Monday, August 5, 2019

Attenuverter Part I: Prototyping the SEM way to do this--Pot with Center Tapped Ground

All hail the mighty attenuverter!  This time I build one--and I break one of my cardinal rules and strip board most the prototype--spent 2 days doing that--UGH!  But now it all works, sounds good I think, and should lead to an interesting PCB project, maybe this month? We'll see!

Here we go:

Backroundverter: I laid eyes on my first attenuverter in an Oberheim SEM (circa 1977?). I was barely old enough to drive and also lucky enough to be doing synth programming for a local recording studio; they had an unbelievably cool Obie with x8 SEMS, customized to chooch on 2 banks of 4 SEMS, giving me 4 stacked polyphonic voices. Wow! Mono was still king back then for the most part; the Obie 8 voice sounded frigging GREAT!

The idea of choosing between + (normal) and - (inverted) mod for a CV was super cool; especially for the VCF when set to bandpass, highpass or notch. The SEM synth sounded, well, they are SEMs! Damn, it's freaking Oberheim!!!

Be gentle--it's my First Attenuverter: The SEM is chock a block full of 'em!

How it works: an attenuverter is basically a single pot mixer. Noon is zero gain; CCW is 0 to 100% of one signal, CW is 0-100% another.  Most often the CCW signal is an inverted form of the CW signal or rarely, visa versa.

You can find euro attenuverter modules all over: for instance, here, here, and here, and of course MATHS has one--yep, 40+ years later these puppies are everywhere!

The basic circuit fragment looks like this--it's easy: one op amp, one pot, and a few resistors:



I have created an online SIM of this circuit fragment here if you want to mess with it. Or lay it out on a breadboard, this is easy stuff.

But be warned: on the bench the above fragment doesn't sound as good as the SEM's. There is bad control w/ this fragment close to noon, and maybe from 10AM to 2PM; it's like someone screwed up the taper--you get very little, then you get too much. Good enough? Nope.

There is a blog post that hops up this circuit fragment here to address: here, before going down the SEM road, check that out?

But I took a different and more difficult approach, since I wanted to 100% nail Oberheim's take on the attenuverter. Why ever make things easy?

The SEM schematic is online here; from that, turns out Oberheim uses (used?) a relatively obscure part for the SEMs attenuverters: potentiometers with center tap ground.  This is a normal pot but with an extra leg halfway down the pot's winding. Using this "fourth leg" you can hook the middle of the winding to ground or whatever else you want and then wipe to your heart's content:

Take a deep breath:ALPS RK11K114F30C0B104 WITH CENTER TAP



Having never heard of this part, I confirmed with Tom Oberheim via email (and yes, he emailed me back! Yeh!) that a CT-ground pot is key to the classic SEM's attenuverter. yes it is!

And good news: Mouser has an ALPS version here (data sheet is here) so I bought a few.


The ALPS center tap ground pot!


Bread boarding this part on my trusty Radioshack learning lab thingy, yep, works as advertised.

Basic layout is this:

Front view of the CTpot

If you hook the center tap to ground, as you approach noon with the pot's wiper more and more of the + in is shunted to ground, same idea for in - (which I mistakenly label as the other IN +).

From here it's an easy matter of having an input buffered and inverted and presenting each signal to the appropriate lug:



A and B are op amp buffers, B inverts while A doesn't. I used some jelly bean SMT boards I created for other projects, you can find details here, but really any buffer/inverter will work, and there are lots of variations to choose from.

I also added Bias Offset/Amplitude modulation independently for each side (the CT pot makes that easy!); more information on that idea is here.

And the output follower/buffer--which arguably isn't needed at all if your design disallows the op amps to short to directly ground--is the same unity buffer amp you see everywhere:





Building the Prototype: I was in a rush to get my Center Tap ground pot into play so I went against one of my cardinal rules and strip boarded this prototype (but isn't everything we DIYers build a prototype?). This led to the usual flubs, screw ups, cursing, fear and loathing, and reminding myself how much I hate strip board, but I did finally get it to work.

I'm sick of the thing now though, and probably won't create a front panel for this messy bunch of wires and daughter boards; rather, come up with an improved PCB and front panel and go from there.
Update: yes I did! Long live the Mike Matthews antistrip boarding society! Go to part II of this post here.)

If you want me to post more details about the this circuit I can; comment below.

OK, some obligatory build photos:

The pot/knob "tall trimmer" thingys are available from Modular Addict here. The red boards are SMT op amp inverting buffers, more info here.



Finished!  That's a messy module, but finally, it attenuverts.

OK, now that my strip board catastrophe works, the attenuvaterization--I doubt there is such a word--is satisfying. Onward! I am not done with center tap pots by any means!

Update 9-7-19 PCB of this same design is back, stuffed, and works!  Post is here.

Until next time, don't ehtaerb the semuf's.

No comments:

Post a Comment

Minimalist Atmel 328 Development Board--How to Build One

Hello again:  if you've been following the last few posts you see that I am trying to use Arduino's IDE a lot less, and learning to ...