[sdiy] EH frequency analyzer

[ASSI]

On Freitag 08 August 2008, harrybissell at wowway.com wrote:
> The LM13600 has a buffer amp whose impedance is fixed at an
> 'average- good for most applications" level.
>
> The LM13700 has a buffer whose impedance is changed with the Gm
> bias. This means the 13700 will perform better in wide range
> applications, especially at really low Gm biases...

Actually it is the other way around, the LM13600 is the one where 
the buffer bias follows the Iabc (with a 1:4 1/6 mirror ratio 
that is not mentioned in the data sheet).  Whether or not 
thumping will be a problem depends on how you use the buffer (if 
at all).  It can be better with constant bias (no CV feedthrough) 
or with varying bias (less impedance change over output range).

> If you want to dig in, look carefully at the spice model
> listings for each and note the differences in the output
> transistor parameters. This are the only clues I have found to
> what is really going on.

The two datasheets at least mention the difference and show the 
bias mirror for the LM13600.  Don Sauers great website 
http://www.idea2ic.com has the saucy bits the data sheets are 
missing including the mirror ratio.  With Don's permission I have 
recently dissected the die micrograph of the LM13600 (clearly 
_not_ the LM13700) in my OTA writeup so you can exactly know 
where each of the transistors is (and why the two halves are not 
exactly symmetrical if you look closely enough).

http://Synth.Stromeko.net/DIY.html#OTA

The NE5517 has yet another biasing scheme for the buffer.  I can't 
be sure because I never had one under the microscope, but I don't 
think they d have identical layout.


Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk]>+

SD adaptation for Waldorf rackAttack V1.04R1:
http://Synth.Stromeko.net/Downloads.html#WaldorfSDada