No Grid Leak on Input Stage??

[KevinOConnor]

Hi Guys

The use of a ferrite bead at the input of a tube amp is a bad one. Grid-stops work much better and are less expensive and have no anomalous distortion effects in the audio band (unless you use carbon resistors - always an exception to every rule... hehe). Mesa is notorious for not using grid-stops and have unstable layouts as a result.

The grid current of a typical 12A_7 is so infinitesimally small that any grid-leak up to 33M is entirely safe. Ampeg didn't put those 5M6s in their amps with their fingers crossed - they knew what they were doing. Ampeg needed a high-z input to match the transducer in the peg of the upright basses they wanted to amplify, hence the basis for their company name: "amplified peg".

In most gain stages, the cathode is tied to ground through a resistor and this is as "grounded" as it needs to be for the discussion above. Rk then provides the grid-cathode bias voltage based on the cathode current through Rk. The grid-leak carries basically zero current and is tethered via the grid leak to the bottom of Rk. As long as the tube can generate a bias voltage between its grid and cathode it will be happy and maintain a stable idle current.

For proper grounding, it is best to have a grid-leak close to the rest of the circuitry for the given tube stage. You should never rely on the guitar cord and parts within the guitar to provide this important current path.

Since guitar pickup damping was mentioned, you can over-damp the pickup and sacrifice a bit of "sparkle" and then have a level pot on the guitar that has much less tone change over its sweep. Most guitar pots provide "critical" damping, which makes the tone vary with the pot sweep as the cable and amp inut are isolated from the pickup one way then tied directly at '10'. In the over-damped situation, a low-value pot loads the pickup all the time and the presence of the cable and amp input makes little difference to the tone.

Have fun

[Mike Sulzer]

No, it is far better to have a standard (1 meg) input impedance, especially when especially since you can not do much by making it bigger, as I explained earlier.

Also, attempting to alter the sound of an amp by changing the pickup resonance will have different effects with different guitars since the pickups vary significantly in resonant frequency. An amp is an amp; a guitar is a guitar. They meet with a 1 meg impedance.

The input resistor, in combination with your guitar's pots, affects the sharpness and width of the resonant peak of the pickups. The higher the value, the more narrow and larger the peak.

I measure every pot I buy (which is hundreds a year) and label them, then, if necessary, use a resistor across them to achieve the exact response I want given the tone of the instrument in question and the pickup.

If you raise the value of the grid leak, it's like raising the value of the pots on the guitar, but only for the amp in question. So if the amp seemed a little dull (or a little sharp) across multiple guitars, changing that value might be all the tailoring one needs without other circuit changes, or as part of a recipe of other changes.

[KevinOConnor]

Hi Guys

Mike, I don't see why you think this is a debate?

For most players and purposes there is no reason to change from 1M.The whole point of the thread is actually that there MUST be a grid-leak of SOME value. Period.

The one time it is useful is as I already stated: capturing the sound from a raw piezo pickup. With such a pickup, 1M loads it heavily and there is no bass at all; with 10M+ there is full tone capture.

Otherwise, there is no disadvantage or harm in using a >1M value.

As far as using the grid-leak to tweak the guitar tone, for most players with standard-wired guitars it makes no difference unless one goes low with the grid-leak resistance. Where it would make a difference is for a unique player who has no pots in his guitar.

As far as the preamp tubes red-plating goes - it never happens and not because of the plate resistor. In tests I did around 1990, preamp tube current did not become unmanageable without a grid reference unless the plate voltage was excessive. In a real amp where there is a plate resistor, there is even less chance of the tube over-heating. It's just not something to lose sleep over.

Have fun

[Mike Sulzer]

Hi Guys

Mike, I don't see why you think this is a debate?

As far as the preamp tubes red-plating goes - it never happens and not because of the plate resistor. In tests I did around 1990, preamp tube current did not become unmanageable without a grid reference unless the plate voltage was excessive. In a real amp where there is a plate resistor, there is even less chance of the tube over-heating. It's just not something to lose sleep over.

