12ax7a cathode bias resistor variance discussion

[hewo]

By mere accident discovered 1st issued vc3112 during 1994 installed 820ohms halfwatt on both stages channel one.
Yet schematic assigns fender's 1.5k to both, with 10uf cathode bypassing.
What 820ohms does is reduce impedance?
Tube runs more hot, sound runs more hot?
That's because the tube is operating in higher idling (like higher rpm in an engine?)
Later motherboards have the legacy fender topology of 1.5k, the sound changes, to replicate fender's classic appeal.
Yet the slm designer 1st intended 820ohms for a good reason, he she may have been targeting vox signature, is this hypothesis viable?
What the 820ohms does is shift the preamplification upwards a bit such that the output waveforms will near or encroach tube saturation (clipping threshold).
Am I misconstrued?
So I swapped out 1995 board's to said 820ohms, and also took liberty to swap the 1k cathode bias of channel two's tertiary stage to said 820ohms. The 1994 boards don't use channel two 820ohms tertiary stage, it uses 1k.
The tension supplied resistor plate feed do not vary in either years of boards.
What's the general consensus regarding elevating the ecc83's throttle position? Will favortism be afforded to different frequencies and their coloration?
My own intuition tells me the low end gets favored, correct me on this.

[nickb]

By mere accident discovered 1st issued vc3112 during 1994 installed 820ohms halfwatt on both stages channel one.
Yet schematic assigns fender's 1.5k to both, with 10uf cathode bypassing.
What 820ohms does is reduce impedance?
Tube runs more hot, sound runs more hot?
That's because the tube is operating in higher idling (like higher rpm in an engine?)
Later motherboards have the legacy fender topology of 1.5k, the sound changes, to replicate fender's classic appeal.
Yet the slm designer 1st intended 820ohms for a good reason, he she may have been targeting vox signature, is this hypothesis viable?
What the 820ohms does is shift the preamplification upwards a bit such that the output waveforms will near or encroach tube saturation (clipping threshold).
Am I misconstrued?
So I swapped out 1995 board's to said 820ohms, and also took liberty to swap the 1k cathode bias of channel two's tertiary stage to said 820ohms. The 1994 boards don't use channel two 820ohms tertiary stage, it uses 1k.
The tension supplied resistor plate feed do not vary in either years of boards.
What's the general consensus regarding elevating the ecc83's throttle position? Will favortism be afforded to different frequencies and their coloration?
My own intuition tells me the low end gets favored, correct me on this.

Not having posted a schematic or even a manufacturer name makes it hard to be specific. Is this Crate we are talking about? Also, it's not clear to me which stages your are referring to. Better to say "V2 pin 2" and so on to be specific. I'm also guessing that English isn't your first language - not much we can do about that

So, generally speaking, using the 820 ohm resistor gives you about half the gain in the stage when compared to the 1.5K + 10uF. There is no significant effect of the frequency response in the range of interest. The change in the DC operating point is not going to make much difference unless the signal on the plate is very large.

[hewo]

halfing 12ax7 preamp gain, desirability

R3, R7, R18
Won best amp award year 1995
Saint Louis Music (SLM, gone, the designers)
Schematic resolution poor, attached.
Thanks nickb, and since gain depreciates using 820ohms, bandwidth gets enlarged somewhat for retaining frequency strengths that would have otherwise suffered. Dials would be upped a bit more, but without drowning out the upper range because of offending lower frequencies that plague power sufficiency. The ot has to be upgraded to encompass power sufficiency, as this is where the holdup usually occurs, correct me if I'm wrong.

[nickb]

R3, R7, R18
Thanks nickb, and since gain depreciates using 820ohms, bandwidth gets enlarged somewhat for retaining frequency strengths that would have otherwise suffered. Dials would be upped a bit more, but without drowning out the upper range because of offending lower frequencies that plague power sufficiency. The ot has to be upgraded to encompass power sufficiency, as this is where the holdup usually occurs, correct me if I'm wrong.

You're wrong. It makes no significant difference. The -3db point on the low end is only 10Hz with the 1.5k/10uf. Far too low to be of interest. The change on the high is similarly of no consequence. You do not need to change the OPT, not least since it will never see those frequencies. You would only need to change the OPT to get more power output at a low frequency and that is determined by the (final) power output stage.

PS: Sorry - I missed the SLM in the original post

[hewo]

Ok, so ears will not detect the 820ohms change?
Realize the gain is downed, so dials will be upped to regain what is downed.
Operating with dials increased will move these operating points rather margins to a different location, that which you say is of mere insignificance insofar as frequecy coverage strengths.

