Is there a dual equivalent to the quad LP324?

(To add to NT's list of applications) I saw a 4069 used as a four channel matched voltage controlled variable resistor in a tunable filter. One of the remaining inverters was in the feedback path of an opamp to linearise the control voltage - resistance characteristic.

Regards, Allan

Hello Allan,

Sweet. Do you remember where you saw that? If public domain it sure would help in convincing engineers and others that this isn't voodoo. I can't show other client's schematics for obvious reasons.

Klaus, there is an unbuffered inverter on the 74 series: 74HCU04. Also a buffered one for more gain, the 74HC04. Then there is the open drain variety 74HC05 which can have other interesting semi-analog uses. What is quite sad is that the venerable CD4007 did not find its way into the

Just a word of caution. The 74HC series is much more 'snappy' than the CD series. Their devices are stronger, meaning they can draw large currents when mid-biased and run at the higher end of their supply voltage range, possibly too much.

Regards, Joerg

This was in a guitar effects pedal. It was a phaser, and the phase shift network had a number (I'm fairly sure it was four) of single pole phase shifters in series. Each stage had a series C and an R to ground which was the N-ch fet in the inverter output. The inverter

-ve rail was at analog ground ( = mid rail for the opamps). The inverter inputs and positive rail were connected to the control voltage, which was driven from the linearising circuit.

The first time I looked at it, I thought WTF? How can that work? But it sunk in after a few seconds.

Regards, Allan

... Because it is not appear to be drawn on a scrap of gridded A5 paper with a 6 mm crayon while driving to work ;-)

Thanks Joerg :-)

Yes - the downside is the added current consumption of these circuits. But if that can be tolerated then very elegant circuits can be done this way. I just wish someone would write a book on these special uses of CMOS ICs and other special uses of other devices (like using the LED for a simple current source to cancel out most of the thermal drift (recent discussion on SED))

My problem is though convincing my colleagues that this is ok for large scale production. If they cannot read a spec for it in the datasheet it is not allowed. That obviously kind-off makes sense since characterization of a device cannot provide a guarentee that the manufactor won't change a process (upholding the spec'ed data) that will significantly change the characterization/measured data.

But still - when looking for the last cents to be saved - un-conventional circuits are often called for

(in my last project I made a low-power SMPS from LM139 and CD4093 and discretes (great savings there). My two most favorite parts :-)

Regards

Klaus

Hello Klaus,

Too little market for it. Many kids coming off universities these days can't design analog anymore. Some can't even solder. All they did was calculate and simulate.

Your colleagues are right. Check Philips, they have about the best and most detailed family specs for CMOS logic I can think of. But be patient, their web site is horrible IMHO. Not so bad from Europe though because of less latency.

I only afforded my leanest switcher two sections of a hex Schmitt trigger (needed the other four elsewhere) but no LM139. Cost even less than a TL494 plus mine can go to 90% duty cycle (the TL can't). It has current mode control, power sense, voltage cap and current ramping. And it turns itself off when there is no load and back on when there is one, to conserve battery juice.

Regards, Joerg

Very nice indeed - looking forward to your first book release *G*

Cheers

Klaus

Widlar had nothing to do with the 741... that was designed by Dave Fullagar (who later founded Maxim with Jack Gifford).

I believe that Widlar thought his LM101 was far superior to the 741 and would have been offended to be given credit for it.

Don't you think the LM10/LM11/LM12 are pretty clever?

His bio says "more than 20" - the LM331 and the LM337 are 2 that he has mentioned.

Hello David,

Thanks, now we know. Hats off to Bob Pease, the LM331 is a mighty fine chip.

Regards, Joerg

I remembered the circuit. There were four of these in series.

___ .--|___|--. | | | | ___ | |\\ | Vin --o---|___|---o---|-\\ | | | \\ | | | >-o------ Vout | || | / '----||-----o---|+/ || | |/ | Vcontrol | + ---. | | | | | |' | '---| >O--' |/ CD4069 (1 of 6) | | GND

(created by AACircuit v1.28.4 beta 13/12/04

Hello Allan,

Cool, thanks. This won't be very linear but sometimes that is exactly the desired effect.

Regards, Joerg

The amplitude linearly was pretty good, but that doesn't matter much for a guitar effect.

The control voltage linearity was improved by a separate lineariser circuit, using one of the spare inverters and an opamp, like this:

|\\| V in------|-\\ ___ | >----|___|--+ I ref >-----+-----|+/ | | |/| | | | | | Vcontrol | | + ---+-------------------------+ | | | | |' | '---| >O--' |/ CD4069 (1 of 6) | | GND

(created by AACircuit v1.28.6 beta 04/19/05

Here, the 'R' synthesised is Vin / Iref

Regards, Allan

I should point out that the signal amplitude was at most a few hundred mV p-p.

Regards, Allan

Hello Allan,

Then it should be fine. If they played rock it won't matter anyway. I have a Datong RF clipper that also uses CD4000 chips for much of it's analog stages. It can clip without mashing up the signal and when I loaned it to a friend to try it on his E-guitar he almost didn't want to give it back. He bought one right away.

Regards, Joerg