Author |
Message |
ropa
Joined: Sep 03, 2012 Posts: 2 Location: Sweden
|
Posted: Mon Sep 03, 2012 2:48 pm Post subject:
Temperature compensated log converter from Osamu Hoshuyama Subject description: Request for experience of sspecific circuit |
|
|
Hi,
I'm new to this forum but I hope it is the place to post my question.
I have a crazy idea to fulfill my childhood dream of designing a modular synth only using BJT's.
I have come up with some designs for various parts but this question is about a technique to temperature stabilize the exponential of a exponential pair by feeding it with a CV that has passed a log/antilog converter that has a bias difference of ap 25mV (in my simulation I ended up with 21mV as optimum) and thereby be able to skip the Tempco.
This is a method I have taken from Osamu Hoshuyama site on Unusual circuits.
From a Spice simulation it is working well but I have a hunch that it will have some real life downsides but I can't figure out what so before I carry on I thought I'd ask if some one tried it IRL.
Obviously there will probably be some IRL need for HF bypassing of the high gain EF stage in the expo biasing feedback loop.
This is the circuit under test. Not completely refactored so there might be some parts that will go away in the end.
Description: |
|
Filesize: |
212.96 KB |
Viewed: |
256 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Download |
Filename: |
BJT LOG 3.asc.txt |
Filesize: |
5.2 KB |
Downloaded: |
257 Time(s) |
|
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Mon Sep 03, 2012 3:15 pm Post subject:
|
|
|
ropa - seems about right to me to post it here.
There are a couple of people here who might be able to say something sensible about your circuit .. not me though ... too long ago I was into transistor models and stuff _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
ropa
Joined: Sep 03, 2012 Posts: 2 Location: Sweden
|
Posted: Sat Sep 08, 2012 1:35 pm Post subject:
|
|
|
Maybe I should be more specific.
Osamu Hoshuyama has presented a method of compensating for the temperature dependence of the exponential factor in the diode equation by using a antilog/log circuit where the later side is de-bias by 25mV.
An antilog/log circuit is also known as a current mirror.
Since I had a current mirror in the exp. converter of my design idea of a discrete VCO I thought I'd use his method instead of the Tempco based compensation.
Since the de-biasing of a current mirror not only changes the scaling transfer function but also the gain it became obvious that it would be hard to adjust such a circuit.
My question is if any one has any experience of this method in real life, in Spice it is fairly easy to step the temp and run some simulations to find the sweet spot, but temp stepping in real world is a different ballgame. |
|
Back to top
|
|
|
|