Dropping String Values VS Plate Voltages

[dorrisant]

When tweaking a clone amp (ODS for argument sake) which is more important? Dropping string values or plate voltages? Is it a mix of the two, with diminishing results at a certain point? At what point? I know... get the proper PT and this is not an issue. But lock that in place and what are your real options?

The reason I ask is that when building, I run into this issue. The plate voltages for the preamp may not be right at all with the given (installed) dropping string values. In order to get the plate voltages to the right level, sometimes I have to lower a dropping string resistor to a very low value, maybe down to 1K. I can't remember where I read that 1K is about as low as you should go, otherwise the isolation between stages starts to diminish. Probably from Merlin.

What are your thoughts on this? What happens when you've lowered the string values down to 1K (or whatever value they've been dropped to)? What happens if you go too low? What happens if you go to high and what would you consider the upper limit, if there is one?

If there is a fundamental idea that I'm missing, feel free to enlighten me. Maybe a general rule of thumb, or a calculated formula.

I have built clones before that had vastly different string values in order to get the plates (and related) to what I consider the right voltages. I'd like to have a grasp on what is changing, besides the plate voltage. For example, if the string value is supposed to be 22k and I end up at 1k, what has this changed as far as feel and response of the amp?

[bepone]

adjust plate voltages correct and dont worry. if you get oscillations through power suppy line ie motorboating you need to increase values. also if you get humm like result.
PI can still be run under low filtered B+ but preamps not anymore.. so lower value R acceptable until PI , after you need to put at least 5k

[Phil_S]

I've found the need to adjust the B+ rail to suit. I built an Express variant in 2007 or 2008 using a pair of 6AQ5's. Naturally, the PT was not as high voltage due to the power tube choice. I wanted to get the preamp running at voltages as close to typical as possible. I recall doing some estimates of load to help me guess at appropriate choices. With not too much trial and error, I finally got close enough. I checked my files, but all I seem to have are photos. The dropping string was done with those reddish resistors, which, it turns out are low contrast. I can't see what values I used. I wish I had made a schematic, but it appears that didn't happen. Anyhow, I wouldn't be too concerned. If you find the decoupling effect is not adequate, you can always rework it. I don't think a few volts will make that much difference.

[Stevem]

If your able to stick to no more then 2 gain stages per filter node you should have a pretty good leway with things.

[martin manning]

When tweaking a clone amp (ODS for argument sake) which is more important? Dropping string values or plate voltages? Is it a mix of the two, with diminishing results at a certain point?

Keeping dropping resistors at 1k or more is a good rule of thumb for 22u filters. Lower R means higher RC corner frequency, and reduced attenuation of audio frequencies from adjacent nodes. For an ODS, assuming you have too much B+, I would keep the 22k and 2k2, and adjust the first dropping resistor after the choke to get the V1 plate voltages where you want them (180-190V). If you are short on B+, you can minimize the first dropping resistor to get the PI voltage as high as possible, and then reduce the 22k to bring the V1 voltages up. In other words, live with a lower PI voltage, but get the V1 and V2 voltages right.

[R.G.]

When tweaking a clone amp (ODS for argument sake) which is more important? Dropping string values or plate voltages? Is it a mix of the two, with diminishing results at a certain point? At what point? I know... get the proper PT and this is not an issue. But lock that in place and what are your real options?

They're both important, at least as far as they have to be enough for the situation.

Dropping string resistor values have to be enough so that with the bypass caps every second stage or so there is not enough feedback through the power supply line to get oscillation, or even just conditional stability. They're also contribute to the plate voltage on following tubes not being too high for the tube. One possible effect of changing series string resistors so you get close to but not over the edge of oscillation is that it will make the amp's time response a little "slurred" as notes excite the almost-ringing, which alters the audible tone.

To a certain extent, rejection of 120hz power supply ripple is affected by the RC filtering of the series R and shunt C between stages. For a given frequency, like 120Hz, you can thumbnail the ripple rolloff per RC section as |Xc|/( |Xc| + R ). Xc = 1/(2*pi*f*C), so for a 10K resistor and a 10uF cap, the reduction in 120Hz is a factor of 133/(133+100000) = 0.0131. That's a bit over -37db if I did the math right. It is then -43db at 240, etc, getting bigger as the harmonic of the power line goes up. A couple of these interstages accumulates a LOT of power supply ripple rejection. But between any two interstage decouplers, you only get one stage of reduction and that has to be bigger than the gain between the two stages to keep you out of maybe-oscillation mode.

Plate voltages are important in that they have to be high enough to get the desired voltage swing on the plate. The smaller the plate signal swing, the lower the plate voltage can be for a given amount of distortion. This directly contributes to the "feel" of the amp because it's a way to push the amp to being as linear as it can be, or banging into cutoff, or dropping into squashy "saturation", or all of the above.

[dorrisant]

Thank you guys... that gives me a very good direction. I have always tried to adjust the string to get the playes as close as close as possible to published plate voltages. I have never worried that much about the difference in string values and will continue to slightly disregard.

You guys are gold!