geofex ota application

[domenico]

Hi I'just build a volume control using geofex application of an ota article.
I can't control the lm13700 with buffer (a quarter of a tl074) but only with a pot directly connected to amp bias input.
Is there any possibily to damage the ic?
Thanks,
Domenico

[R.G.]

Hi I'just build a volume control using geofex application of an ota article.
I can't control the lm13700 with buffer (a quarter of a tl074) but only with a pot directly connected to amp bias input.
Is there any possibily to damage the ic?

??

Do you mean you're trying to feed the output of an opamp directly into the bias control pin?

If so, yes, it can be done, but it is not simple, and not a beginner thing to make work well and reliably. In particular, if you *ever* put more than 1 or 2 ma into that pin, the chip will self destruct. I think the number is 1ma for the CA3080 OTAs and 2ma for the LM13700s, but read the datasheet, don't rely on my memory.

The Iabc input on an OTA is a **current** input. If you feed it a voltage, not a current, then you must make sure that the current resulting from that voltage never goes over the sudden-death current for the chip. Otherwise, you wind up buying a lot of replacement chips and writing lots of posts here.

[Keppy]

Adding to what R.G. said, if you are using the opamp as a variable voltage source, you need to add a resistor from the opamp output to the 13700 bias input. Varying voltage / constant resistance = varying current.

[R.G.]

Yep, what Keppy said.

On the other hand, you can make the gain per current act exponential, not linear like it is normally, by forcing a voltage onto the pin. But if you do this, you need to scale it so your input control of whatever kind be linear in voltage from 0V to no more than the voltage which makes the Iabc be less than the destruct current at maximum input. And the destruct *voltage* is knowable, but not known for any specific chip. What works better is forcing a linear *voltage* onto the pin with a maximum voltage of X and a current limit of Y, where Y is the destruct current minus 100uA or something.

It gets complicated.

Synth-y.

[domenico]

Hi I'just build a volume control using geofex application of an ota article.
I can't control the lm13700 with buffer (a quarter of a tl074) but only with a pot directly connected to amp bias input.
Is there any possibily to damage the ic?

??

Do you mean you're trying to feed the output of an opamp directly into the bias control pin?



If so, yes, it can be done, but it is not simple, and not a beginner thing to make work well and reliably. In particular, if you *ever* put more than 1 or 2 ma into that pin, the chip will self destruct. I think the number is 1ma for the CA3080 OTAs and 2ma for the LM13700s, but read the datasheet, don't rely on my memory.

The Iabc input on an OTA is a **current** input. If you feed it a voltage, not a current, then you must make sure that the current resulting from that voltage never goes over the sudden-death current for the chip. Otherwise, you wind up buying a lot of replacement chips and writing lots of posts here.

I mean to feed lm13700 bias control pin in a similar way like anderton volume pedal i.e. a pot connected to 9volt/ground and a resistor to control pin, is it a stupid idea?

[R.G.]

No, not stupid at all. It's just that you have to be aware of some things.

I mentioned most of them. The Iabc control pin always sits between 0V and two diode drops, about 1.4V, above the most minus power supply voltage pin. Inside, there is the equivalent of two diodes. So you can't pull it more positive than that without a big current flowing into the pin, and killing the chip. You can drive the pin with a voltage source through a suitable resistor.

"Suitable" means that the resistor limits the current flowing into the control pin to less than the 1ma (... or is it two ma for the LM13700? Somebody ought to look that up.) that the chip can take. So if you drive this from a pot between ground and +9V, and your "+9V" will really not be more than 9.5V, which is what a fresh alkaline batter puts out, then you can calculate some resistor which does kind of what you want. The value of the pot gets into it, as it acts like a resistance between 0 ohms and 1/4 of the pot's nominal value in series with the voltage it puts out, so you tend to either get some imprecision from high value (like, 100K maybe) pots and a lot of current pulled through the pot with low value pots like 1K, 5K or 10K. The resistor between the pot wiper needs to limit current to 1ma (... or is it 2ma?) with the pot wiper at 9.5V and the Iabc pin at, say, 1.2V. That's (9.5-1.2)/0.001 = 8.3K. Or 4.15K for a 2ma limit.

