enlighten me please

[iknowjohnny]

I'm just curious why i'm getting the result i see by lowering the cathode R on the first stage of my high gain build. I have no bypass cap on this stage, and from all the results i have seen so far in tweaking these stages and trying everything from a 820R up to a 10K cathode R, i always see a DROP in gain as i lower the value. Well, i just tried something i saw on a bad cat schematic....No resistor at all....just a straight wire to ground from the cathode. The gain went DOWN, and thinking this was because it was a dead short to ground, and value of R should then take the gain back up. I figured 820R seemed to be the most gain, so what if i used a 470R. would that be even more. So i tried it and still less gain than 820. So it seems like at some point between 820 and zero ohms to ground, the gain turns around and goes down instead of up as you decrease resistance. Maybe i'm spacing here, but can someone explain whats going on?

And that said, my goal has not been to get more gain, but less. And this has been great because all other methods of gain reduction so far have been ok up to a certain point, but once the gain gets down to where i want it the tone suffers. (even a split plate load method) Using a 470 here has allowed me to lower it w/o that loss of tone. I don't get it?! It should be screaming with a 470R, but instead it got me a lower amount of gain just like i'd been shooting for w/o the loss of tone. Somebody enlighten me please. I'm glad, but i'm also confused.

[Phil_S]

My theoretical knowledge on this topic is weak, so give me a bit of latitude.

As you change Rk, you should also change Ra. As you tweak, check for both cathode voltage and plate voltage. You will see there is a relationship.

I suspect, as you lower Rk and keep Ra constant, you are allowing Va to rise. This probably explains what is happening to your amp. It Rk that biases the tube, and you are not compensating appropriately as you fiddle with it.

[Andy Le Blanc]

the cathode resistor creates bias at the grid.... and the bias in relation to the
plate voltage can move the operating point well into non-linear territory....
which can result in a lose of gain...... the plate voltage needs to be adjusted
to keep the operating point of the valve at the most linear portions of its
transfer curves........ you can view it with a scope...... if both tops of a
sign-wave begin to clip at the same time the tube is at a optimum operating
point...... if the distortion is a-symetrical it shows either too much or too
little bias.......

placing the cathode at ground potential is called zero-bias operation ......
you still need to provide a fixed bias, a battery bias, or use a cap and large value
resistor to use grid leak to make a bias potential.....

[Andy Le Blanc]

your gonna have to use transfer curves to solve for optimum gain, plate load,
and plate voltage.... for your application....... get a tube cad.... it beats the
heck out of trial and error....

[drz400]

your gonna have to use transfer curves to solve for optimum gain, plate load,
and plate voltage.... for your application....... get a tube cad.... it beats the
heck out of trial and error....

I second that and according to the creator is is more accurate than spice

[iknowjohnny]

As i said earlier, i WANT less gain, not more. So this is a good thing. I don't have a scope nor do i know what tube cad is. I was just curious as to why the gain starting with the cathode going straight to ground and gradually adding resistance goes up, then at some point goes down again. Seems somewhere between 470R and 820R is the optimum gain with a 220k plate R. Going either way, be it up or down results in less gain.

[Jana]

You really do need to get a scope if you're going to be tinkering the way you are, it really helps to see what is going on.

Measure the voltages, see if you find some clues.

My guess is this, when you lower the cathode resistor you are changing the bias. The bias is going more positive which makes the tube conduct more. With a 220k plate, the increased conduction lowers the plate voltage quite a bit which results in less gain. Just a guess.

If you want less gain, try using a smaller plate resistor.

[Andy Le Blanc]

look on the web .... if nothing else.... for available RCA receiving tube manuals
there are many out there in PDF..... in the back is a chapter called
"Resistance-Coupled Amplifiers"..... that contains a few small schematics
and accompining charts for various tube types with componant values......
voltages and volts gain... they hold the info your looking for...... less gain
is found by lowering the supply volts and/or reducing the value of your plate
resistor..... the charts give a range of each...... and an optimum value of the
cathode resistor......

