Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: BlackCrowe604 on August 16, 2021, 04:11:06 am
-
Hi All, I hope everyone is doing well in these new times....
I came across a '69 Princeton (non-reverb) and have been messing around with it.
The tremolo is weak and the bias was super hot (PV = 361V @ 38mA @ one side, and PV = 362V @ 42mA other side), leading me to believe the output tubes are not closely matched.
I have a new set of current production Tung-Sol 6V6, and have now found myself in a rabbit hole...
The OEM spec sheet (old) says 12W max plate dissipation, but all charts I have in books state 14w PD for 6V6s, so which is it?
The OEM spec sheet (old) also says max plate voltage is 315V. Is 360V up to 400V normal on the plates of 6V6s? Were higher PVs typically used in guitar amp circuits?
To wake up the tremolo I plan to change the 1M oscillator resistor to 470K.
I plugged in the new matched 6V6s I have and measured 35ma @370V on each which is still too high, I am thinking this should be closer to 25ma and that the bias needs to be cooled.
To cool the bias, I plan to change the bias voltage divider resistor from 22k to 27k or 30k as I have heard a cooler bias wakes up the tremolo as well.
Appreciate any input. Am I on the right track here?
I know a lot of members here have Princetons and have built Princetons so any tips appreciated!
Thanks!
-
The OEM spec sheet (old) says 12W max plate dissipation, but all charts I have in books state 14w PD for 6V6s, so which is it?
Both. Google "Design Center," "Design Maximum," and "Absolute Maximum" ratings systems.
The OEM spec sheet (old) also says max plate voltage is 315V. ...
Some of them have a figure up to 385v. Clearly, the tubes could handle more than was claimed on the data sheets.
... Is 360V up to 400V normal on the plates of 6V6s? Were higher PVs typically used in guitar amp circuits?
Yes, my 1967 Princeton Reverb had a 6V6 plate voltage of around 420vdc.
The tremolo is weak and the bias was super hot (... 38mA @ one side, ... 42mA other side) ... I am thinking this should be closer to 25ma and that the bias needs to be cooled.
... I have heard a cooler bias wakes up the tremolo as well.any tips appreciated!
My Princeton Reverb was biased for ~21mA per 6V6. The tremolo was very prominent; yours will be too when you get the idle current closer to 20mA.
-
You're on the right track. Get the bias right first. You may want to add this simple bias pot mod. See attachment.
Once you are happy with the bias, your tremolo will likely be stronger too, but if still too weak for your liking, I suggest replacing the 3.3K/25µF cathode components with a red LED (cathode to ground). This will give you very strong tremolo regardless of your bias setting.
-
I will second what others have said. Hot bias will neutralize all the bias goodness. Sluckey’s bias mod suggestion with potentiometer is the same scenario as a Deluxe Reverb bias scheme. Can reference that if that helps. It’s a great mod. I did it on my recent Princeton Reverb build.
I did a 470K in place of the 1Meg in the oscillator. Actually it’s 433K (measured low). I wouldn’t go any lower. That and .02 caps in bias circuit get it super swampy. Almost swallows all the attack of the note with intensity above 7.
-
FWIW, I have read that factory bias settings for the Princeton Reverb was about 55%. Cold bias to make the trem work.
-
FWIW, I have read that factory bias settings for the Princeton Reverb was about 55%…
That seems unlikely, how could anyone know?
From the schematic g2 voltage of 400V, and the bias voltage of -34V, the 6V6 triode chart indicates an anode current of about 30mA.
With an anode voltage of 410V, that works out to 12.3W idle dissipation, 102.5% of the 6V6 design centre limiting value.
It seems a bit whacky to me, but that’s what the numbers indicate :dontknow:
https://el34world.com/charts/Schematics/files/Fender/Fender_princeton_reverb_aa1164.pdf
https://tubedata.altanatubes.com.br/sheets/093/6/6V6GT.pdf
-
Thanks all! Interesting discussion.
Seems "design centre" was a nominal value once upon a time.
Thanks for the bias mod suggestion, I don't have any pots on hand, but have a resistor sub box, so I will dial in what's good and then put a fixed resistor in for now.
Appreciate the insight, seems the numbers are a bit wacky and I might not be as crazy as I thought!
-
You're on the right track. Get the bias right first. You may want to add this simple bias pot mod. See attachment.
Once you are happy with the bias, your tremolo will likely be stronger too, but if still too weak for your liking, I suggest replacing the 3.3K/25µF cathode components with a red LED (cathode to ground). This will give you very strong tremolo regardless of your bias setting.
Thanks Sluckey! Yes, original bias voltage resistor was 22k (even though schematic states 27k). I replaced with 30k (actual measured @ 28k) and it cooled the bias to 20-21mA and the tremolo is back to life, much stronger response! Also, the original power tubes had 5-6mA of mismatch between the pair while in use.
I haven't seen an LED used in this type of manner, any quick explanation? Seems interesting.
I will go ahead and experiment w/ dropping the 1M oscillator resistor value and change the middle .01uf cap to .02uf and see what happens to the speed & intensity.
-
Read the last paragraph on this page...
http://www.valvewizard.co.uk/trem1.html
Although the wizard did not directly say so, the increased gain from LED biasing will greatly increase the tremolo signal, resulting in a much stronger tremolo effect, especially on bias wiggle circuits such as yours. Give it a try. The results are much stronger tremolo than you will get by messing with that 1M resistor.
-
... I suggest replacing the 3.3K/25µF cathode components with a red LED ...
... I haven't seen an LED used in this type of manner, any quick explanation? ...
Sluckey's link says the trem goes down to 3Hz. But the bypass cap is 25µF, which has a reactance of 2.1kΩ at 3Hz, which is not a great bypass at that frequency. More like <300Ω would be a good bypass for the 3.3kΩ cathode resistor.
A "poor bypass" means signal current will create negative feedback across the cathode resistor, lowering the gain.
The LED establishes a steady voltage drop as long as it's forward-biased/lit, and maintains that voltage down to d.c., which is 0Hz. Therefore, the LED bias ensures same-gain from the oscillator all the way down to 0Hz, which is perfect for a trem oscillator.
-
Sluckey's link says the trem goes down to 3Hz. But the bypass cap is 25µF, which has a reactance of 2.1kΩ at 3Hz, which is not a great bypass at that frequency. More like <300Ω would be a good bypass for the 3.3kΩ cathode resistor.
What value bypass cap would get you down to <300 ohm? Just curious, I like using the LED. Nice touch on the front panel.
-
What value bypass cap would get you down to <300 ohm? Just curious, I like using the LED. Nice touch on the front panel.
Using XC = 1/2πFC
177µF will give XC of 300Ω @ 3Hz.
-
Thanks. :icon_biggrin:
-
...What value bypass cap would get you down to <300 ohm? Just curious, I like using the LED. Nice touch on the front panel.
>"25µF, which has a reactance of 2.1kΩ at 3Hz, which is not a great bypass at that frequency. More like <300Ω would be a good bypass"
If 25u is 2100Ω at 3Hz, then 250uFd is 210Ω at 30 3Hz.
So 300Ω at 30hz would be a little less than 250uFd, maybe 220uFd, but hardly worth close figuring.
The other thing is that the grid R-C and the cathode R-C interact while bias is established, at start-up and at any sudden change of rate or load. Not a big deal, but the LED is real darn constant.
-
...then 250uFd is 210Ω at 30Hz.
My calculator says 250µF is 212Ω at 3Hz.
-
I don't use tremolo much, but the idea of an LED to bias it intrigues me.