Author |
Message |
Thomas Henry
Joined: Mar 25, 2007 Posts: 298 Location: Southern Minnesota
Audio files: 2
|
Posted: Fri Jul 13, 2007 5:48 pm Post subject:
An XR-2206 Non-Hint Subject description: ...or what not to do... |
|
|
In my VCO Chip Cookbook, I mention that the four quadrant multiplier in the XR-2206 would be lousy as a VCA. In fact, I had a chance to find out just how bad it is this past week.
I've developed a rather unusual electronic percussion synthesizer with no commercial equivalent, and among the chips in it is the XR-2206. Somewhere along the way in the design cycle I thought I might use the internal multiplier as a VCA and save a chip or two. I figured, even as bad as I knew it to be, it would be okay for percussive sounds.
In a nutshell, no way! Even for simple VCA applications, this chip stinks. There are several problems. First, "off" means down -55dB which is still very audible. Second, it's hard to find "off" in the first place---it corresponds to a control voltage around 1/2V+, meaning that a trimmer will be needed. Third, even in the "off" position, the amount of background hiss is quite objectionable.
Like I say, I thought I might get away with it in this situation, but it's clear the unit is abysmal even as a simple VCA for percussion.
However, the story has a good ending. I wound up using the multiplier as a ring modulator, and in this context it's great! I probably wouldn't have considered that option before, but didn't want to waste part of the chip.
So again, in the XR-2206: VCA no, ring modulator yes. I hope this saves someone else a wasted evening pursuing that which doesn't exist! By the way, the new circuit is the most difficult one I've ever pulled off. It took a solid 14 twelve-hour days to get it right. See what I do for summer vacations!
Finally, I figured out another way to put the XR-2206 under exponential control. (There's one version in the VCO Chip Cookbook that is particularly powerful, but this new method is ideal for non-critical applications like tuning the drums via a keyboard). When I get a chance, I'll draw it up and share it here.
Thomas Henry |
|
Back to top
|
|
|
StephenGiles
Joined: Apr 17, 2006 Posts: 507 Location: England
|
Posted: Fri Jul 13, 2007 11:38 pm Post subject:
|
|
|
That sounds terrific Thomas, I've had a 2206 stashed away since the early 80s which I bought for an Elektor circuit. Do you see anything of Craig Anderton these days? |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Jul 14, 2007 9:06 am Post subject:
|
|
|
I'll drop an XR-2206 hint: that VCO Chip Cookbook is great - I've even found myself actually using the easy math in it <gasp>. It's a real lifesaver to have around when working with these chips.
Exar, Philips and Intersil all missed the boat when they were writing their apps for these ICs - if they had Thomas writing them, I think you'd find NE566s in the checkout counter impulse buy section, along with the chewing gum and candy bars.
Sweet!
Cheers,
Scott _________________ My Site |
|
Back to top
|
|
|
State Machine
Janitor
Joined: Apr 17, 2006 Posts: 2809 Location: New York
Audio files: 24
|
Posted: Sat Jul 14, 2007 12:10 pm Post subject:
|
|
|
I find the following chips to be the best for building a VCA:
In order of my preference
1) SSM2164 (VCA)
2) THAT2180 (VCA)
3) LM13700 (OTA)
4) CA/LM3080 (OTA)
The first two amplifiers are inherently LOG responsive but can be easily be made to respond LINEARLY. I know for sure about the SSM2164 because I have hand rolled my own VCA modules using these chips and they have +20 to -100 dB dynamic control! The second two are, and you all know this, are linear within a small operating range but can make excellent VCA's.
I think the SSM2164 is by far the best, IMO, VCA chip out there right now and it has four sections or "gain cells" built in. With one SSM2164 IC, you can build 4 LOG VCA's or 2 LIN VCA's since you have to use two gain cells in the feedback path of the other. By putting the gain cell in the feedback path, you essentially get a amplifier with a output response of (10^(Log x)) with essentially distills down to "X" (example 10^Log100=100) thus the function becomes linear by eliminating the Log function. I can supply the derivation if anyone wants it ... just PM me.
I agree though, the XR2206 really is cool as a ring modulator !!
