to improve on those of good manufacturers. If you want to tweak your NFB, do so in small increments paying attention to HF stability and distortion.
I have had some good results in subjective sound quality by replacing the cheap factory R/C with premium parts, notably replacing ceramic caps with silver mica ones.
There are good treatments of NFB loop designs in the popular books on tube amp design going all the way back the Audio Cyclopedia.
Good luck. And be cautious.

richwalters 11th March 2009 04:01 PM

Quote:

Originally posted by Binaural
Richy,
What is this phase correct cap you speak of?
You also say that the OP stage is biased for class A? Also, what would happen if i completely removed the negative feedback? Any ideas?


You will need a oscilloscope and 1KHz square wave for optimising the phase comps cap on first stage. The value all depends on o/p tranny quality.
My guess is as no compensation caps are fitted in the amp, the front end triode HF roll off including the HF o/p tranny roll off (-3dB down point) is probably quite low, i.e 15KHz. In this case no small caps are fitted.
If you remove the global nfb, the loop gain will be at max. Ideal for guitar. In many amps typically 50mV for full o/p instead of 500mV (20dB nfb). Also the noise, microphony will be noticeable and in flighty amps, the amp might howl if LS is close by the 1st stage tube.
The annoying aspect without nfb loop is one hears everything in the house turning on/off, i.e fridges, light switches etc.

Failing experience and test equipment I would choose nfb resistor values +/- 25% from nominal. An amp should accept the variation. Ceramic caps have a tendancy to resonate, and any move to polyester or others is an advantage. The coupling caps to the output stage grids be a good quality type with decent voltage ratings.

Hope this helps

richy


It's worth keeping your eye on the value of the 12AX7 heater volts for the first two stages.

Binaural 12th March 2009 04:37 AM

Great, thanks for the info. I had some time over the last couple of days to really sit down and dig into this. What I did was take a scope, with a calibrated 1K sine input and hooked up a 50K linear potentiometer in place of the NFB resistor. I set it so the amp will start clipping at 1.5V input. Does this sound about right? I then took the value the pot was at and fitted a resistor of close value. Is the capacitor really necessary in the NFB loop? If so how would I go about finding an optimum value. What should I look for on the scope? Also how can you tell by looking at the schematic that this amp is a class A amp? This is one thing I haven't learned yet. I know the basic differences between the classes, but not how to identify them on a schematic (if possible). Thanks, everyone has been more than helpful to me!


All times are GMT. The time now is 10:44 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2