Gyrator from CMOS stage ?

[Brymus]

Is it possible to use a CMOS stage (4049UBE) to make a gyrator for an EQ band like it is an op amp or transistor ?
Does anyone know of any examples if it is ?
thanks,Bryan

[Brymus]

Do I want an SE inductor or a floating indutor since both require two CMOS stages ?
I still cant find any actual circuit examples of a tunable circuit like I want.
I want to make a couple of EQ bands in the mid band area with half a CMOS IC
Everything ,info wise seems to be PDFs they want you to buy.

[Paul Marossy]

I know that you can use a CD4066 in an EQ section. The "Bouteek Preamp Distorter" uses a pair of them, one acting as a low pass filter and the other as a high pass filter. The tone control kind of pans between the two. It's actually a very cool design, never seen anything like it before. I breadboarded the complete circuit, and it's a good sounding pedal.

I would share that EQ part of the schematic that I made from reverse engineering it, but it's been sworn to secrecy.  

[earthtonesaudio]

If you MUST use a 4049, you can replace the inverting op-amp with one of the CMOS inverters and construct a low-pass section like this:

http://www.millertechinc.com/pdf_files/MTI%20TN094%20Zero%20DC%20Offset%20LPF%20and%20the%20D%20Element_files/image009.gif

...then add a capacitor from "out" to ground, making the net impedance seen from "in" into a bandpass.  You could basically sub that in for the whole kit that attaches to each of the wipers in a graphic EQ.

[Ben N]

I know that you can use a CD4066 in an EQ section. The "Bouteek Preamp Distorter" uses a pair of them, one acting as a low pass filter and the other as a high pass filter. The tone control kind of pans between the two. It's actually a very cool design, never seen anything like it before. I breadboarded the complete circuit, and it's a good sounding pedal.

I would share that EQ part of the schematic that I made from reverse engineering it, but it's been sworn to secrecy.  

The ROG Mr. EQ has high and low pass CMOS sections, and bandpass too, that could probably be adapted to what you're talking about, Paul. I don't know about gyrators, though.

[Paul Marossy]

I really don't know if that pedal I reversed actually uses a gyrator or if it's just some kind of active filter arrangement. I have to confess that I am not an expert in this area, at all. But in any case, it does seem to work well.

The CD4066 section I mentioned does appear to be a variation of what is in Boscorelli's "Stompbox Cookbook", Project No. 2, the "Distort-O-Matic". But Boscorelli uses a CD4069 instead, and it's not panning between a high pass filter and a low pass filter. But the essence of how it works is the same from what I can see.

[Brymus]

thanks guys thats helpful, the notch on the Mr EQ is close to what I am after ,I just want it adjustable from extreme notch to extreme hump. :icon_confused:I was hoping to have two mid bands with extreme settings,
And I want a low boost/cut and high boost/cut .

Paul does the one your talking about use the bias to adjust the effect of the filter ?
I ran across an article that suggested that,just no examples...

I am basically using voltage 4-15V to control gain in this pedal,so the filters would follow suite I suppose.

If it helps I am trying to replace a tweaked BMP tone control and mid control with a 4 band EQ,maybe 3 band we'll see.
I still need to get the notch on the mid control deeper or the peaks on the highs and lows bigger either way.

I tried just using 4 bands from the EQ I have been working on and it just doesnt sound the same,even when compensating the center frequencies.
I still have alot of reading to do I guess.
Any more suggestions greatly appreciated.

[Paul Marossy]

Paul does the one your talking about use the bias to adjust the effect of the filter ?
I ran across an article that suggested that,just no examples...

Not sure to be honest.

[Ben N]

thanks guys thats helpful, the notch on the Mr EQ is close to what I am after ,I just want it adjustable from extreme notch to extreme hump. :icon_confused:I was hoping to have two mid bands with extreme settings,
And I want a low boost/cut and high boost/cut .

I'm sure you'd be better off if one of the ROG guys, or Mark H or RG or some such smart person, chimed in here. But it seems to me fairly trivial to sub a pot for SW2 to make your mid-ad adjustable. As for adding a second mid-band, that would probably be no more difficult than using another inverter section and selecting cap values to taste. I think all the resistor values for the second bandpass could stay the same. I do't know how to make the HP and LP sections do cut as well as boost.

[Brymus]

Thanks Ben ,yeah I was hoping RG, PRR, or anyone who understands this would throw an example up I could tweak to taste or at least use as a starting point.
The Mr EQ is a good example ,so thank you for pointing it out.I at least know it can be done...
I guess the question I should have asked is ,
> Can a CMOS stage be used as a direct replacement for an op amp in filter design ?

[PRR]

> Can a CMOS stage be used as a direct replacement for an op amp in filter design ?

The question I would ask is: WHY?

[Brymus]

> Can a CMOS stage be used as a direct replacement for an op amp in filter design ?

The question I would ask is: WHY?

Because I have a pedal that has unused CMOS stages and I would like to use them to make active filters to replace the passive tonestack rather than adding another IC or transistors.

[PRR]

> construct a low-pass section like this: LINK ...then add a capacitor from "out" to ground,...

