Not sure how this analogue circuit works... exactly

[chuckstarck]

I need to add the following level control circuit to an existing device and am wondering how it works...

**broken link removed**

Here we have audio in two places, one going through Q1301, the other above on a buss called signal as well.

The issue here is how to control the level of that signal in both places. Basically, I"m adding components in the circles.

My understanding of the circuit. C1310 and C1219 perform somewhat of a shunt to both signals via diodes 1301 and 1201 WHEN they are forward biased. The biasing voltage is created via the voltage divider at the R 1239 Level Control. Some simple math, just looking at the voltage divider/control in blue, it looks like the result of the wiper 15v to 0v would be about 1/10 through the divider

I've taken a 4mfd cap and attached it from SIGNAL to ground and it almost kills the signal so that is my assumption of how the signal is reduced. Where the issue comes in is the diode which in 1969, a silicon diode was just about that and no more, nothing special. The parts list shows it as a silicon diode and a duplicate circuit like this also has the same simple glass silicon diode.

So, how is it that these diodes can be "kinda" turned on or off. Is it because of the knee of the diode?

Next picture...

**broken link removed**
I always thought a diode either was or wasn't, never kinda.

My next quandary is why wouldn't the diode be on 1/2 of the time anyway since the signal through the 4mfd is a +/- going audio signal. I'd think it would be toggling the diode anyway. Perhaps it's C1311 and C1217 caps across the diode that keep that from happening.

Maybe I don't see the circuit correctly. You decide. It seemed very simple and logical, then it became unlikely in my head.

 

[audioguru]

The diodes act as an audio compressor so that very loud sounds are not as loud. I had a tape recorder that used diodes like that and the distortion created was not too obtrusive.

 

[chuckstarck]

Thanks for the reply. I've seen equipment with your response as well, although, this circuit is actually the volume/level control. And, yes, those are "all the parts" that are used in it. I'm stumped as to why or how this works.
This is all in a big three manual theater organ. Originally the volume control was typical circuit but after 4 years of models, this was the last iteration to control the level of the 1st manual. The second manual already had this technology, I never noticed it before. The third manual is typical volume control.

 

[ccurtis]

I have no idea why it was designed the way it is, but the diode is acting as a voltage variable resistance that attenuates the input signal to an extent dependent on the pot that adjusts the bias current through the diode and thus the resistance of the diode. The voltage across the diode is always positive so long as the input signal is fairly low, so the diode doesn't switch on and off.

It looks like there are two diodes controlled by the same pot that attenuates two separate audio signals in tandem.

 

[chuckstarck]

Yup, that's what I said. I assume the design is to remove the possibility of a noisy pot being in the actual audio signal. This isolates the control from the audio signal totally and also attenuates two or more signal lines with one pot. Clever. I'm assuming that the caps across the diodes are there to filter out the spikes that would cause rapid conduction changes of the diode.

The question, is this because of and does it occur in the knee portion of the diode response? 58 years as an electronic tech has always had me believe a "plain" diode was on or off, never a gray area in it's function. Maybe this is assuming perfect world results and hence the second drawing. If it is the knee portion, seems the control of the voltage would be real touchy. I've read many college lectures on this to see if someone mentions this and they always skirt the gray area. Today, almost 50 years later, we do have some devices that would do this easily but this is 1969, just past geranium devices that worked better in a crystal set, it amazed me that they would come up with this.

 

[audioguru]

A diode conducts slightly when it has low current and it conducts a lot when it has high current. It has a high active resistance when it conducts slightly and it has a very low active resistance when it conducts a lot. It also conducts in between.
Modern compressor circuits have much less distortion and use a Jfet.

 

[chuckstarck]

That makes good sense. Thanks, it's never described this way. I appreciate your time and effort!!! So as that voltage goes up from the pot, the effective resistance of the diode goes down. Neat.
This device has 6 gajillion diodes in it for switching but only a couple places in signal like this.

 

[AnalogKid]

I always thought a diode either was or wasn't, never kinda.

In the real world, a diode *always* is "kinda". There is no point on a diode's conduction curve that is flat. For any particular current through the diode, there is corresponding voltage. At that operating point, the diode appears as a resistor, with the equivalent resistance from Ohm's Law. As you vary the current through the diode, the forward voltage and the corresponding effective resistance change. That is what your circuit is doing. The two diodes are the shunt legs of two simple two-resistor voltage dividers. The two capacitors keep the slowly varying DC control voltage out of the audio downstream. Note that this works well only if the control current is much greater than the audio signal current, because the audio current through the diodes also affects the effective resistance of the diodes. An old commercial term for this circuit is a Variolosser.

ak

 

[schmitt trigger]

In the pre-IC electronic design age, engineers had to use tricks like this, where the parasitics and non-linearities of components were actually part of the circuit's operation.

This is a lost art....if you read the history the Roland TR808 percussion synth, they actually employed defective transistors to obtain the required sound.