[iknowjohnny]

Thing is, i have 220k plates because that was one of the things that made for a richer different sort of tone that i was looking for and could never quite nail with 100k plates. thats why i use 220k's now. I've tried everything you can imagine to reduce gain and everything works to do just that, but they also all kill the tone along with it. Dropping the cathodes to a dead short to ground or a very small R seems to reduce gain w/o killing the harmonic complexity, which seems to be the first thing to go when reducing gain the other ways. granted, i'm a complete bozo when it comes to understanding the theory behind all this. But my ears are very educated and i can tell you that at the moment the tone is pretty effing amazing. I have yet to hear any other high gain amp as complex, and i've owned a boatload of marshalls and a smattering of everything else since 1971. But being a perfectionist and knowing perfection is just a few value changes away, i can't help but try and nail every detail i hear in my head. this is the closest i've ever been.

I will admit that there are 2 details that really made this amp jump another rung on the ladder of tone more than most anything else i have tried, and both i copped from a bad cat schematic ! (one of course being the small cathode R) Funny tho....i played a bad cat about a year ago and didn't care for it at all.

[iknowjohnny]

I've learned over the months i've been building and tweaking not to speak too soon...well, i done good lately till now ! Yeah, i should have tested this more because now i'm finding the feel isn't quite right and there is a hiss that wasn't there before. Wonder how BC gets away with this.

[Jana]

The badcat gets away with it because of the .1uf cap on the input along with the 1 meg to ground. What's it called, contact bias? You could do that if you put that cap in there to prevent your guitar of anything else plugged into the input from affecting the grid DC.

[iknowjohnny]

I see. Thanks.

[Andy Le Blanc]

its called zero-bias operation...... its a form of grid leak bias......
the cap and a large value resistor.... up to 10 or 15 meg. ......
an approach left over from the earliest days of radio.... all there was, were diodes
and the heater was powered by battery..... a fleming valve..... for the
detection of radio waves..... Edison first noted the diode in his experiments to
perfect the light bulb but never thought it was useful so never sought a patent
when Lee De Forest stuck a grid in the tube he got an amp..... anyway......
everything was done with battery..... it was found that the battery that supplied
the grid bias could be replaced with a capacitor..... a potential accumulated
upon the grid ..... because if the blocking action of the cap......
the leak is the high value resistor, either parallel with the
cap, or to ground to prevent the accumulation of electrons on the grid from
becomeing large enough to shut the the valve off.....

[drz400]

As i said earlier, i WANT less gain, not more. So this is a good thing. I don't have a scope nor do i know what tube cad is. I was just curious as to why the gain starting with the cathode going straight to ground and gradually adding resistance goes up, then at some point goes down again. Seems somewhere between 470R and 820R is the optimum gain with a 220k plate R. Going either way, be it up or down results in less gain.

You are probably getting into a non linear range of the tube and running into cut off on part of the signal. You can damage the tubes if you dont run them where they want to be, you can play with values on the plate and cathode to a point but for lowering gain try a different tube or adjust the pads between stages and run the tube where it is designed to run, learn about bias Tube CAD has a "make it correct" button, it is a $40 program that is worth every penny.
http://www.tubesandmore.com/

[jjman]

I'm wondering if some people are treating the word "gain" differently than others. Strictly speaking, gain is not the same thing as clipping or distortion. Although "gain" is slang for such.

"Gain" is the amplification factor. If stage A provides "more gain" that otherwise, it is putting out a stronger signal than otherwise. Whether it's distorted is a separate question and it’s easier to think of it as not clipping. However, it’s stronger-than-otherwise gain is more capable of “overdriving” the next stage into clipping/distortion. So stage A’s “more gain” puts out a stronger signal which (over) drives the input of stage A+1 and manifests as distortion on the plate (and cathode) of stage A+1.

Not an important distinction to a player but very important to a builder or tinkerer.