Hi Chip,
I just finished watching the movie "The Cincinnati Kid". Whereas in
former years I would identify more with Steve McQueen nowadays I tend
more to the Edward G. Robinson character. It seems a little an
allegory to this LM1600 matter.
I am perfectly confident that I am able to understand and calculate
this VCA stage. But it is late today now (11 p.m.) and I hear my bed
calling my name :-)
I take the challenge and I'll come back later, hopefully with some
figures!
Johannes
P.S. I am not sure whether to answer to the list or your blog. What do
you prefer?
On 2013-03-03 21:41, chipaudette wrote:
> What I find so difficult about the LM13600 is my inability to calculate what the transconductance should be for a given circuit configuration.
>
> For the VCA shown in the datasheet, the datasheet says that:
>
> Iout = Iin∗(2∗Iabc)/Idiode
>
> Seems simple enough, except I can't seem to use this equation to get anything reasonable. For example, the datasheet says that their example circuit results in Vout/Vin = 940 x Iabc. I can't see how they got that value from their circuit.
>
> Then, when i try to generalize to slightly different circuit setups (like the one that we're discussing in the polysix), I can't get any reasonable answer that comes close to the experimental result.
>
> Part of the reason could be that the assumptions in the datasheet, while accurate for the specific circuit that they drew, has important (critical) differences with how people (such as the Polysix designers) actually end up implementing their VCA. This page has a very interesting discussion...
>
> http://lushprojects.com/blog/2012/08/lm13700-missing-forumla/
>
> ...but it doesn't have quite enough detail to help me through all the way so that I can compute the gain vs Iabc relationship that I'm looking for.
>
> Bummer.
>
> Johannes, are you able to correctly calculate the gain of the LM13600 for the example in the datasheet or for the setup in the Polysix? It thwarted me....
>
> Chip
>
> --- In PolySix@yahoogroups.com, Johannes Hausensteiner <johau@...> wrote:
>>
>> The LM13600 as VCA works without (negative) feedback. If you think about
>> it this makes sense. The operational amplifier with negative feedback
>> forms a control loop: the output voltage is increased or decreased until
>> the two inputs have zero voltage difference - as long the circuit can
>> reach and maintain this condition. The purpose of the LM13600 (and
>> similar circuits) is that the gain can be controlled via a control
>> signal (current in this case). You would not be able to control the
>> gain and hence the output level of a closed-loop amp. (you would just
>> be controlling the loop gain, which is not what is desired). So the
>> classical VCA application of the LM13600 must not have a negative
>> feedback. In order to not overdrive the input stages the signal must be
>> rather low. Therefore you see a voltage divider at the LM1360' input.
>> In the case of the schematic snipped it is about the factor 100 (if we
>> neglect the pre-emphasis R156/C80).
>> Furthermore you may have noticed that the LM13600 is dubbed as
>> "Operational Transcoductance Amplfier" in the datasheet (OTA), not as
>> simple "Operational Amplifier". "transconductance" means the ability
>> of conducting current. It is the reciproc (1/x) value of the electrical
>> resistance. Ohm's Law is well known to mankind; it says R=U/I. This is
>> the resistance. We get the transconductance as I/U (reciproc value). In
>> the case of an amplifier this denotes a voltage-type input and a
>> current source output. The transfer function (output / input) becomes
>> Iout / Vin. Normally we are used to a simple voltage rational transfer
>> function Vout / Vin, which commonly is denoted as amplification or
>> (voltage) gain. Btw. to achieve a low-source resistance voltage output
>> the LM13600 features am emitter follower which transofrms the high
>> impedance current output to a low-impedance voltage output. It does so
>> by means of a parallel resistor. The output current creates a
>> proportional voltage on that resistor and the voltage follower makes
>> this low (source-) impedance.
>>
>> Conclusion:
>> To be able to calculate the voltage gain of an LM13600 stage both the
>> input divider and the parallel output resistor have to be known and
>> put into the equation.
>>
>> Here at the KLM-368 input stage the LM13600 injects the current
>> directly into the following op amp IC17. R164 does not matter for the
>> DC amplification, I expect the LM13600 source impedance one or two
>> magnitudes higher than 4.7kOhm. Probably R164 together with C75 form
>> an additional low pass characterstic of the stage.
>>
>>
>> Everthing clear ;-)?
>>
>> Johannes
>>
>>
>>
>>
>>
>>
>> On 2013-03-02 22:31, chipaudette wrote:
>>> OK, I took some measurements of the resonance-controlled VCA. You can see my data and analysis here...
>>>
>>> http://synthhacker.blogspot.com/2013/03/polysix-resonance-controlled-vca.html
>>>
>>> My conclusion is that this VCA compensates for the natural drop in volume from the voice board that occurs when you start turning up the resonance knob. On my Polysix, as I turn the knob from "1" to "3", I get an increase in gain up to 6 dB, which almost perfectly components for the drop in volume from the voice board.
>>>
>>> When I turn the resonance knob past "3", there appears to be no further change in gain from this VCA.
>>>
>>> I found this very interesting!
>>>
>>> In doing this analysis, I really struggled with trying to predict the behavior of the LM13600. In the end, I just accepted its behavior based on the data. I would prefer to have good working equation for the gain of the LM13600 given the values of the circuit elements and given the value of the control current. The datasheet wasn't helpful enough. Does anyone have a good quantitative understanding of the gain of the LM13600?
>>>
>>> Thanks,
>>>
>>> Chip
>>>
>>> --- In PolySix@yahoogroups.com, "chipaudette" <chipaudette@> wrote:
>>>>
>>>>
>>>> Well, I still think that it's a VCA, but I found that the gain goes ∗up∗ with increasing resonance. The gain increases until you reach a value of about 4 on the resonance knob. Above that, the gain stays the same.
>>>>
>>>> More analysis still forth-coming...
>>>>
>>>> Chip
>>>>
>>>>
>>>> --- In PolySix@yahoogroups.com, "chipaudette" <chipaudette@> wrote:
>>>>>
>>>>>
>>>>> Hi Johannes,
>>>>>
>>>>> A few posts ago, you said:
>>>>>
>>>>>> I always wondered why the Resonance control is part of this.
>>>>>> Similar to the output LPF there is a controlled (by the
>>>>>> Resonance knob) LPF at the input.
>>>>>
>>>>> Looking at the schematic, I believe that this LM13600 (the first half of IC20) is actually wired as a Voltage Controlled Amplifier, not as a VCF. Yes, there is a cap to ground after the OTA, but there is no negative feedback. Without the feedback, I don't think that it can be a VCF. So, I think that it is just a VCA.
>>>>>
>>>>> Looking below this chunk of circuitry, you'll see that the other half of this LM13600 is wired in a similar way (though with different component values). It, too, has no negative feedback. This second half of the LM13600 is implementing the VCA MOD and the Program Volume, so one can probably assume that they definitely intended this LM13600 stage to be acting as a VCA.
>>>>>
>>>>> Due to the similarity, it's my conclusion that both halves of IC20 are acting as VCAs.
>>>>>
>>>>> Going back to the first half of the VCA (the part that seems to be controlled by the resonance setting), it seems to me that this circuit is simply adjusting the amplitude of the audio based on the resonance setting. To my eye, it appears to reduce the amplitude of the audio as the resonance setting is increased. I guess that the idea is to keep the amplitude of the self-oscillation under control. Just my conjecture...
>>>>>
>>>>> Chip
>>>>>
>>>>
>>>
>>>
>>>
>>> ------------------------------------
>>>
>>> Yahoo! Groups Links
>>>
>>>
>>>
>>>
>>
>
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>
>