Is 1 Meg ohm REALLY the most for a grid leak resistor?

[anson]

Hi everybody,

I was wondering is anybody knows if the recommended grid to ground resistance of 1 mega ohm for a 12ax7 can be exceeded with out problems. I've read in a few places that using a higher value might lead to oscillations, but I was wondering if any of you can back this up with experience. I would like to try using a higher value grid resistor because according to this web site......


http://www.ozvalveamps.elands.com/pickups.htm


.... piezo electric pickups seems lose allot of bass response when the grid resistor values are below 2 meg or so. So, I was thinking of boosting the value of the grid resistor in my amp for that reason. Any thoughts?
Anson

[MWJB]

Harp players regularly use 5-10meg in typical cathode biased preamp in puts. Try 5m and see what you think? If you go too large the amp might start to sound more overdriven and lose some fidelity.

[Tubis]

the datasheets usually says 2.2M but a lot of limiting values are exceeded all the time in tube amps so.....

[Alex/Tubewonder]

High values above ca 2M are used when the tube is grid leak biased i.e. no cathode resistor. Usually in the 5-20M range. Yes, they do it in harp amps.

[MWJB]

Just to clarify (sorry Alex, some could get the impression that larger values are only used in grid leak circuits), there are currently no grid leak biased harp amps being manufactured in any kind of scale, that I'm aware of (though grid leak bias was the norm up to the mid 50's, so many vintage amps obviously still have this arrangement).

All the well known, currently manufactured, harp amps are cathode biased regarding the preamp. Many use around 5meg for the input load. Some old grid leak biased Fenders only used 1meg for the grid load...there aren't any schematics to support this however.

Ampegs also often had 5.6Meg inputs with cathode biased preamps.

[Alex/Tubewonder]

Thanks for the clarification. I may know a bit about tube electronics but harp amps are sort of terra incognita for me. A day when one doesn't learn something is a day wasted. Thanks again.

[Enzo]

Part of the confusion I think comes from people calling the grid resistor a "grid leak" resistor even when it is not a grid leak bias arrangement. In real grid leak bias, the resistor can be much larger than 1 meg.

[Alex/Tubewonder]

Part of the confusion I think comes from people calling the grid resistor a "grid leak" resistor even when it is not a grid leak bias arrangement. In real grid leak bias, the resistor can be much larger than 1 meg.

Well, splitting hairs here, but a grid resistor or any dc path from grid to ground is the path for grid leak current, therefore it is by definition a grid leak path.
The secondary effect that we most often look at is this same resistors loading on signal source, be it a guitar pickup or a preceding gain stage.

In DC coupled amp grid leak current is sucked up by preceding stage.

As I said, splitting hairs

[Enzo]

Yeah well, only hair I got is on my chin.

When words have more than one sense, I try not to use them. It's just me, but I reserve the term for grounded cathode stages.

For example I switched over to the term "mains" instead of "AC line" because some get confused thinking I refer to a line level something.

[cbarrow7625]

The information offered here is all good. But the one thing I don't think was mentioned was the noise penalty that you may pay by using a large "grid load" (?? is that a better term than grid leak?) resistor. All resistors (and pieces of wire) are antenna, they pick up noise & other radiation. Large resistors are usually better antenna (especially true in this arrangement). Not only that, but there is also the inherent resistor noise added plus the inherent tube noise (that will generally increase with larger grid resistors).

The best way to get a high input impedance and not pay the noise penalty is to use a bootstrapped input. Bootstrapping raises the effective input impedance by a factor of 10-20 (or more) times depending on how it is constructed but allows a smaller resistor to be used (thereby not violating tube maximum limits).

Yes, the gain is lower (or possibly only 1, depending on your design) and you do have to use an input capacitor (although this is a requirement in grid leak arrangements as well) but you get a relatively quiet stage with a very high input impedance.

A good example is a phase inverter. Both the LTP & cathodyne style PI use bootstrapping. As with the LTP, the gain of the stage can be somewhat controlled by the lower cathode resistor. THe cathode bias is set by the upper resistor in the usual manner.

Didn't want to leave this option out of the discussion.

[MattT]

FWIW, the Dumble ODS has an FET input that is just for piezos. Perhaps a pedal/acoustic preamp of some sort?

[anson]

Thanks evrybody. Also I just wanted to mention my additional interest in the large input impedance because it would be useful for daisy chaining the inputs of several amplifiers together with less loss as a result of multiple grid resistors (one for each amps) in parallel.



All that info is really helpful, and now I have no inhibitions about giving a bigger grid resistor a try in my amp. As far as the boot strapping is concerned, I want to be clear in my understanding of how it works. So I think it's a circuit where the grid resistor and cathode resistor are brought together, as they normally would in a typical 12ax7 preamp stage, but instead of being connected to ground directly, that junction is connected to ground via some other resistor (the "tail" resistor of a LTP). Is that right? And I was curious if anyone knows how to calculate the bootstrapped input impedance of a circuit. At Aiken amps he briefly described Bootstrapping for a LTP circuit, but I don't remember anything quantitive.
Anson

[raiken]

The problem with using too high a "grid leak" resistor is that any grid current that flows in the tube will develop a voltage across the grid resistor, which will in turn affect the bias of the tube and may cause other effects.

In a typical grid-leak bias stage, you will notice the cathode is grounded (no cathode resistor), because all the bias voltage is developed by the grid current flowing through the grid-to-ground resistor. This method of biasing is not as stable as cathode biasing, because it doesn't have the inherent negative current feedback at the cathode to stabilize the operating point.

If you use a large grid resistor in conjunction with cathode biasing, you will likely find that you need to lower the value of the cathode resistor to compensate for the grid voltage that is developed across the grid resistor.

You will also need to use an input coupling capacitor to isolate the guitar volume control potentiometer from the DC voltage developed across the large grid resistor, otherwise you will get scratchiness when you rotate the volume pot, as well as a bias that will vary with changes in guitar volume or pickup type.

Lastly, you will probably find that the input stage is very sensitive to tube variations, because the amount of grid current varies widely in different tubes, even from the same manufacturer. In fact, you can sometimes even get enough grid current with a 1Meg or lower grid resistor to develop a high enough voltage to cause scratchiness on a volume pot. I have had to change out preamp tubes of several occasions because they developed excessive grid voltage in a circuit.

Randall Aiken

[MWJB]

"If you use a large grid resistor in conjunction with cathode biasing, you will likely find that you need to lower the value of the cathode resistor to compensate for the grid voltage that is developed across the grid resistor." lowering the value of the cathode resistor will change the tone of the amp and may even exacerbate problems that have been introduced by the oversized grid load resistor, such as the amp hesitating or choking when hit with a big signal.

"You will also need to use an input coupling capacitor to isolate the guitar volume control potentiometer from the DC voltage developed across the large grid resistor, otherwise you will get scratchiness when you rotate the volume pot, as well as a bias that will vary with changes in guitar volume or pickup type." Possibly, but I've never encountered these symptoms with loads of between 5-10Meg, the input coupling cap may also adversely affect tone.