TS9 LED clippers... jump in volume

[madboy]

Hey everyone! I did search the archives, but I couldn't find what I was looking for... it's sometimes difficult to know which key phrases to use. Anyway, I have a TS9 with a switch between the 1n914s and LEDs, with the LEDs being significantly louder. Is there a simple way to equalize the volume between the two settings? Or would I lose the clarity that the LED brings? Thanks in advance!

[thelonious]

I've always just turned down the volume knob on the pedal when I switched to higher forward voltage clippers like LEDs. I guess you could do something complicated like use a 3PDT toggle as the clipping switch and have it switch in a voltage divider that drops the volume at the same time... but why go to that trouble when you could just turn down the volume knob?   

[GibsonGM]

+1   You wouldn't want to lose the 'goodness' that the LEDs are bringing to the table, would ya?  ) 

[MrStab]

i'd do what Tristan suggested: use a spare set of poles on a switch to go between unaffected output level on the 1n914s, and a voltage divider to ground (to simulate a volume pot turned down a little) for the LEDs. or even better, a trimpot. it's what i'd do, at least. in my experience, you only need a SPDT to switch between diodes, so a DPDT could do the job.

i'm thinking this would be best on the effect output - you could maybe put it somewhere else, but then you'd risk affecting the operation of the rest of the circuit.

[madboy]

Alright... I will see if I have a spare switch in my parts bin... thanks for the help everyone!

[GibsonGM]

One suggestion:   Try it out by wiring it up loose (breadboard...) before you make it permanent. You might not like it.   If you find a loss of highs that you don't like, try bridging the 1st resistor with a cap (try 1n, 10n and other values to taste).

[Pojo]

Could quite possibly be overlooking an unwelcome interaction with the tone control here, but what about going the other way and have it add resistance to the feedback loop of the 2nd opamp when the 914's are selected to add some gain at the output?

Either way, attenuating or boosting, just a dpdt toggle should do the trick (3 poles not needed).

[teemuk]

Different forward voltages for clipping. The one with higher sounds louder because clipping happens at much higher amplitude level.

You need to figure out which effect you ultimately want from using different diodes: 1) to capture their different knee characteristics or 2) to exploit their different forward voltages.

Because from that depends where and how you attenuate.

In scheme 1 you want to "normalize" the signal level to match different clipping thresholds. Silicon starts to clip around 500 mV, LEDs have plenty of variation but for the sake of example lets say that clipping could be expected at about 1V. In such case you wish to drive the LED circuit with approximately twice lower signal level. You can either switch in an external attenuator before the gain stage, or switch in resistors that decrease the stage gain. Either way works. Post compensation is not needed in this setup. You may find that the difference between diodes this way isn't possibly even great enough to warrant a switching feature.

Scheme 2 is practically the opposite because it deliberately exploit different forward conduction thresholds to clip a fixed signal level less or more, depending on Vf. In such case you obviously need to hit the clipping diodes with fixed signal level and the attenuation to compensate perceived loudness differences must happen post clipping. Since signal gain doesn't really change, only parts of the signal are clipped away, such gain compensation may be difficult to realize adequately in practice.

[MrStab]

Could quite possibly be overlooking an unwelcome interaction with the tone control here

as the 4k7 resistor and the 47nF cap before it pretty much end up at vbias or ground, and are in series, it's possible to just switch both and change the cap proportionately so the high-pass filter has the same cut-off frequency as it would have had with the stock values. so if you were to (roughly) halve the resistor to 2.2k, you'd double the cap to 100n (example values). you could do this with just a SPDT. there are probably a million posts explaining this, but it's good to keep it in the public eye! er, ear...

i haven't done enough reading on the TS tone stage to comment on how that could be affected, though, as i have a tendency to replace it altogether these days.

[PRR]

You can series-string 0.6V diodes until you get a threshold similar to your 1.7V LEDs. Same curvature at higher output.

[ashcat_lt]

You can series-string 0.6V diodes until you get a threshold similar to your 1.7V LEDs. Same curvature at higher output.

And I'd be willing to bet you won't hear enough difference between the doubled diodes and the LED to be worth the trouble.  In fact, you could probably just leave it with the silicon diodes and cut the opamp gain to somewhere between half and a third and get much closer than close enough to the "LED sound" without the extra output.

[PRR]

> won't hear enough difference between the doubled diodes and the LED

I suggested it mainly because it is a 20-cent experiment. Anybody interested can use his ears without going broke.

Most of the usual models for LEDS are exactly the same as for simple diodes, with a higher multiplier (we need a taller "cliff" for electrons to fall-off and splatter photons). And relatively more series resistance but I don't think the usual clippers ever get close to the zone where that would matter.

If you like less-abrupt, try 10 Ohms to 500 Ohms in series with whatever diode. That "softens", though soft clipping is higher output so you may run into hard op-amp clipping if you beat it hard.