OTA Based tremolo-Success! And questions

[ExpAnonColin]

I have been fooling with OTAs (2 CA3080s burned so far, 0 13700s) and decided to try them with a tremolo.  So I used my handy dandy dual OA schmitt triger/integrator LFO and made a fairly simple system where the current was dropped to what I thought was an appropriate level and a single OTA on the 13700 was used as a basic light amplifier to be modulated...  I got mixed, and bad, results.  So I snooped around and looked to J Hollis's frobnicator for some help, and hooray! it's working now.  But the things I changed didn't occur to me at first and I want to know why it helped.

1) Why the LED/diode in series after the LFO?  Why isn't a resistor enough to provide the appropriate current?
2) The trim, the carrier null trim, it works perfectly but why?  And why do any of them work for that matter?
3) The 47n cap at the input-it sounds cracky and misbiased without it, but why?  Is the guitar's impedance improper for the OTA's input?
4) Why the trannie buffer at the output?  I still don't have that on my trem, and I love how it sounds as is.

Fortunately it's working very well, I am just going to rig up a nice buffered mixing system and a better control over slope and I will be posting some schems... the final version will be a pan box (with the LFO simply inverted and the other side of the 13700 used for the "other" side of the pedal.)

-Colin

[R.G.]

I'm assuming you also readApplications of an audio Current Controlled Amplifier at GEO:   http://geofex.com/Article_Folders/VCA%20Applications.pdf

1) Why the LED/diode in series after the LFO? Why isn't a resistor enough to provide the appropriate current?

I believe that the LED is there to show visibly the LFO speed, and the LED + diode raises the minimum voltage that the LFO output has to do so it's closer to the 1.2V that the Iabc pin has to sit at.

2) The trim, the carrier null trim, it works perfectly but why? And why do any of them work for that matter?

It injects or sucks out just a whiff of current from the input so that any imbalances in the input diffamp pair are balanced. Quoting from the GEO article, it prevents thump and LFO feedthrough when the LFO changes quickly.

3) The 47n cap at the input-it sounds cracky and misbiased without it, but why? Is the guitar's impedance improper for the OTA's input?

It is misbiased without it. The OTA input floats somewhere in the middle of the +9/ground supply. Leaving out the cap pulls one side to ground. Bad juju.

4) Why the trannie buffer at the output? I still don't have that on my trem, and I love how it sounds as is.

OTAs are current source outputs, so their gain is dependent on loading. John chose to use an output buffer to insulate the OTA output from loading further down the chain. That's why buffers are integrated on the chip on LM13600/13700s, the CA3280, and the NE5517 OTAs. They usually need a buffer for predictable results every time.


Try some of the other apps from the GEO article as well. They're fun!

[Transmogrifox]

Perhaps I should give that one a try.  It seems simple enough, but I was having trouble getting the LM3080 to work for me.  I guess I ddin't spend enough time gathering info from the datasheets.  My problem, though, was that I was simply not getting enough rejection on the low level.  I was making a triggered volume swell effect that was starting at an audible level.  I would be surprised if I was getting 20 dB rejection.  I think RG's OTA circuit, with 80dB rejection should be more than enough to make a nice sounding triggered swell circuit.  Thanks for the link to the article, RG

[ExpAnonColin]

I completely missed that article I'm afraid but I will give it a good read.  Thanks for the help RG.  Can you go any more in depth as to *why* the trimpot works as it does?  Why does it have to be such a specific value?

I actually did try using the LM13700's onboard buffer (sans any resistors or anything of the sort) but the signal dropped dramatically.  I will have to look into it more...  or just use the same sort of thing that hollis is doing because it works brilliantly

-Colin

[puretube]

Colin: check LM13600/-700 datasheets and appnotes: lots of basic schems in there - they all work!

hey: and where is GEZ ?

[ExpAnonColin]

It turns out I HAD read the VCA applications sheet... just had forgotten about it.  I am going to put another buffer in front of the ordeal as well I think, as you did, RG.  

-Colin

P.S. I keep getting "Server temporarily busy" notices.

[puretube]

BTW. R.G.: the texts are incomplete @ "application 1" & "circuit 2"... (FYI)


P.S.: yes, those "503" messages occur often, lately...

[R.G.]

Thanks guys. I'll go take a look.