The LM3900 is a quad Norton op amp, not an OTA. They’re very different.
Well, I knew the names of the thingies and those do not explain a lot.
What is it that makes them different?
As I understand it (very minimally) an OTA takes a differential voltage and outputs a current; a Norton amplifier takes a differential current and outputs a voltage.
Which sounds like they are 1 resistor apart. @All: Maybe someone with some expert knowledge could comment on this?
Well to stay simple CFA/Norton amplifiers/transimpedance amplifiers are like normal opamp but faster because they directly work with current. You can even find some fast normal opamps that have a CFA inside, with the correct input stage wrapped around.
The main advantage of transconductance amplifiers (OTA), is that you can bias the current output stage with an external current, hence easily creating a VCA or even a full multiplier.
I never heard of current operational amplifier (COA), though… But they seems to exist.
Edit: I just realized that actually COA are basic building blocks for normal opamp (VOA), with the corrects input and output stages added on it.
@Dud le Bricoleur built another copy of this, which reminded me of this thread and that strange power supply connection, and after skimming through some 70s opamp literature I think I’ve figured it out.
Here’s the Iabc input on the CA3080:
The Iabc input (pin 5) controls the current through the differential input pair that handles the input signals (pins 2 & 3), via a transistor connected to the negative supply rail (pin 4). Note that this means that Iabc is held one diode drop above the negative rail, so if you feed it through a passive resistor you have to dimension things for CV +
10.6 V 11.3 V to get the right current (for a ±12 V supply, that is, adjust accordingly for other supplies).
Most designs work around this issue by using an active current source to drive Iabc; either a transistor configured as a common base amplifier (see e.g. the Henry BD++) or better an opamp + transistor (see e.g. Electric Druid’s vintage VCA). This lets you ignore that 11.3 V difference and think only in terms of current, but it adds some external components, and if you’re sneaky you can save one whole transistor by using one that’s already there:
By putting Iabc in the feedback loop, the incoming Icv current (CV / the input resistor) has nowhere to go other than through pin 5 (Iabc), and for that to happen the opamp needs to pull pin 4 down roughly two diode drops (D2 + internal circuitry).
Of course, the output stage uses the same supply rail but since the output is current-driven that’s not a problem as long as you drive into ground or higher (e.g. the current to voltage converter with a virtual ground used here).
You can use the same trick with the LM13700 but it has two diode drops to the rail so you need to add one more diode in series with D2 to have some margin, and of course both halves use the same supply, so you’re probably better off using an external transistor so you can use the second amplifier for something else…
(if anyone wonders about the D1 in the original schematics, it’s there to protect against reverse CV polarity).
Something like this should work:
The transistor can be any small signal PNP, e.g. 2N3906, BC556 through 560, etc. The 10k resistor is optional but will make sure the Iabc current stays well below the 2 mA max rating no matter what the CV circuity does.
(with Tayda prices, this change will make each VCA around 2 cents more expensive)
(Now you’ve tricked me into ordering a bunch of 3080s off eBay, from a few usually reliable suppliers I’ve used before. Will report back if they seem legit, or if they’re repainted 358s or something
Ok, the first batch arrived today (faster than expected), and my first impression is that they’re recycled. But I can live with that. Now I’m just waiting for the PCBs to show up…
Ok, got the second eBay batch today, and it looks a lot less recycled. Maybe they pulled them from sockets this time
(now I just have to figure out which seller sent me what package…)
There are some things Man was not meant to know
Single board stripboard by @Dud here:
but not yet tested
have you tested yours ?
I did some smoke tests IIRC but don’t recall if I did any more functional testing (my build discipline isn’t what it’s supposed to be )
What’s that power supply module?
An USB plug ?
btw i havn’t seen the title
It’s an industry standard SIP DC/DC converter. Nice for testing, since they’re well protected against shorts/overload.
IIRC the one in the picture is a MEAN WELL DPAN02E-12 but they’re available from a bunch of manufacturers, and for different input voltages (5 V USB in this case) and power requirements.
I edited my stripboard layout
I saw that i wrote 100k for all pot
it’s allways not yet tested but it’s now the good values for pots
I still have 2 of these modules to build, maybe tomorrow …