That's not an inductor, it is a resistor feeding a super-capacitor (12dB/octave). I can't see how "a capacitor from "out" to ground" does anything except make the chip work harder.

>Is it possible to use a CMOS stage (4049UBE) to make a gyrator for an EQ

Probably; "anything is possible". Everything has a Dual. However the common gyrator topologies (grounded and floating) use high-input low-output impedance amps, often as unity gain followers.

For more fun there are 1-port and 2-port gyrators.

http://techpreservation.dyndns.org/beitman/abpr/newfiles/The%20Gyrator.pdf
http://inst.eecs.berkeley.edu/~ee100/fa04/lab/lab10/EE100_Gyrator_Guide.pdf
http://focus.ti.com/lit/an/slyt134/slyt134.pdf

The CMOS inverter, ah, inverts. OK, use two and it non-inverts. But to get a set gain the input impedance winds up lowish. This and the nature of CMOS leads potential hiss-noise issues.

I'm not seeing a way simpler than adding a transistor.

[Ben N]

There you go: a smart person! The Mr. EQ was designed from the ground up to do what it does, but from my rudimentary grasp on these things, filters, especially steep, multi-order filters, need gain--something opamps supply in abundance, but inverters must be tortured to produce even small amounts of. So, aside from the obvious differences like having single inputs rather than +- inputs, inverters cannot be substituted for opamps because they are not really amplifiers, and certainly not very good ones.

[earthtonesaudio]

> construct a low-pass section like this: LINK ...then add a capacitor from "out" to ground,...

That's not an inductor, it is a resistor feeding a super-capacitor (12dB/octave). I can't see how "a capacitor from "out" to ground" does anything except make the chip work harder.

Ah, you're right.  I think what I should have said was "add a capacitor in series with the input resistor."  That would give you a bandpass-to-ground, which you could use in place of a typical gyrator, although the frequency response would be asymmetrical (6db/oct on the bass side, 12 on the treble side).


I agree with this:

I'm not seeing a way simpler than adding a transistor.

If you just want some adjustable midrange bands, there are other, perhaps more applicable (to CMOS inverters) ways to do it than with gyrators.

[Brymus]

Thanks Paul,those links are good the first was way over my head from the start though.
I think people used much bigger words 60 years ago.

Yeah Alex,I think your right especially after PRR's post.
I am open to any idea on making an adjustable active filter from a CMOS stage or two even,although a recovery gain stage with the passive tonestack is looking like an easier solution.(for me at least)
I just thought it would be better to make use of the un-used stages rather than add more active components.

[PRR]

> should have said was "add a capacitor in series with the input resistor."

I suspected as much, and plotted it. I do get a resonance but weak and asymmetric.

This basically is NOT an inductor because at low frequency the input is R1 plus the load, and at high frequency it is just R1. So the impedance falls, does not rise as an inductor should.

I'm not saying there isn't a good resonance in there. But I can't find it. And I suspect if there is, there's a simpler way.

> gain--something opamps supply in abundance, but inverters must be tortured to produce

Inverters, generally, can be very good amplifiers.

And in this case: He can do a modest Gyrator with a single BJT using only the current-gain of ~~100. The CMOS inverter is not any "infinite gain" thing, but has a voltage gain near 100 and also considerable current gain (maybe over 1,000 in real life).

And to look at it another way.... if he could slice-and-dice his CMOS, he could swap the P and N devices, re-wire as a cathode-follower and a current source. That could be a pretty good gyrator engine for modest audio uses.

We got enough magic. It just isn't packaged conveniently for this use.

If you have a million-dollar contract hanging on getting a gyrator out of a 4049, it would be worth more thought. Everything has a dual, and there probably is a gyrator you can build with only inverters. I won't rack my brain for a one-off with an obvious alternative, but it tickled my mind all night and nothing bubbled out.

What do you really want? Does "active" sound better? Or are you just low on signal? The 4049 can boost-up the weakened signal out of a passive tone-stack. OR.... wire a Baxandall bass/treble stage. Nominally unity gain, uses an inverting amp, which wants gain much more than 10. Just remember that the input impedance, like most tone controls, tends to fall at high frequency.

[JDoyle]

Everything has a dual...

Really? What is the dual for a transformer?

[PRR]

OK, staring at the stars I "got it".

This "is" an inverter-based gyrator; maybe not general, but it gyrates a cap into a coil:



BUT: the inductance and the series resistance vary directly with the inverter gain.

Note that gain=-100 gives 1H and 2K, -200 gives 0.5H and 1K.

Since a 4049's gain is highly variable, with part, temperature, voltage, and signal level, the "inductance" is unpredictable and generally not doing what you want.

Sure, you can "fix" the 4049's gain with two more resistors. But only to a lower value. And the resistor adds more shunt resistance. Raising all resistances will probably raise noise voltage.

I also suspect it will behave horribly for even small signals.

This is just brain-play, not a practical circuit.

R12 R15 are dummy-points for SPICE's tiny brain to probe.

This is just the inductor; a full bandpass needs a cap.

[Nasse]

If it is not perfect, perhaps it has some kool distortion...