This sounds like a version of a conjuntive filter. Only that's a cap and resistor in series.
I know where you're coming from, but remember "filters" have to have something that's frequency-dependent. So this is similar in being applied to the OT primary, but different.
As Silvergun said, mostly this thing is just an attenuator.
Think of it like a cross-line master volume at the outputs of a phase inverter. Basically, + and - voltage signals are being combined to cancel each other out to some degree. I don't like Twistyneck's non-tech explanations of what's going on, but this is what he's talking about when he saying the tubes are fighting each other. Regardless, smaller resistor equals more + added to -, and less output voltage at the speaker so less volume.
The question I have for everyone thinking about doing this is that if a Hotplate (or some other external attenuator) isn't the magic bullet for you, why would this mod be better? If all this did was lower volume, we could say this mod is on equal footing with all other forms of attenuation.
Look at the chart attached. Along with reducing volume by allowing more voltage cancellation, the resistor across the OT primary changes the load impedance that each tube sees. Rather than try to analyze a specific amp/resistor combination, look at the trend that occurs as the resistor is made smaller (more attenuation).
If you look at the gridlines on the graph, they all wind up merging together around a plate voltage of 30v and below. Gridlines also bunch together and are more closely-spaced for a given change in gird voltage at low current (look how close the -20v and -25v lines are compared to the -10v and -15v). Where grid lines bunch together, you will get distortion. If the lines are bunched together (distorted) on only one end of the loadline, the distortion will be predominately even-harmonic; if the lines bunch together at both ends, there will be increased odd-harmonic distortion.
We can conclude that for lower values of load impedance, odd harmonic distoriton will be reduced and even harmonics will predominate.Which seems like a good thing until you realize that push-pull operation cancels even harmonics generated in the output stage.
Look at the graph and note the dashed curve: this corresponds to voltage and current which results in 25w of plate dissipation. If the loadline strays above this curve, plate dissipation rating is being exceeded. If a small portion of the line goes above the curve, you might not notice redplating because it is a small percentage of the total signal cycle.
However, notice that as we keep reducing the load to the tubes by using this mod, pretty much all of the resulting load lines stay above 25w dissipation most of the time. This is why you get redplating, and why EVH had to use a variac to attempt to reduce the voltages in the amp. The lines drawn assume a 450v B+, and is why all lines converge at 450v, 0mA. If you reduce B+, these curves get shifted to the left to the new value of B+.
If we argue that the variac-reduced voltages as a result of redplating were a key part of the "brown sound" we could get that more safely with power scaling, VVR or some such technique. If we also assert that slamming a 100w amp with all controls at 10 is also key, we could reduce the volume if needed by using more conventional attenuation between the speaker and OT secondary.
Again, I'd argue we don't know
for certain (maybe someone more knowledgable can say) EVH had to use attenuation in the studio, because volume from a 100w amp is generally not an issue.
I'll generally agree with those that would say nothing sounds like a 100w on 10 than a 100w on 10, even after trying every attenuator trick out there. This is why chasing the sound of a recording will usually leave you frustrated unless you are also making a recording and can duplicate the conditions of the sound you're chasing.
Topic: EVH mod ?????????? (Read 11469 times)