Etherwave output amplifier for tuner question

Hello,

Today I build the output amplifier for tuner circuit from the etherwave hot-rodding guide.(p12) The output i get from it is a some what rounded square wave. When I trun the brightness and waveform potmeters the frequency is stable but the shape of the squarewave changes. Is this normal behaviour?

The shape of the sine wave on the audio out pin from the auxiliary header also changes when I turn the brightness and waveform potmeters. When the LM13700 ic is unplugged the shape of the sine wave stays stable.

I believe this would be normal behaviour (I havent checked or simulated) - The LM13700 deliberately distorts the waveform, and some of this distortion will be right at the input - the same point you are tapping.

If you look at the schematic, you will see that the top of C23 goes directly to the LM13700 +ve input, pin 2 of this IC goes internally to a pair of linearizing diodes which are being used in a peculiar way to create distortion - these diodes are directly connected to the +ve and -Ve inputs (download a data sheet for the LM13700 and you will see what I am talking about).

Removing the LM13700 one is removing the diode "loading" and will end up with a waveform that isnt affected by the tone controls.

These are just guesses - but I think theyr right.

Fred

Fred, you are right saying that this is normal behavior. But your "analysis" of what the LM13700 does here is unfortunately not fully correct.

The inputs of the LM13700 (pins 3 and 4) are first of all connected to the bases of a differential transistor pair. As each differential amplifier without negative feedback, even small signals will drive this diff amp into saturation, clipping will occur. That's how waveshaping is basically done in all Etherwaves, including the PRO. The waveform pot allows to add a small amount of variable negative voltage (from -0.25 to -0.5V) to the (-) input which allows to play on the symmetry of the (initially symmetrical about +/-40mV) clipping. The input signal at the (+) input is a full rectified sinus which goes from 0V to -0.6V. 

Now these diodes which are connected from pin2 to the (+) and the (-) input come into play. When there is no voltage at pin2, they have no effect at all (but "zenering" when the input differential voltage exceeds +/-5V, which never happens in an Etherwave). If you make a current flow into pin2, there will be the same bias current through both input transistors, reducing the gm of the differential pair and thus "linearizing" its operation which makes that the clipping limits raise from +/-40mV to up to +/-200mV, making the output signal less clipped, thus softer and rounder. That happens when you turn the brightness pot counterclockwise, making more current flow through R27 into pin2. At the other extreme (brightness fully clockwise) there is no biasing and thus the most sharp clipping.

With the absence of negative feedback, the differential input resistance of the OTA is relatively low which makes that changing the offset voltage on pin4 and the current into pin2 will have an impact onto the signal on pin3. Thus it is normal that the waveform at the raw audio terminal pin varies slightly with the waveform and brightness pot settings. In practice this has no noticeable effect on a tuner which is connected through the ca3080 signal former stage as suggested by Bob Moog. 

Thank you Fred an Thierry, that's more or less clear to me now.

Dous anyone know of a ca3080 soic replacement? (For the moment i used another LM13700 instead of a ca3080.) What IC/circuit are they using in real etherwave plus theremins to generate that squarewave?

Would it be possible for someone to take me a high detail picture of the etherwave+ pcb so that i can read the text on the ic's.

The plus add on circuit board is full of smds, I tried already to figure out what they used for components, but it was impossible since many of them are hidden below the three 1/4" connectors.

I was looking for this in higher detail. i didn't know the connectors had smd's under them.

"Dous anyone know of a ca3080 soic replacement?" - wannes_dn

I dont see that a 3080 is needed.. A simple opamp voltage follower is all that is needed AFAICS.

Thanks for the LM13700 details, and details on the tone control operations - As someone who isnt that impressed by the EW sound I have never really bothered with this bit of the circuitry..

Fred.