I may be a bit biased in my opinion but the VCA Chip Handbook really does have a very good section on the XR2206 IC
Bill |
|
Back to top
|
|
|
Thomas Henry
Joined: Mar 25, 2007 Posts: 298 Location: Southern Minnesota
Audio files: 2
|
Posted: Thu Jul 26, 2007 7:24 pm Post subject:
|
|
|
StephenGiles wrote: | That sounds terrific Thomas, I've had a 2206 stashed away since the early 80s which I bought for an Elektor circuit. Do you see anything of Craig Anderton these days? |
Craig is the person who gave me my start in electronic music writing. All told, he probably edited some forty articles of mine, and always very generously gave me hints and encouragement. And something probably no one knows, he's also the reason I ended up as a college professor, which radically changed my life. For in fact, he wrote a glowing letter of recommendation for me that turned the tide when first getting into academia.
And now the surprising answer to your question, Stephen. No I don't see much of Craig these days, for the very simple reason: we've never met! I've always considered him a great friend and colleague, but our communications have been exclusively through the US mail!
In fact, I don't believe I've met any synth people in person. But that'll change in a couple weeks when I undertake a trip to the hotbed of creationism in the US.
Thomas Henry |
|
Back to top
|
|
|
para
Joined: Oct 16, 2006 Posts: 276
|
Posted: Fri Jul 27, 2007 4:33 am Post subject:
|
|
|
"the hotbed of creationism" please be careful we don't want to lose you.
steven |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Jul 28, 2007 3:45 pm Post subject:
|
|
|
The welcoming committee has fancied up the scaffolding with some purty red, white and blue bunting, the Ladies Temperance Union has generously donated 123 torches, and the Farmer's Grange contributed not only two and a half bushel of chicken feathers, but also a gallon and a half bucket of tar. Old man Jenkins even pulled a rail out of his lower pasture fence for the occasion. Col. Brownback should be on time to represent the side that is all righteous and true, and Praise-The-Lord Unified Brethern has agreed to loan their King James for all the evidence required.
This is going to be the biggest thing to hit these here parts since Thomas A. Edison dropped by to toast that elephant back in ought 3 (do that with AC, Tesla!).
This is gonna to be a hootenanny! _________________ My Site |
|
Back to top
|
|
|
Dan Lavin
Joined: Nov 09, 2006 Posts: 649 Location: Spring Lake, Mi, USA
Audio files: 21
|
Posted: Mon Nov 05, 2007 5:19 pm Post subject:
|
|
|
Thomas (or anyone else who has tried sync with the XR2206),
Can you explain the difference between sync with the XR VCO (page 9 of the 21st Century Synth book) and figure 89 in the VCO Chip Cookbook? Obviously the XR circuit is easier to implement (less parts) and I know figure 89 is the traditional way to do hard sync. I guess more importantly, does one sound better (more intense) than the other? |
|
Back to top
|
|
|
Thomas Henry
Joined: Mar 25, 2007 Posts: 298 Location: Southern Minnesota
Audio files: 2
|
Posted: Mon Nov 05, 2007 6:33 pm Post subject:
|
|
|
Aha! Someone finally noticed. I think that the XR sync is one of my more clever ideas, if I may be so immodest.
The traditional method is self-explanatory. The switch shorts out the cap.
The XR method uses the internal mechanism that switches between timing resistors at pins 7 and 8. Pin 7 is a virtual resistor, being the main current control input. But pin 8 is locked at 1k (the smallest legal value); it guarantees the effective period of oscillations is way shorter than anything pin 7 can come up with. The upshot is that the timing cap is also shorted out lickety split (perhaps more than once, but you won't hear it).
Truthfully, due to the timing constant imposed by the 1k, the waveform probably isn't perfect at sync time. However, your ears will not detect the difference, I guarantee it. For oscilloscope art the circuit would create noticeable visual distortion, but for electronic music you'll never know you aren't hearing true hard sync.
It sounds good, so why open a can of beans with a stick of dynamite? As if you didn't know that was coming...
Thomas Henry |
|
Back to top
|
|
|
Dan Lavin
Joined: Nov 09, 2006 Posts: 649 Location: Spring Lake, Mi, USA
Audio files: 21
|
Posted: Mon Nov 05, 2007 8:15 pm Post subject:
|
|
|
Interestingly enough, I was going to ask if the XR-VCO approach was an "aha" moment! Fortunately, I didn't steal your thunder! |
|
Back to top
|
|
|
|