• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

12ax7 data with positive grid voltages

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
^ ^
| |

Never heard of any such characteristics. Don't see why this would be necessary as the 12AX7 is a small signal type (and one designed for operation at very low plate currents) not a power tube. If this has something to do with some sort of RF project, types like the 12AU7, 12AV7, 12AT7 were made for that, and also quasi-digital applications like multivibrators, one-shots, Schmidt Triggers, R-S latches. These types can operate with vgk > 0

The grid structure of the 'AX7 could be too delicate for that.
 
citizengain,

What kind of circuit are you designing that will use a 12AX7 to have positive grid voltages (grid more positive than the cathode voltage?).

It is one thing to have a few volts positive on the grid, and have a few k Ohms cathode self bias resistor. With proper design, the cathode voltage will be more positive than the positive grid voltage. That is OK.

The 12AX7 grid wires are extremely small diameter. Grid current will destroy the grid.
 
Thank you for the responses. I'm trying to figure out some guitar amp circuits. Eg. the Marshall 2203 preamp where the stage before the cathode follower is biased quite warm Rk = 820 Ra = 100 k).
In LTSpice I've been using the Koren and the Cardarilli model of the 12ax7. Reading this paper lead me to thinking the grid current more carefully:
http://recherche.ircam.fr/pub/dafx11/Papers/76_e.pdf
 
Thanks!
Now I know why you asked the question.
This is an electric Guitar application, not a Hi Fi application.

A. The electric guitar:
1. I have never measured the output voltage of even one electric guitar.
And, at even one or more strum levels, finger, pick, etc.
2. I have never measured the output impedance of even one electric guitar.
And I suspect that impedance is different for different frequencies.

Even if I had done all the above, there would be lots more measurements to do:
# 1. and 2. above all over again, with all the combinations of electric guitar switch settings, and potentiometer settings (all the controls that are on an electric guitar).

And, I have never seen a schematic that shows the pickup, switch(s), potentiometers, and whatever else might be in there (capacitors) of any electric guitar.

This makes the issue of how a 12AX7 in a guitar amplifier responds a very complex study, not just because of the 12AX7 circuit parameters, but also of the guitar.

B. The 12AX7:
I know of an ongoing project to measure the grid current of 12AX7 tubes under conditions such as low plate voltage.
One thing that was discovered by the ongoing study, is for the very large number of manufacturers of good tubes, there are lots of differences in the grid current with all those tubes in the same circuit.
When I say good tubes, I mean ones that are very good for Hi Fi applications, as well as a newly designed model of a 12AX7.
The many different constructions of those different manufacturers 12AX7 tubes, which work well in Hi Fi applications, respond differently when in a guitar amp input circuit (input stage).

I hope I was able to list some of the issues that you may, or may not have thought of.

Have fun with your study.
 
Last edited:
1. I have never measured the output voltage of even one electric guitar.

Several years ago I stuck a scope across the input jack of a DIY guitar amp which has the usual 47K resistor to ground, and a 1K grid stopper right into the grid of a 12AX7 which has its cathode connected to ground through a 1.5 K resistor bypassed with a 22 uF cap. I blasted it with every guitar I could find under every possible playing condition and combination of volume, tone and pickup selector looking for the largest possible available input signal.

Here is the scope shot. There is a 10X non Tektronix probe attached. It does not tell the scope to scale the readings, so the actual voltage is 10 X larger that want the screen displays, IE, 1.41 Volts peak to peak. An old Guild guitar with twin DiMarzio humbuckers, both on and up full, was the maximum voltage winner for a single plucked note. The usual single coil Squier Strat makes 0.7 volts P-P or so unless you just thrash on all 6 strings where you might hit a volt.

The grid voltage must obviously swing a bit positive, and nobody is blowing 12AX7's with a straight guitar. In fact I have never seen anybody blow a 12AX7 even with some seriously overdriven pedal boards with several VOLTS of output.

As to the actual curves of a 12AX7 under these conditions, I don't know, haven't looked. The Vacutrace doesn't do positive grid voltage steps. I'm building something that will, but it is on hold again.
 

Attachments

  • SingleNoteGuild_A.jpg
    SingleNoteGuild_A.jpg
    193.5 KB · Views: 138
Tubelab_com,

When I initially answered the question about 12AX7 grid current, I had no idea this was for a guitar amp.
The OP post # 1 did not indicate that.

Of course an electric guitar's output is not going to melt a 12AX7 grid.

