Six-stage hex-inverter-based phaser?

[Fancy Lime]

Hi all,

I seem to remember that there are phasers that use the bottom transistors of hex inverters as variable resistors. The logic being that these are supposed to be well matched (laser trimming or some such heathen wizardry). I just cannot remember which models that were. I thought there was a version of the EHX Bad Stone using some old hex-inverters but I cannot find a reference at the moment (only the discreet version with JFETs). Anyone tried this sort of thing with a CD4049 or 4069?

Cheers,
Andy

Edit: never mind, I found it: Bad Stone with a CD4009.

[PRR]

> well matched (laser trimming or some such heathen wizardry).

Not heathen-level technology. More like: the caveman steal the bird's eggs, all the eggs are the same size, more or less. While any chip-fool can step-and-place 'identical' masks on a chip and make 'identical' FETs, there has been no long-term need for such a chip. Multiplexers? No, they quickly grew logic and buffers and high prices. But the '009 is reasonably matched by layout and cheap as chips.

Was it Anderton or Simonton who had a hex VCA the same way?

[Ben N]

The logic being ...

I see what you did there.

[Mark Hammer]

There was a 4049-based six stage project in Electronics Today, way back when.  Mike Irwin told me that it was quite noisy, because use of the inverters required attenuating the input significantly, and then boosting on the output to provide effect/bypass level matching.  I seem to recall he had a modest fix for that, but in its absence we're stuck with the stock circuit.

[Fancy Lime]

Yes, those pesky body diodes... No way to get around them on a 4049 or 4069, so signal attention is about the only option there is. Or companding with a NE570 or so. A 4007 would give us two N-FETs on which we can sidestep the body diodes but I have no idea how well they are matched. I was wondering if we could use a 4016 or 4066 quad switch instead. The switch channels are matched as per datasheet and I think it may be possible to tie all controls to V/2 and then use the inputs for signal and the outputs as controls. Not sure though, need to think some more. Depends on some details that the datasheet refuses to tell us. Has anyone tried that?

[ElectricDruid]

There was also a 4007UBE-based phaser in Penfold's "More Advanced Electronic Music Projects" book. Each pair of phase-shifters used one 4007, using two of the N-channel MOSFETs. There is another N-channel MOSFET and three p-channel MOSFETs unused in each chip.

I built the circuit years ago, but never got it working. It was ok with two stages, but increasing it to four or six stages stopping it from working properly. I never did find out why. I've still got it in a box somewhere, so perhaps one day I'll take another look.

[puretube]

(Heavy) AC-feedbacking of the *to-be-shifted* signal from the FET`s Drain to its Gate (where the DC controlvoltage also resides) helps linearizing the *virtual resistance* of the FET to ground, and extends the acceptable signal-drive voltage range.

Done it with 4069 in discreet-transistor Phasers & Filters, where the signal-voltage needed no attenuation.


Mike Irwin mentions (RC-)feedback it in this advanced article here, too:
http://www5b.biglobe.ne.jp/~houshu/synth/mike_irwin/cmos-vcr.gif
e.g. in this thread:
https://www.diystompboxes.com/smfforum/index.php?topic=44515.0

Backgrounds:
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3131312A&KC=A&FT=D&ND=3&date=19640428&DB=EPODOC&locale=en_EP (fig.4)
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3213299A&KC=A&FT=D&ND=4&date=19651019&DB=EPODOC&locale=en_EP (fig.4)
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3452290A&KC=A&FT=D&ND=3&date=19690624&DB=EPODOC&locale=en_EP (fig.1)
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3577019A&KC=A&FT=D&ND=4&date=19710504&DB=EPODOC&locale=en_EP
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3702443A&KC=A&FT=D&ND=4&date=19721107&DB=EPODOC&locale=en_EP (fig.2)
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3798562A&KC=A&FT=D&ND=4&date=19740319&DB=EPODOC&locale=en_EP (fig.7)
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3835418A&KC=A&FT=D&ND=4&date=19740910&DB=EPODOC&locale=en_EP

An up-to-date copy of National Semiconductors Application Note 129 (orig. 1975):
https://www.onsemi.cn/pub/collateral/an-6603cn.pdf

An old thread, similar content:
https://www.diystompboxes.com/smfforum/index.php?topic=88745.0
For details about the quad-fet mentioned there (AM9709CN), go to page 160 of the handbook you can download here:
https://usermanual.wiki/Document/1977NationalFETDatabook.1783250853/help
The circuit in question can also be found in this nice 40 yr. old book of "Modern Circuits" ...
/Markus_ModernElectronicCircuitsR%20[Electrical4u.ir]/Markus_ModernElectronicCircuitsR%20[Electrical4u.ir].pdf]http://dl.electrical4u.ir/links/Markus_ModernElectronicCircuitsR%20[Electrical4u.ir]/Markus_ModernElectronicCircuitsR%20[Electrical4u.ir]/Markus_ModernElectronicCircuitsR%20[Electrical4u.ir].pdf (p.639)

[Mark Hammer]

Leave it to Ton to use the search function like a pro, and remind us that it's all been done before. 

