Preamp Voltage in Fender Amps

[Toxophilite]

Various versions of Fender amps have different preamp voltages on the 12AX7 plates. It seems to vary from 170 up to 230 I think


I have an amp (not the ones I've been building lately but another build ) with a fender tone stack on a 12ax7 and it's only getting about 140 vdc on the 12AX7 plates, I can easily adjust to raise this amount to the more standard 180 vdc found in the AB763 circuits, however..




I'm curious what differing voltages do


Is it simply a matter of higher voltage  = more gain in the preamp??

[PRR]

> 170 up to 230 I think

To 380V on some.

> higher voltage  = more gain in the preamp??

No. Supply voltage has very little effect on voltage gain. Read the R-C tables for 180V and 300V.
http://www.tubebooks.org/tubedata/HB-3/Receiving_Tubes_Part_1/Resistance-Coupled_Amplifiers.PDF

Assume typical Fender values: 100K plate resistor "Rp", 220K load "Rg".

At 180V -- V.G. = 47
At 300V -- V.G. = 52

Not even 11% difference. This will be nearly inaudible in your pick-hand, and can easily be matched with a slight turn of the Volume knob.

The difference is in overload and color. In typical Fender topology, mostly in the first stage.

At 180V -- V.G. = 47, Eo = 25V, maximum clean input = 0.53V
At 300V -- V.G. = 52, Eo = 57V, maximum clean input = 1.09V

Guitars sometimes touch 0.5V but rarely touch 1V.

At maximum "clean" input your distortion is around 5%. At half of that level your distortion is about half, like 2.5%.

Even when not overloading, the 5% is perceptable on a naked steel string (or chord), the 2.5% not-really. So if you want a trace of "sweetness" before overload, use *lower* supply voltage on preampery; if you want dead-clean on very straong signals go toward 400V.

[Toxophilite]

Wow
Interesting stuff
thanks for the thorough answer, I think I obviously need to do a little more reading on the subject


What I'm understanding is higher preamp voltages  are better for dead clean sounds ( more my normal are of playing)
whilst the lower supply voltages are better for transitional disortion qualities(sweetness)


Do higher voltages sometimes make for a noisier preamp circuit?

[SoundmasterG]

Wow
Interesting stuff
thanks for the thorough answer, I think I obviously need to do a little more reading on the subject


What I'm understanding is higher preamp voltages  are better for dead clean sounds ( more my normal are of playing)
whilst the lower supply voltages are better for transitional disortion qualities(sweetness)


Do higher voltages sometimes make for a noisier preamp circuit?

Any time you increase gain, the noise will go up too. So even though the gain difference is only 11% in PRR's example, the noise will likely increase around the same amount.


Playing around with preamp voltages can do a LOT to get your amp to sound the way you want....get the stages to overdrive the way you want to make a sweet sounding amp at all volumes.


Greg

[tompagan123]

PRR's comment about plate voltage versus headroom is interesting.  I assume that the input voltage continues to increase at every successive stage.  Wouldn't it be cool to be able to scale the plate voltage (headroom) at each stage?   Does anyone bother with this on Fender style circuits?-Tom

[shooter]

scale the plate voltage (headroom) at each stage

I do my "scaling " at the grid with voltage dividers, not sure what the masses do.

[jjasilli]

@ tompagan123: "I assume that the input voltage continues to increase at every successive stage."



As shooter says, the gain in signal can be attenuated to deliver a large or a small input signal to the next gain stage, at the choice of the designer (within the operating parameters of the tube).  A 12ax7 has an operating gain factor of about 60. 1V in becomes 60V out; then 360V; then 2160V!!! Also, a 12ax7 wants to see a max of only about a 2V input (4V peak -to- peak).  Clearly, attenuation of output signal is required. 

@ tompagan123: "Wouldn't it be cool to be able to scale the plate voltage (headroom) at each stage?"


The voltage is always scaled at each stage.  That is, the design of the Power Supply is pre-planned to deliver a specific Supply Voltage at each stage.  Next comes the Loadline. This is determined by the value of the plate resistor in conjunction with the current draw of the tube.  This sets the Plate Voltage.  The Loadline is plotted as a straight, slanted line across the tube chart.  Then an Operating Point point must be selected on the Loadline. Bias must be set to achieve the Operating Point.  As AC signal voltage swings, there may be totally clean headroom; or compression and saturation; or cut-off.  All this is pre-planned scientifically.


@ shooter:  "I do my 'scaling ' at the grid with voltage dividers, not sure what the masses do."  Adjusting the amount of swing of the signal voltage at the tube's input is good; but, again, it's only one factor in the operation of the tube. 


So, you can change: the Supply Voltage; plate resistor value & plate voltage; bias & operating point; and/or input signal strength.  All of these are "co-dependent".