But in a Hi Fi application, the 12AX7 was used as a phase invertor with the cathodes tied together, and a 3 or 4 mA current sink from the cathodes to a negative supply.
One grid was tied to the input connector, and a 47k Ohm to ground.
The other grid was tied directly to ground.
There was an instance when the filaments and cathodes were warm, and then B+ dropped out.
That means ALL of that 3 or 4mA from the current sink ALL went to the grounded grid.
Burnout time!

Just a warning to those who may try and draw a few mA of steady state DC current out of a 12AX7 grid.
Your mileage may vary.
 
Last edited:

PRR

Member
Joined 2003
Paid Member
Guitar Amplifier Overdrive, Ulrich Neumann, Malcolm Irving, 978-1329596658

But this is not a precision calculation because of the unpredictability of the signal and the listener. Usually a 1k resistor and a diode will tell all you can usefully know.
 
V20 . . . 40 triode sections?

If you ran 1 mA per grid, that is 40mA.
And then what do you get in the cathode, 40 x 3mA plate current, plus 40 x 1 mA grid current, . . . 160mA peak?

20 12AX7 at 12.6V = 3A filament current.

1600uMhos per cathode x 40 cathodes = 64,000 uMhos, now we might just be talking something here.

Have fun balancing 40 triode sections (or 20 per side in a push pull amp).
Can you tell if there is a "Sleeper" among the 40 sections?

Has anybody seen the schematic?
Has anybody actually heard the amplifier?
I have done neither.

Not my cup of tea.
Oh, that makes me thirsty.
Now I am going to have a cup of Twining's English Breakfast Tea.
I might even make it civilized too.
 
Last edited:
Thank you for the answers guys. I think I'll do the measurements myself at some point and see at what precision I'm able to get the data.
The variation between different tubes of the same model are huge indeed, as 6A3sUMMER mentioned. It is pretty evident even comparing the waveforms of different specimens.
 
I had no idea this was for a guitar amp.

I didn't either until he mentioned "Marshall."

Of course an electric guitar's output is not going to melt a 12AX7 grid.

I just posted my scope shot because the question of "how much output does an electric guitar have?" comes up quite often, even in guitar amp related threads.

The average output impedance of a guitar varies a lot with frequency, knob settings, and pickup design. Under everything up full conditions it can be from 10K to 50K at high frequencies (3 to 5 KHz). That's why cables do make a difference with some guitars.

As a kid I often made guitar amps out of old radio and TV parts. The schematic I had used a pentode (6SJ7) in grid leak bias (2 meg grid resistor). I found that you can indeed create blocking distortion in an input stage with a hot germanium guitar preamp. I would not however understand what was happening for several more years though.

Today's guitar pedal boards might indeed have enough output to blast the grid in a 12AX7, since many have a good sized opamp feeding the output jack.

the cathodes tied together, and a 3 or 4 mA current sink from the cathodes to a negative supply.

I have a driver board design that uses a CCS from the negative supply to the cathodes to the input LTP. I place a NE-2 neon lamp from the cathode to ground to clamp the maximum negative cathode voltage so that the H-K spec is not exceeded under cold start. I have never used a 12AX7 in the first stage, but have used the similar 5751. So far there have been no issues.


V20 . . . 40 triode sections?....Has anybody seen the schematic?

I have tried to create a "massively parallel" amp design twice, both times ended in failure, and we were only talking 12 tubes, 6 per side in a push pull amp.

The first experiment used 12 6AQ5 tubes each with it's own cathode resistor. Granted all of the tubes came from my box with several hundred used pulls of varying types and condition. I selected several 6 pair sets that worked well individually in a fixed bias amp with similar bias voltages. Each time I tried to make this amp work, one tube would eventually red plate and run away. I eventually gave up.

Several years later I went down a similar path with 13GB5 / XL500's. This time they were in fixed bias with a separate bias pot for each tube. Again current hogging was an issue, as was random spurious oscillation.

I think that regulated power supplies might have helped in both cases, since there was quite a bit of interaction between all the tubes.
 
Is such data available somewhere? Meaning the plate characteristics curves that would include also positive grid voltages. And also the grid current measurements.
Thanks in advance.

It may be interesting for you that I did some quite accurate triode LTspice models, especially regarding grid current.

More or less similar (in terms of mu) to what you request are the following of my models:
Sylvania 6948
RFT ECC81
electro-harmonics 12AT7EHG

You will find a zip of my models here:
http://adrianimmler.simplesite.com/440956786

The code of the last two models is also visible here:
Vacuum Tube SPICE Models

all the best, Adrian
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.