Author |
Message |
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Wed Apr 12, 2006 12:54 pm Post subject:
Amplifier circuit with logarithmic scaling? Subject description: How does one make an opamp get non-linear in a controllable way? |
|
|
I was just reading in the Tape Op bonus articles section about the dbx 386 preamp. What got me really interested was this part:
The Type IV system uses logarithmic scaling from -4 dB to 0 dB allowing high level transient signals to pass through without clipping. Signals below -4 dB are unaffected by the scaling. This creates a system where, because of the mapping, the 386 can accept very hot signals and still not clip. Very much like how tape can handle hot signals with out causing undesirable effects.
I like the idea of that a lot. It got me thinking that something like this, but with an ajustable amplitude point to start the log scaling, would be really really nice as a module. Possibly with a selectable "equation" that controls the scaling, too. How easy/hard is it do make a log or inverse log amplifier? Having it selectable between 1/x or x^n (with adjustable mantissa and exponent) would turn it into a nice compressor or expander. I'm sure something like this has been done before, right?
Being fairly new to this, and not really knowing how to do such a thing, I was hoping somebody here could give me a head start. How does one make an opamp get non-linear in a controllable way? |
|
Back to top
|
|
|
jksuperstar
Joined: Aug 20, 2004 Posts: 2503 Location: Denver
Audio files: 1
G2 patch files: 18
|
|
Back to top
|
|
|
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Wed Apr 12, 2006 9:00 pm Post subject:
|
|
|
Well, we have the Schmooze forum for general discussion on topics not related to anything else. .
Are discussions like the one I started not welcome on the DIY board? My intent is to make a module for my fledgeling modular, and I figured the people on this board would know a thing or two about circuit design and could get me started. I'm looking for an unusual effect that might not be made in just a basic compressor.
But thanks for the tip on compressors. I already kind of knew this was a sort of compression circuit. I guess I wasn't actually sure what I was looking for specifically, just kind of curious in general. |
|
Back to top
|
|
|
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Wed Apr 12, 2006 9:07 pm Post subject:
|
|
|
Oh... oh oh. I misunderstood. Mosc's second post, not the second post on the thread which happened to have been made by Mosc. Ooops.
Ok. That's a good basic circuit to get me started as well. Thanks!
or
I guess they are essentially the same... with one using the transistor as a diode. It's been about 5 years since electronics class, but collector to emitter can be a diode, right?... or is that only emitter to base that can be a diode? With the base grounded, there wouldn't actually be any current flow in that circuit... unless reverse biased... oh it's been too long and I just can't remember it right. Last edited by bigtex on Wed Apr 12, 2006 9:09 pm; edited 1 time in total |
|
Back to top
|
|
|
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Wed Apr 12, 2006 9:17 pm Post subject:
|
|
|
Okay, now in this circuit, which Mosc linked to on that referenced thread, it looks like Q2 is being used backwards.
I don't think I've ever seen the base used as the output like this. I thought the current flow from the emitter to the base determined the current flow from the emitter to the collector, making it a current amplifier.
Can there be current flow from the emitter to the collector without a voltage at the base? I am confused. I see that the collector and the base are tied together, but with the 1M resistor between them and Vcc, there couldn't really be enought current flowing to do anything, right? Where am I wrong about this? |
|
Back to top
|
|
|
Macaba
Joined: Jul 13, 2005 Posts: 160 Location: UK
|
Posted: Thu Apr 13, 2006 8:34 am Post subject:
|
|
|
I would point out, that, as I said in that thread, Ray Wilsons scaling circuit seems to be based on the above, but differs in some crucial aspects. The thing that clinched it for me, is that the above circuit when simulated does not work, whereas Ray Wilson's one DOES. |
|
Back to top
|
|
|
mosc
Site Admin
Joined: Jan 31, 2003 Posts: 18197 Location: Durham, NC
Audio files: 212
G2 patch files: 60
|
Posted: Thu Apr 13, 2006 12:13 pm Post subject:
|
|
|
I'm not sure if the circuit I linked to works or not. It's just something I found on the internet and pointed to it FYI.