[PRR]

> I assume that the input voltage continues to increase at every successive stage.

Guitar amps can be funny. It may be instructive to compute the signal level at each stage, input levels of 20mV 200mV and 1V, gain knobs adjusted to bring the output stage to just-clipping.

It is often true that the lowest level in a g-amp is at the volume pot wiper. The AA-Champ, when not dimed, can have the 1st and 3rd stages working hard and the 2nd stage hardly working; indeed the 2nd stage is never strained until the 3rd stage is all flat-lined. The distortion profile reflects this 1st+3rd topology.

There are designer preferences. Fender tends to alternate gains and losses. Some 1970s Ampegs took a good bit of gain in front, a lot of loss in the middle, and a lot of gain at the end.

[tompagan123]

Great discussion.

Shooter, you said the following:

"I do my "scaling " at the grid with voltage dividers, not sure what the masses do."


To make sure  I understand - Are you referring to a voltage divider on the plate to scale down the AC output signal to the next stage, or are you referring to scaling down the DC voltage on the plate?  I assume it's the former, correct?

[shooter]

Yup, the former.   When I 1st started I couldn't figure out why starting at like V1B's output I'd always have a square wave, The smart ones here explained n told me to read stuff
They also pointed out you can split the plate R up and take your signal from the middle of 2 plate R's, which you don't see done much.
My brain still stores the correct data but when I access it there are a lot more checksum errors, so I could be way off

[Toxophilite]

Hey
Does anyone know why the ab763 deluxe has 180 vdc on the pates of the normal channel preamp and only 170 vdc on the vibrato channel?

[sluckey]

In my mind, 170v = 180v. It's only a 6% increase. Readings on the attached schematic are +20%. However, I see 170v on both channels...

    http://el34world.com/charts/Schematics/files/fender/Fender_Deluxe-AB763-schematic.pdf

[VMS]

I think toxo meant the reverb version:

http://el34world.com/charts/Schematics/files/fender/Fender_deluxe_reverb_ab763_schem.pdf

[sluckey]

I think toxo meant the reverb version:

Probably so. But...

[Toxophilite]

Ha I didn't notice the +/- 20%, like old caps
Yes I did mean the reverb version
I was curious if there was a design reasoning? Why have them different on the schematic? all else being roughly the same

[sluckey]

There are plenty of mysteries and outright mistakes on the old Fender layouts/schematics.

[Toxophilite]

Ah the mysterious wisdom of the ancients!
Who knows what the strange markings in their crumbling manuscripts mean?


I guess it could've even been a cut and paste sort (not actually cut and paste of course) of mistake by a draftsman
Okay, curiousity satisfied
Thanks

[tompagan123]

For what it's worth, I just increased the final series dropping resistor on a fender style build bringing the preamps from the 192 range to the 175-180 range and it made a nice difference in touch sensitivity and added a little more harmonic complexity into the overall mix.  Probably due to the increase in compression down the line. 
A related question.   That final dropping resistor went from a 18K metal film to a 27K 2W carbon comp.   Does anyone have any opinions on resistor type for large series dropping PS resistors?   Like metal oxide, metal film, carbon comp, wirewound etc?     Just curious..

[jjasilli]

This is a recurrent issue.  Some like the dependability of modern resistors; they don't tend to fail or drift as much and are less noisy.  Some prefer carbon comp for old time's sake.  Some claim to hear a difference.  Others say hogwash.  I'm with Dave Funk's Tube Amp Workbook:  use carbon comps where the vintage amps used them.  With a vintage amp topology, that's all CC's.  With a more complex modern amp there's more places to put modern resistors.  YMMV!

[PRR]

In the Reverb-- it is the same 2-triode bottle, with the same 100K plate resistor both sides, and the cathode resistor is *shared* on the 2nd stage (though separate on the 1st stage). All coming off the one B+ line.

If the tube is sides-matched and the 100K are matched, it must be the SAME voltage on both plates.

+/-10% on the 100K resistors will cause ~~+/-5% difference worst-case, which more than covers 170V-180V.

The tube-halves were made on the same day, with mostly the same machinery and workers, but the halves never come out the SAME, and +/-20% variation is entirely expected.

You'd think the note-tester might have noted the circuits are identical and could have fudged the observed numbers to be the-same. But then again I just missed a delivery because I can't even get my phone number right, so I can't sneer at number mix-ups.

The numbers on the 12AT7 driver can't be test-best either. The side with the larger resistor has the smaller plate-resistor drop.

Take off your good glasses, do not use digital meters or mirror-scales. Tube plate should sit about yo-high on a cheap needle meter, +/- a finger or two, and that's good enough to rock.