Author |
Message |
uniqview
Joined: Jun 15, 2020 Posts: 8 Location: Silicon Valley, California
|
Posted: Wed Jul 15, 2020 11:49 am Post subject:
Single Supply Antilog Generator Subject description: How to build an antilog generator that works on a single-supply with the LM3900 |
|
|
Over many years of research, I've been progressively putting together a unique SDIY synthesizer based on the inexpensive single-supply voltage "Norton" op amp, the LM3900. This one device is being used for most of the system functionality, along with transistors, diodes, and probably some 4000-series CMOS logic. This effort is about obtaining a New Sound, getting good performance out of inexpensive components. So, not using audiophile circuit designs, nor extremely high-precision op amps, etc. A single power supply helps a lot with PCB layout complexity too, because it's more "digital" like, yet is analog.
Sections of this system are starting to come together, which I'm covering in my blog Beyond Application Note 72.
Recently, I was able to construct an antilog circuit, which uses the LM3900, a +15V supply, and a PNP matched-pair transistor. There is an older design that Serge developed in the 1970s, which cleverly uses the LM3900 current mirror for antilog generation. My design is newer, patterned after antilog circuits using NPN matched-pairs and bipolar opamps. A key challenge here is around obtaining good results with only a single-supply voltage. My design is intended for more DC precision with the LM3900, a wide dynamic range, and to allow calibration. In that sense, this design less flexible than the Serge design, but it does seem to work well as a calibrated 1V/octave generator.
My blog post Improved Single Supply LM3900 Antilog Generator details the circuit design, with a scope foto.
I hope this circuit, and maybe the eventual synthesizer may be of some interest to others. _________________ Creating a New Modular for New Sound Last edited by uniqview on Thu Jul 16, 2020 10:20 am; edited 1 time in total |
|
Back to top
|
|
|
elektrouwe
Joined: May 27, 2012 Posts: 143 Location: Germany
|
Posted: Thu Jul 16, 2020 10:11 am Post subject:
Re: LM3900 Antilog Generator Subject description: How to build an antilog generator that works on a single-supply with the LM3900 |
|
|
uniqview wrote: | Over many years of research, I've been progressively putting together a unique SDIY synthesizer based on the inexpensive single-supply voltage "Norton" op amp, the LM3900. This one device is being used for most of the system functionality, along with transistors, diodes, and probably some 4000-series CMOS logic. |
Nice, looking forward to seeing the full synth !
PS: just complaining about the title: IMHO you built a PNP expo and replaced the opamps used in the standard configuration with Norton amps. No one would call a PNP expo built with TL074s a "TL074 Antilog Generator" because the antilog comes from the PNPs and not from the TL074. Same with your circuit. I think it's misleading, because Serge T., as you already mentioned, made a nice VCO using the LM3900 as antilog gen. in a true sense.
PS 2: I'm just developing a "true LM3900 expo" tri-core VCO; that's why I'm complaining |
|
Back to top
|
|
|
uniqview
Joined: Jun 15, 2020 Posts: 8 Location: Silicon Valley, California
|
Posted: Thu Jul 16, 2020 10:25 am Post subject:
Re: LM3900 Antilog Generator Subject description: How to build an antilog generator that works on a single-supply with the LM3900 |
|
|
elektrouwe wrote: |
Nice, looking forward to seeing the full synth !
PS: just complaining about the title: IMHO you built a PNP expo and replaced the opamps used in the standard configuration with Norton amps. No one would call a PNP expo built with TL074s a "TL074 Antilog Generator" because the antilog comes from the PNPs and not from the TL074. Same with your circuit. I think it's misleading, because Serge T., as you already mentioned, made a nice VCO using the LM3900 as antilog gen. in a true sense.
PS 2: I'm just developing a "true LM3900 expo" tri-core VCO; that's why I'm complaining |
Thanks much for the well-wishes , this work has taken a long time.
Is your LM3900 VCO done? Do you have a link you can share? Definitely would enjoy seeing what you did !
Sorry about the title.
It came about because many of these LM3900 circuits really are quite different, and this seemed more of the same. Because while this circuit is superficially similar-looking compared to the NPN and bipolar op amp approach, I have found in general that unipolar circuit design is a very different system framework.
But you raise a very fair critique, thank-you. One reason I keep emphasizing LM3900 is that it seems to be a non-ordinary way to construct a synthesizer mainly with that one inexpensive and medium-performance device, for both signal generation and signal processing. For sure, the LM3900 is used in coalition with transistors and diodes, which do provide antilog, switching, and other auxillary functions. As you point out, of course others do this with TL074 and a myriad of other bipolar or CMOS analog devices. Agree with that. I think though that representing a form of analog computing (sound synthesis) within a unipolar framework has interesting issues and unusual opportunties. The unique operating mode of the LM3900 is an integral element to this enterprise,
However, in deference to this feedback, and also wrt honoring Serge's earlier innovation, I actually changed the title here, and to the linked blog post. Your point is well taken.
I do wish you much luck and fun with your SDIY adventures! _________________ Creating a New Modular for New Sound |
|
Back to top
|
|
|
|