[nickb]

Ok, so ears will not detect the 820ohms change?
Realize the gain is downed, so dials will be upped to regain what is downed.
Operating with dials increased will move these operating points rather margins to a different location, that which you say is of mere insignificance insofar as frequecy coverage strengths.

The speaker doesn't reproduce those frequencies. Even if it didn't you would not hear the difference. The change in the DC operating point is not enough to affect any coloration introduced by these stages as the signal amplitude is fairly small - perhaps 2.4Vpp or so for full power out at V1 pin 1.

[J M Fahey]

Changes in frequency response and gain are very small, measurable but not much to human ear.

The main change when varying cathode resistor and keeping everything else the same (otherwise there's too many variables to explain in a simple way), is that the idle plate voltage changes.

The datasheet suggested values (100k plate + 1k5 cathode, etc.) in general try to have plate at 1/2 V+ , for maximum clean signal and symmetrical clipping.

Now the top waveform clipping is sharper (buzzier) and the bottom peak clipping is rounder, smoother, "sweeter" if you wish, so by varying idle point you favor one or the other.
And also create even distortion.

So varying cathode resistor is also a designer tool to tweak amp sound and feel.

Remember the famous JCM800 "cold biased" stage with unbypassed 10k cathode resistor, yet plate still 100k as an extreme example.

To be precise, 820r is "wrong" , 1k5 "right" (with 100k plate) , but the designer chooses one or the other for a specific reason, generally "by ear".

[Dave H]

in general try to have plate at 1/2 V+ , for maximum clean signal and symmetrical clipping.

Are you sure about that?

[hewo]

Scope shows no indifference, ears hands down do.
Blind test too.
Got both 94/95 factory models to compare.
Changing 95 into 94 confirms 820ohms superior vox signature. And dirty channel reminiscent "sgt pepper's" snazz, but that wasn't vox, right?

[frus]

Are you sure about that?

yeah, wasn't it something like 2/3rds of V+

[Malcolm Irving]

yeah, wasn't it something like 2/3rds of V+

Typical values for a 12ax7: B+ = 300V, Ra = 100k, Rk = 1.5k, Va = 200V, Ia = 1mA, Vgk = -1.5V
So 2/3 is a good estimate. (1/2 might be better for a pentode.)

[nickb]

Scope shows no indifference, ears hands down do.
Blind test too.
Got both 94/95 factory models to compare.
Changing 95 into 94 confirms 820ohms superior vox signature. And dirty channel reminiscent "sgt pepper's" snazz, but that wasn't vox, right?

Huh? This doesn't make sense. Well, to me at least, but that wouldn't surprise anyone

You seem to me saying in the first sentence that you can neither see nor hear any difference. Yet, in the last sentence you claim to hear a difference. So its not clear what you mean.

I would suggest you need to look at the frequency spectrum rather that volts/time of a the 'scope using a variety of different level test signals to really examine it. You would have to have a gross difference to spot it with the guitar.

I think we all agree that the difference in frequency response is negligible so that just leave non-linear effects due to the change in DC levels and voltage swing.

The first stage has an output of 1.2Vpp for 10mV rms in (from the notes). I would also accept that if you easily get 200mV from your pickup giving us 24Vpp for the 1.5K and 12Vpp for the 820 ohm. These are probably big enough that the coloration due to the inherent distortion might be audible by a sensitive pair of ears. Particularly so if this clean channel is then turned up enough to clip the power amp. On the other hand, if the guitar and gains are turned down then the difference would be inaudible. BTW, the output swing on the second stage is unchanged for the same output level. In other words it all depends how you drive it.

At the end of the day if it suits you then it's good.

[Dave H]

Changing 95 into 94 confirms 820ohms superior vox signature. And dirty channel reminiscent "sgt pepper's" snazz, but that wasn't vox, right?

"As for amplification, the Beatles had at their disposal in 1967 a Fender Showman and a Bassman head with a 2x12 cabinet, a Selmer Thunderbird Twin 50 MkII, a Vox Conqueror and the Vox UL730, 7120 and 4120 bass amp used on Revolver."

There's more to recording a guitar sound than an amplifier with an 820R cathode resistor, never underestimate the studio trickery involved.

guide-recording-equipment-and-instruments-featured-beatles-sgt-peppers-lonely-hearts-club-band

[Dave H]

Scope shows no indifference, ears hands down do.
Blind test too.

Nick, I read that as - There's no difference visible on the scope but there's a clear audible difference (even in a blind test).

[nickb]

Nick, I read that as - There's no difference visible on the scope but there's a clear audible difference (even in a blind test).

Thx DH. I agree. On a second reading that is what is meant. A combination of punctuation and phrasing threw me off, I guess.

I remain skeptical on the difference as there are just too many other variables. I would only trust a statistically significant properly conducted ABX test.