Contour control- anyone ever use it?

[imaradiostar]

But, the impedance of the source affects the frequency response of the circuit, so you probably still need that CF.

The 4.7k resistor is the "source" impedance in the circuit. When driven from "perfect" low impedance you can twiddle the value of the 4.7k and it mostly effects the high frequency response. If you reduce the value of that resistor too much you lose the ability to roll of the highs at one end of the control.

I suppose a cathode or source follower is ideal because it'll give you a uniformly low source impedance. Scaling the control and driving it from the plate of a triode is likely to end in a circuit the performs differently depending on the health and specs of the triode- which could be frustrating. Perhaps you could call it a "feature" but I'm firmly in the camp of designing for readily available modern tubes and changing part value and design to alter sound rather than simply "tube rolling" to get "the sound." Build it right and you'll still hear differences in how different tubes distort but it won't be the result of wildly changing EQ.

Jamie

[J M Fahey]

Hi Jamie. You're right but to remember to add the internal impedance of the generator in series with those 4K7 you see.
The output impedance of a classic "X7" stage ends being around 47K, adding a CF you can lower it to a couple KOhms or even less, but still must be accounted for.
If you need to pad "hundreds of volts" and have low driving impedance to boot (not to mention saving a triode), you may simply follow the last triode with a 20:1 attenuator: .022uF/100K /4k7.
You will also save the contour input 4k7, because "you already have it "

[imaradiostar]

Thanks for the tip, J M.

So you're saying make a voltage divider off of the preceding tube- 22nF/100k from the previous tube to 4k7 to ground, input to contour control at the junction of the 100k and 4k7, omit the 4k7 in the contour control because it's no longer needed.

At first I didn't see the point but I just simulated it and I see what you're getting at. This could make for more selective control over the various elements and pad down the level substantially. Most times when I'm working with a high gain lead channel I prefer to have some sort of effect send after the distortion so I don't mind the level being down between 30 and 50 dB at that point. By playing with the value of the 22nF, 100k and 4k7 you can vary the EQ curve which is helpful.

Theory guys, please help me with this. The plate of a 12ax7 with a 100k looks more or less like a 40k source, right? So if my software has a "perfect" signal source with a near zero source impedance, I would have to add roughly 40k to whatever value I use at the front of this circuit. Am I doing this correctly?

If I am simulating it correctly I really like the way the EQ curve looks when I'm using 47nF, 50k at the top of the divider and 4.7k at the bottom. The output level is down between 25 and 45dB depending on control position and frequency. This assumes that I did it correctly- plate impedance

On the flip side- it increases part count in a minor way and I prefer a cathode follower on a high gain channel anyway. Given the low source impedance of a cathode follower (under 1k) one could easily sub a pot for the 4k7 resistor and mess with the value to achieve the proper balance of sparkle and treble rolloff at different knob positions. A 10k pot would work well for this and in simulation it seems to offer a decent range for an internally adjustable "presence" setting for the control.

Jamie