Have fun

I guess it is not a debate; you are now agreeing with nearly everything I wrote.

The exception is what limits the current in a preamp tube. Despite what you write, you can find curves (for example here: http://www.drtube.com/datasheets/12ax7eh-ehx.pdf) that show that you can exceed the allowed plate dissipation in a 12AX7. And these curves appear to continue straight off the page. It sure looks as if you can put 300 volts across a 12AX7, no plate resistor, grid cathode voltage set to zero, and exceed the allowed dissipation by a lot. So it is the plate resistor that limits the dissipation.

[Mike Sulzer]

Hi Guys

The use of a ferrite bead at the input of a tube amp is a bad one. Grid-stops work much better and are less expensive and have no anomalous distortion effects in the audio band (unless you use carbon resistors - always an exception to every rule... hehe).

How does a ferrite bead cause audio distortion on a high impedance input? Let's consider a case where ferrite can cause distortion: on the output of a class D amplifier. Figure 9 of this reference: https://hal.inria.fr/hal-01103584/document shows distortion for various cases. One case is open circuit, that is, no load, essentially zero current. In this case the distortion is very low.

[wizard333]

Hi Guys

Since guitar pickup damping was mentioned, you can over-damp the pickup and sacrifice a bit of "sparkle" and then have a level pot on the guitar that has much less tone change over its sweep. Most guitar pots provide "critical" damping, which makes the tone vary with the pot sweep as the cable and amp inut are isolated from the pickup one way then tied directly at '10'. In the over-damped situation, a low-value pot loads the pickup all the time and the presence of the cable and amp input makes little difference to the tone.

Have fun

As far as using the grid-leak to tweak the guitar tone, for most players with standard-wired guitars it makes no difference unless one goes low with the grid-leak resistance. Where it would make a difference is for a unique player who has no pots in his guitar.

If you do that you aren't losing just "sparkle", you are losing the articulation of the pick attack and if you are a player with a fast picking style, that is the kiss of death.

That is also why a larger grid leak for tonal purposes could be useful. It's the equivalent of raising the value of your volume pot, since they are in parallel as far as your pickups are concerned, but at the amp instead of the guitar. That DOES have an effect, I tweak pot values all the time to achieve desired tonal effects from a guitar, not infrequently I wire them so that each pickup "sees" a different pot value, this is no different. I do this kind of thing at least 100 times a year, if you think you can't hear a difference, I encourage you to compare the tone/response of a 400k load vs. 500k, or 330k vs. 550k, etc. It's not a subtle difference if your high end hearing is intact.

In terms of having the pot work without massive tonal change over its range, the best solution is a 150k resistor bypassed by a .001uf cap across the input and wiper of the volume pot. Works fine and doesn't clamp your high end detail like using a small pot or grid leak.

The rest of that is very interesting; thanks for the post!

[pdf64]

Apologies for the red herring I put forward regarding a tube amp's input stage potentially red plating should it lose its grid dc reference.
I had in the back of my mind an occasion when a preamp tube red plated in those circumstances, however it was a 12AT7 reverb driver (with a transformer coupled load), clearly a very different arrangement than a regular input stage.

[Mike Sulzer]

Apologies for the red herring I put forward regarding a tube amp's input stage potentially red plating should it lose its grid dc reference.
I had in the back of my mind an occasion when a preamp tube red plated in those circumstances, however it was a 12AT7 reverb driver (with a transformer coupled load), clearly a very different arrangement than a regular input stage.

Transformer coupled would do it! And thanks for the experimental confirmation that preamp tubes can draw too much current and overheat.

[KevinOConnor]

Hi Guys

The fender reverb driver operates at nearly the full plate voltage and is a small power amp with little resistance other than rp to restrict current. Overall the tube can dissipate 5W without red-plating but that is almost where it idles here. The tube manufacturer's 300V rating is for transformer loading, and Fender is exceeding that in every amp I've seen of theirs.No surprises the tube goes wild without a grid-leak - so does any power tube with transformer loading.

Have fun