[puretube]

Leave it to Ton to use the search function like a pro, and remind us that it's all been done before. 

Those patents were easier to find than the box with my scribbles (& breadbord & perfboard-prototypes) from 7 years ago of the 2-transistor filter-stages, that could be configured to be HiPass, LoPass, AllPass, all using 4069 "HEX-VCR"s


or 2 for BP (Wah),


or in combination with a 3-transistor VCA into a pot-sweepable filter to seamlessly voltage-control the filter-characteristic from Lo- through All- to Hi-Pass ...

[puretube]

"Hammond" mentions the drain-to-gate feedback here in 1972:
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=3737555A&KC=A&FT=D&ND=3&date=19730605&DB=EPODOC&locale=en_EP
(okay - no cmos bodydiode involved ...)

[anotherjim]

The body diode thing is an advantage for the 4007 chip as they are turned away into the substrates to a supply rail - but I'd have thought they do in the hex inverter chips too and it's the protection diodes effectively on all 3 ends of the MOSFETs that might trouble you, but the 4007 has those too but it has a few independent transistors and yes, you might only use one transistor in the 4007 for that reason.
The 4049 has weak positive protection for the gate inputs so it can be driven from full CMOS levels while running on 5v to suit TTL on its outputs which can help some with isolating the 6 sections if you only want the NMOS devices.

[Crater Dylan]

It's fun that I found out about this recent thread just as I was wrapping up my last project, a bunch of CMOS phasers from 2 to 6 stages! Since it's a lengthy article with pictures and sounds, I'll just link to the site I made for these things https://bentfishbowl.wixsite.com/electronics/post/pup-a-host-of-perfectly-usable-phasers
I hope it's just what you were looking for.

[Fancy Lime]

It's fun that I found out about this recent thread just as I was wrapping up my last project, a bunch of CMOS phasers from 2 to 6 stages! Since it's a lengthy article with pictures and sounds, I'll just link to the site I made for these things https://bentfishbowl.wixsite.com/electronics/post/pup-a-host-of-perfectly-usable-phasers
I hope it's just what you were looking for.

Wow, nice! That is exactly what I was originally looking for. Thanks for putting in the work and for sharing. I need to try this as soon as I have a little time.

Andy

[StephenGiles]

Looks very interesting!

[Mark Hammer]

A very auspicious debut here.  Nicely documented, AND representative samples.  I'm impressed...and grateful.  Welcome!

[Crater Dylan]

A very auspicious debut here.  Nicely documented, AND representative samples.  I'm impressed...and grateful.  Welcome!

Thanks Mark. Can't say I'm really new here, even if haven't posted much. You can check some of my other projects too, or maybe I'll put some time into translating them into posts here if that's welcome. I'd sure enjoy the discussion.

[Ben N]

A very auspicious debut here.  Nicely documented, AND representative samples.  I'm impressed...and grateful.  Welcome!

Thanks Mark. Can't say I'm really new here, even if haven't posted much. You can check some of my other projects too, or maybe I'll put some time into translating them into posts here if that's welcome. I'd sure enjoy the discussion.

Dylan's been posting quite a few great original projects over at the other forum (as "Dylan159"), all beautifully documented and explained. Very worthwhile stuff, IMHO.

[Crater Dylan]

A very auspicious debut here.  Nicely documented, AND representative samples.  I'm impressed...and grateful.  Welcome!

Thanks Mark. Can't say I'm really new here, even if haven't posted much. You can check some of my other projects too, or maybe I'll put some time into translating them into posts here if that's welcome. I'd sure enjoy the discussion.

Dylan's been posting quite a few great original projects over at the other forum (as "Dylan159"), all beautifully documented and explained. Very worthwhile stuff, IMHO.

That would be me thanks. Honestly, I still don't know where my kind of content would go here (we're in "Building your own stompbox - post questions about building your own stompboxes" now), or I'd appreciate the useful insights I've seen from people here in other occasions. Cutting this short because I know I'm bordering on OT.