Accurate log scaling is a difficult thing, especially if you want it to track over more than one octave and be stable over time and temperature. This is why good oscillators on modular systems can be expensive. Not only are the designs difficult, but they must be made with precise and well-matched components. When DSPs came out, I personally lost interest in that level of circuit design, so I'm in no way an expert. _________________ --Howard
my music and other stuff |
|
Back to top
|
|
|
jksuperstar
Joined: Aug 20, 2004 Posts: 2503 Location: Denver
Audio files: 1
G2 patch files: 18
|
Posted: Thu Apr 13, 2006 12:37 pm Post subject:
|
|
|
bigtex wrote: | Can there be current flow from the emitter to the collector without a voltage at the base? I am confused. I see that the collector and the base are tied together, but with the 1M resistor between them and Vcc, there couldn't really be enought current flowing to do anything, right? Where am I wrong about this? |
Q2 is more or less acting like a variable resistor, when compared to the 1M resistor above it, more or less becoming a voltage divider. The emitter voltage is varying based on input signal, so the difference between the collector and emitter will vary. This "output" of the division feeds back into Q2, so I'd think it'd be a non-linear signal that feeds the op-amp UA1 in the end. At least, that's my best guess at the theory behind it. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24081 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Thu Apr 13, 2006 1:10 pm Post subject:
|
|
|
Ok my go :-)
Q2 is used as a diode for temperature compensation of Q1. _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Thu Apr 13, 2006 1:12 pm Post subject:
|
|
|
Hmmm. I think I liked jksuperstar's answer better, bacuse I could almost understand it.
Please do elaborate. |
|
Back to top
|
|
|
mosc
Site Admin
Joined: Jan 31, 2003 Posts: 18197 Location: Durham, NC
Audio files: 212
G2 patch files: 60
|
|
Back to top
|
|
|
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Thu Apr 13, 2006 1:29 pm Post subject:
|
|
|
Ah! Now that I understand. Thanks!
So how do you adapt this to work over a higher voltage range, short of buffering and attenuating the input and amplifying the output? I bet it must be a much more complicated circuit to do that... perhaps involving lots of diodes/transistors and different value resistors, perhaps as a ladder?
I should stop speculating and just look this stuff up... |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24081 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Thu Apr 13, 2006 1:31 pm Post subject:
|
|
|
bigtex wrote: |
Please do elaborate. |
I'd love to .. unfortunately that would not only cost me this evening but the next weekend as well :-)
But the 2nd sentence on the page the schematic came from hints at it.
Quote: | The second transistor is used as refference and with the PTK resistor minimizes the dT/dV coeficient. |
(my bold)
And it makes sense sort of, both transistors being the same type, having the same temperature coëfficient. To be sure I would have to be calcula it though, unfortunately I have some documentation that needs to be finished before next week ... _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
Pehr
Joined: Aug 14, 2005 Posts: 1307 Location: Björkvik, Sweden
Audio files: 2
|
|
Back to top
|
|
|
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Thu Apr 13, 2006 1:33 pm Post subject:
|
|
|
These circuts only work over a relative narrow range. The point where the cruve on the right take off vertically is 0.7 V in silicon.
And I always thought diodes just turned on and off. So much for theory!
An idea I just had, going along with my original question, would be to use a rail-to-rail opamp, and have that 0.7 volts represent the max amplitudes, so that stuff would get logarithmically attenuated before clipping. Wouldn't that work? Just have the diode be between the output and the power rail.. and then... oh I'm not sure what else. |
|
Back to top
|
|
|
bigtex
Joined: Mar 30, 2006 Posts: 323 Location: Cupertino, California
|
Posted: Thu Apr 13, 2006 1:35 pm Post subject:
|
|
|
Hey, I just bought one of those! I guess I'd better shut up and just put it together |
|
Back to top
|
|
|
Pehr
Joined: Aug 14, 2005 Posts: 1307 Location: Björkvik, Sweden
Audio files: 2
|
|
Back to top
|
|
|
jksuperstar
Joined: Aug 20, 2004 Posts: 2503 Location: Denver
Audio files: 1
G2 patch files: 18
|
|
Back to top
|
|
|
|