Hmm. With this setup, putting a 12V or 18V power supply on the pedal some day to see if it sounds even cooler with "more headroom" might kill the chip. Hmmm.

[Keppy]

2 mA absolute maximum, meaning you don't really want to be anywhere near there. Most of the datasheet applications include a bias resistor that would result in 1 mA of current if the positive supply voltage was applied to it (actually a little less after adjusting for the diode drops). A few applications have a bias resistor that results in 2 mA fed into both bias inputs simultaneously, so I guess that's 1 mA apiece if they're perfectly matched. In other words, a 9k-10k resistor should be about right in this application. 4.5k or less will fry your IC if you put 9v on it.

[domenico]

Tried yesterday and everything is ok! I will use only regulated 9v in order to not fry the ota.
Another question, what will happen if I put 12 or 18v in the original circuit (with opamp feeding ota) ?
Thanks!

[ORK]

Tried yesterday and everything is ok! I will use only regulated 9v in order to not fry the ota.
Another question, what will happen if I put 12 or 18v in the original circuit (with opamp feeding ota) ?
Thanks!

This: you will double the resistorvalue beforehand!

[R.G.]

Tried yesterday and everything is ok! I will use only regulated 9v in order to not fry the ota.
Another question, what will happen if I put 12 or 18v in the original circuit (with opamp feeding ota) ?
Thanks!

This: you will double the resistorvalue beforehand!

... or else you will convert the chip into a Light-Emitting_OTA. 

Hmmm. That wouldn't last long before it got dark. OK, it will become a Darkness-Emitting-OTA.

[domenico]

R.G., I really like that circuit that I've built but unfortunately it has hiss and a thump when pot goes is a certain portion of rotation...I've aso built your photo fet mike muter and I would ask you if it's possible to build a volume pedal out of it by driving h11f3 leds with a potentiometer.
obviously I would put a buffer before h11f3 series.
what do you think about it ? do you have an alternative to lm 13700 for low noise operations ?
Thank you,
domenico

Tried yesterday and everything is ok! I will use only regulated 9v in order to not fry the ota.
Another question, what will happen if I put 12 or 18v in the original circuit (with opamp feeding ota) ?
Thanks!

This: you will double the resistorvalue beforehand!

... or else you will convert the chip into a Light-Emitting_OTA. 

Hmmm. That wouldn't last long before it got dark. OK, it will become a Darkness-Emitting-OTA.

[R.G.]

The H11F3 is unfortunately limited in the size of signal it can handle without distortion. The LM13700 is limited by the small size of the signals it can withstand without distortion, so that means you get hiss. Most people don't like working with the H11F3's limitations on guitar signals.

You can change over to using the predistortion diodes in the LM13700 to handle higher signals, getting about a 20db hotter signal and hence relatively 20db less noise in the final signal.

Beyond that, you need to go to some more refined VCAs. THAT corporation makes some nice VCAs with low noise and good specs, but they're harder to get and more expensive.

[domenico]

unfortunately THAT vca is very hard to get...could you please explain me how can I change over to using predistortion diodes?
thanks,
Domenico

The H11F3 is unfortunately limited in the size of signal it can handle without distortion. The LM13700 is limited by the small size of the signals it can withstand without distortion, so that means you get hiss. Most people don't like working with the H11F3's limitations on guitar signals.

You can change over to using the predistortion diodes in the LM13700 to handle higher signals, getting about a 20db hotter signal and hence relatively 20db less noise in the final signal.

Beyond that, you need to go to some more refined VCAs. THAT corporation makes some nice VCAs with low noise and good specs, but they're harder to get and more expensive.

[R.G.]

You run a current through the predistortion input pin. Generally a resistor to V+ works fine. It's shown in the LM13700 datasheet, and explained.