[sdiy] usage of 3080 in micromoog

[Grant Richter]

Just a note to point out the purpose of the differential output.

The ladder filter uses transistors as variable resistances (simple  
explanation) and variation in the control voltage adds a signal to  
the audio portion going through the filter.

Using an identical ladder with out of phase inputs and a differential  
output, the control voltage signal appears as a common mode signal  
and is cancelled out by this scheme.

So changes in the output buffer will mostly affect control voltage  
rejection.

The same trick is used in the diode version or the ladder (VCS3) and  
is similarilly used in the diode version of the Salen-Key filter  
(Stiener).

On Oct 7, 2006, at 1:31 PM, Dave Manley wrote:

> Your comments about the Mini-D made me wonder what circuits
> Moog has used over the years in their ladder filter.  In short, they
> used a different circuit in every product.  A summary below:
>
> Modular 904a - discrete transistor buffer
> Minimoog D - discrete transistor buffer
> Taurus - dc coupled 3080 (Iabc=30 uA) directly connected to ladder
> Micro - dual fet, 3080 (Iabc=450uA)
> Prodigy - two darlington configured 3046, to 3080 for differential  
> to single ended conversion and which is also the vca;  a diff-amp  
> configured 4558 provides the resonance feedback
> Liberation - apparently the same as Prodigy, I don't have  
> schematics, just circuit description
> Rogue - as in the Prodigy, but the two darlington 3046 is replaced  
> with two LF353's,
> Source - dual fet, plus two 3080's one for  vca, one for resonance  
> feedback
>
> -Dave
>> I'd give these reasons...
>>
>> The input impedance of the 3080 is a function of the
>> bias current... run at a very low bias current and the
>> input will be a very high impedance... all in a single
>> stage.
>>
>> Excellent input common mode range allows you to
>> eliminate coupling caps and still have a ground
>> referenced output
>>
>> If you look at the original Minimoog D filter ane
>> compare with the 3080 internal circuit... they
>> are somewhat similar. The 3080 can be considered a
>> well
>> matched differential pair... with the load resistors
>> replaced with the internal current mirrors...
>>
>> Pretty easy and it works well... and its way cheaper
>> than the three opamp config or two (matched) FETs
>> and one opamp.
>>
>> I've tried different ways, the 3080 is "good enough"
>> for rock and roll as they say...
>>
>> H^) harry
>>
>> --- Antti Huovilainen <ajhuovil at cc.hut.fi> wrote:
>>
>>
>>> On Fri, 6 Oct 2006, Dave Manley wrote:
>>>
>>>
>>>> pair.  So I'm not sure it's that much simpler.  Is
>>>>
>>> there some other
>>>> characteristic of the 3080 they are exploiting
>>>>
>>> that an opamp wouldn't have
>>>> provided?
>>>>
>>> I can see two possible reasons: They like the sound
>>> of the OTA distortion (which is certainly case for the VCA) and  
>>> it might
>>> have been cheap. I'd favor the first - the OTA distortion in VCA  
>>> at least
>>> is critical for the Moog sound.
>>>
>>>
>>>> Also was the motivation for the ladder simplicity
>>>>
>>> or that at the time there
>>>> really weren't a lot of choices for making a
>>>>
>>> voltage dependent resistance?
>>>
>>> Considering the time when Moog designer the ladder
>>> filter, the only options for gm stage were differential pair,
>>> exploiting diode nonlinearity to vary the gain with bias voltage,  
>>> using JFETs as
>>> voltage controlled resistors and using vactrols.
>>>
>>> The differential pair is clearly the winner from
>>> engineering point of view (good matching, DC rejection) and the  
>>> Moog ladder is
>>> nothing but 5 differential pairs placed in series. Very elegant
>>> design that eliminates biasing and buffering circuitry from all but
>>> first/last stage and, most importantly, does it without requiring  
>>> a single
>>> opamp.
>>>
>>> Antti - Who has yet to actually build a Moog ladder
>>> filter
>>>
>>> "No boom today. Boom tomorrow. There's always a boom
>>> tomorrow"
>>>    -- Lt. Cmdr. Ivanova
>>>
>>>
>>
>>
>>
>>