Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: firemedic on November 21, 2010, 11:09:21 pm
-
I have looked all over this forum to see if anyone has shown a concrete solution for NFB/presence in an amp with a paraphase inverter. I see the 5C3 has a simple NFB loop feeding back to the control grid of a power tube.
I did try a switched circuit back to the point between the cathode and the cathode resistor/bypass cap of the last gain stage before the PI but with any series resistor less than 100k, which seems to just shut off the NFB, it makes a hell of a racket. The presence pot is a 22k since I had one laying around, shunting a 0.1 cap to ground. Maybe the OT primaries are reversed. Would that be a good place for NFB to be applied? Options?
-
The 22k pot is too high for the 100k feedback resistor's value. That's only a 4:1 ratio! Typical presence ratio is a 100k feedback resistor w/ a 5k pot which is a 20:1 ratio. If you ground out the pi input there's no feedback generated so you see how close you are to negating the use of the control for your presence pot porposes?
-
Check the feedback arrangement of the SD 80. This schem was posted on the SLO-clone forum. Since I have never had an SD 80 I cannot verify that it is correct.
Cheers Stephan
-
So sorry: I got it backwards! :huh: I'm only leaving this here in case someone else can take this idea and flip it around.
Would this work as a Presence Resonance control with a paraphrase PI? (
I'm not sure about the value of the pot or the resistor below it except that the total of the two values should be equal to the stock 1 meg NFB resistor. I do not know what the value of the cap bypassing a portion of the NFB resistor should be either. Someone who really understands the paraphrase PI would have to do the math for the proper cutoff frequency.
Hope this helps,
Chip
-
That's an interesting idea. It shouldn't be too hard to experiment with the cap values and I like that it's already an existing circuit. Whenever I try to do :huh:math for this stuff I am always wrong anyway and get a headache to boot. I will try this arrangement. Thanks chip....
-
Except one thing: the grid leak(?) resistor on the 5C3 is 1MR, and so is the NFB resistor. On the 5E9A driver (which I'm working with) the value is 220k. Maybe I should make the total NFB resistance match this 220k resistor. Or does it matter?
-
Except one thing: the grid leak(?) resistor on the 5C3 is 1MR, and so is the NFB resistor. On the 5E9A driver (which I'm working with) the value is 220k. Maybe I should make the total NFB resistance match this 220k resistor. Or does it matter?
First off - the drawing I posted was a mistake. It's a resonance control, not presence which is what you asked about.
Secondly - you never mentioned the 5E9-A Tremolux circuit.
http://www.el34world.com/charts/Schematics/files/fender/TREMOLUX_5E9A.pdf
Third - YES, the resistance values matter. Negative feedback is always a relatively small sample of the output signal recycled back to an earlier part of the circuit where the phase is opposite. You get the sample using a voltage divider. Wikipedia will show you the formula for a voltage divider. A presence control typically alters the ratio of the voltage divider above a certain frequency so that less of the higher frequencies are cancelled out by the NFB. (A resonance control does the opposite, so less of lower frequencies are cancelled out.) I don't understand this PI well enough to tell you what the NFB ratio would be with a given NFB resistor value.
Fourth - what is your goal? To clean up the sound of the 5E9-A Tremolux with negative feedback, to have a presence-type control, or both? No matter what the answer is, why don't you find a circuit with the features you want and not get so hung up on the type of phase inverter?
Just for giggles, I went searching for a Fender circuit with all of the features you seem to want:
self-balancing paraphrase phase inverter
negative feedback
presence control
The Fender Amp Field Guide (http://www.thevintagesound.com/ffg/) is very useful for this kind of adventure. Take a look at the Wide Panel "Pro" 5D5 (http://www.thevintagesound.com/ffg/schem/pro_5d5_schem.gif). Look familiar? It's got a 1 meg negative feedback resistor, and the power amp looks close to your Tremolux despite the 6L6 power tubes.
The Narrow Panel Pro 5E5-A (http://www.thevintagesound.com/ffg/schem/pro_5e5-a_schem.gif) has a presence control which should work with some value adjustments. 100K NFB resistor with 5K pot => try a 50K pot with your 1 meg NFB resistor.
Since the NFB is being inserted back into a PI circuit I truly don't understand, I can't tell you exactly what would happen but these look like reasonable starting points.
BTW the 5E5-A is one of my favorite old Fender circuits. It's just so damned squirrely!
Cheers,
Chip
-
WHICH "paraphase"? The word gets used on a lot of things. The details are in the devil.
> back to the point between the cathode and the cathode resistor/bypass cap of the last gain stage before the PI
If it were phased correctly for NFB at the PI, going one stage left is inverse phase which is wrong.
You can maybe get enuff NFB to reduce speaker flubbing with a few hundred ohms in one of the two para cathodes, and resistor from the speaker.
You can not get the FULL effect of the 5F6-A Presence network with anything other than that longtail. It's doing several screwy things which change the distortion spectrum. I suspect when not used to the max, the distortion change is more important than the treble boost (which can be had a million other ways).
-
I believe the 5E9a has a self-balancing paraphase inverter.
My goal is to clean up the sound a little at high volumes. It's not farty or oscillating, just muddy & harsh. At low volume (up to 6) this amp is SWEET and I don't want to mess with a good thing. I figure a little NFB may help with the muddiness/harshness at high volume.
In fact I am using 6L6s so the 5D5 looks quite compatible with what I've got.
Since I'm doing the mod, I want to be able to adjust it easily. It will be switchable NFB. If a resonance control is easier to do with the NFB on the most compatible circuit (5D5) that's okay. I'm easy. So even if you mistakenly posted what I wasn't asking for, it still helped me out a lot. Sometimes us newbies aren't sure what we want. That's why this forum is so popular.
-
Okay so I did install the NFB loop, inserted it into the cathode of the 2nd gain stage and removed the cathode bypass cap so it would work. The presence control also works fine except but when turned all the way up,max shunting of highs to ground, there is a parasitic oscillation. Could that be a ground loop? Also when the one-tube reverb is turned up above halfway, "normal" feedback develops. This was tested with a shielded signal from a cd player so it's not the instrument and it's not a microphonic reverb tube, I switched it out with no change.
Other than that I'm pretty pleased.