[sdiy] heated OTA as expo , was Re: (no subject)

[Uwe Schueler]

Tim Ressel <timr at circuitabbey.com>: schreibt...

> I still prefer the heater method of temperature stabilization, a la  
> AN299. Makes it all just go away.

Tim, did you read my posts on muffs about using the LM13700 as heated  
expo tri core ?
www.muffwiggler.com/forum/viewtopic.php?t=194710
for discussion about the principle and
www.muffwiggler.com/forum/viewtopic.php?t=204795
were I publish schematics & layout of the final VCO

it is also discussed by "Guest" in his paper on the last page. He (and  
others) is not a fan of heated
expo designs. Most complaints are:
1) needs extra 40mA heater current
compared to the >100W amplifier behind many synths I can't really see  
the problem. Battery operation is also not a hot topic for analog synths
2) warm up time
True. Using an LM13700 in SMD (PDIP is obsolete anyways) improves both  
power requirements and warm up time.
Personally, I don't mind a 1..2min. warm up time

Another thing I want to mention: Guest shows the error graph of the  
heated OTA (Fig.54) obviously without
HF compensation. It's clear, that if one misuses analog chips with  
small transistor geometries like LM13700
or MC1496 as heated expo pairs, it can't be done without proper Rbb  
compensation.
I've done this with 1 additional opamp as Rossum style compensation or  
recently with a NIC generating -Rbb for compensation. With proper HF  
compensation 10 octaves expo conformity are possible (measured) with  
this low cost circuit.
Meanwhile me and my friends have build 8 Expo-OTA VCOs and we are very  
satisfied with them. I wonder why this
low cost principle doesn't find it's way into the DIY scene. Maybe  
part count is too high