MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Anyone have experience with the V2162 dual VCA chip?
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author Anyone have experience with the V2162 dual VCA chip?
CLee
Looking for someone with a bit of experience with the chip.

I'm working on a design that will be using the THAT2162 dual VCA. I got a couple Coolaudio clones, V2162 to do some breadboarding. They're cheaper and bigger so easier to solder to a SMT to DIP adapter.

I can't get an undistorted output from it. The output was basically rectified (the positive going side of the waveform was flattened) and I finally fixed that by running a 51R resistor from the SYMMETRY pin to ground. I noticed the THAT docs say it should be unconnected if you're not using it but the Coolaudio docs show it always connected and the simplified internal schematic on the docs are different between THAT and Coolaudio.

So, anyone have experience with the chip. Seems it's not a direct clone. I can't get the basic data sheet VCA schematic to work without distortion, now with the SYMMETRY pin going to ground through a 51R the negative going output is nonlinear, curved and attenuated.
J3RK
It sounds like it's going into positive gain (rather than unity and below). Except the negative symmetry thing... Have you checked your CV circuit? Maybe there's an offset creeping in somewhere.

I don't have direct experience with this IC, but I've run into this before with THAT 218x chips.

I'm using the SSI2164 for all of my VCA needs now, so I haven't looked at the 218x or 2162 in quite a while.

Anything different on your output, or just a standard current to voltage converter?
CLee
Hey thanks Dustin, but I don't think it's a CV problem. The distortion is in the AC signal output. With a Triangle wave input the top, positive portion looks good but the lower, negative, part of the TRI wave is distorted. Its consistent throughout various amounts of attenuation.
CLee
guest
do you have a schematic of what youve built? if the impedance of the EC drive is too high, you can see issues like this.
CLee
guest wrote:
do you have a schematic of what youve built? if the impedance of the EC drive is too high, you can see issues like this.


Impedance should be low, driven directly from an op amp out. The circuit looks a bit redundant but I need the 0-5v CV somewhere else too

CLee
The rest is basically this but the 510K resistor and pot aren't included

guest
are both the opamp and 2162 being run at +/-12V rails? and whats the volts/div on your scope shot above? and whats your input signal level? the 20k is typically used with 1Vrms signals, not modular level signals.
CLee
Yes +/-12v supply and the scope scale is 1v/div

I think it's a current drive issue. I replaced the 2 20K resistors in the test circuit with 100Ks and the waveform is much better.

Interesting because the folks at THAT never mentioned the gain cell not being able to handle +/-5V audio at unity gain. The Coolaudio data sheet says the max current is 1.4ma which should have been ok
guest
the 1.4mA is a hard uppper limit, typically its best to keep things down at the 100uA range for lowest distortion. the SNR is a bit better at higher drive, though. but, with +/-2.5V/20k = 125uA, it should be ok.

when the gain goes up, the output drive needs to be higher, which also causes distortion. the max amplitude at either the input or output needs to be kept at a reasonable level. im assuming the above scope shot is for 0db gain, so im assuming thats not the problem here either.

you might give a try with the symmetry adjust. it can have a big effect.
CLee
guest wrote:
the 1.4mA is a hard uppper limit, typically its best to keep things down at the 100uA range for lowest distortion. the SNR is a bit better at higher drive, though. but, with +/-2.5V/20k = 125uA, it should be ok.


THAT seemed to think that would be OK, that I could probably get away with the full +/-12v range at unity gain with a 20K resistor. I don't have genuine THAT2162s to compare yet, but no need to push it anyway.

guest wrote:
when the gain goes up, the output drive needs to be higher, which also causes distortion. the max amplitude at either the input or output needs to be kept at a reasonable level. im assuming the above scope shot is for 0db gain, so im assuming thats not the problem here either.

Yes, unity gain. Like I said, changing the 2 resistors from 20K to 100K greatly improved the distortion. Work will keep me from doing much testing for the next few days...

guest wrote:
you might give a try with the symmetry adjust. it can have a big effect.


Tried the trim before switching 20Ks to 100Ks and it didn't do anything. I'll need to try again now
guest
well, definitely keep us posted. it will be good to know how the cool audio parts compare to the THAT parts. the only other things i can think of are:

1. try shorting the symmetry pin to ground.
2. try connecting the other VCA in the package to ground - maybe its effecting it somehow.
J3RK
This is just a quick side note, and I'm sure you have reasons for going with the 2162, but from my experience, the SSI2164 (Sound Semiconductor Version) is absolutely amazing, very easy to set up, and performs very well. I'm not sure if it matters or not, but the minute I tried it in a couple of applications (exponential conversion, filter cells, and linear VCAs) I stopped using all THAT VCAs completely, and even cut back on my LM13700 use a bit. It took... ...a decade I guess? but I'm a total convert now. Just a recommendation if you get tired of messing with the 2162 w00t
CLee
J3RK wrote:
This is just a quick side note, and I'm sure you have reasons for going with the 2162, but from my experience, the SSI2164 (Sound Semiconductor Version) is absolutely amazing, very easy to set up, and performs very well. I'm not sure if it matters or not, but the minute I tried it in a couple of applications (exponential conversion, filter cells, and linear VCAs) I stopped using all THAT VCAs completely, and even cut back on my LM13700 use a bit. It took... ...a decade I guess? but I'm a total convert now. Just a recommendation if you get tired of messing with the 2162 w00t


Yes, I like the 2164, I've used it too. In this application I'm going to eventually need the EC+ and EC- dual control ports though.
J3RK
CLee wrote:
J3RK wrote:
This is just a quick side note, and I'm sure you have reasons for going with the 2162, but from my experience, the SSI2164 (Sound Semiconductor Version) is absolutely amazing, very easy to set up, and performs very well. I'm not sure if it matters or not, but the minute I tried it in a couple of applications (exponential conversion, filter cells, and linear VCAs) I stopped using all THAT VCAs completely, and even cut back on my LM13700 use a bit. It took... ...a decade I guess? but I'm a total convert now. Just a recommendation if you get tired of messing with the 2162 w00t


Yes, I like the 2164, I've used it too. In this application I'm going to eventually need the EC+ and EC- dual control ports though.


Ah! Makes sense. I did a pan control using both ports for an output mixer module I was working on. I could never get the response exactly the way I wanted, but it was close, and I liked the use of the control ports to do it. I almost forgot about that. lol
guest
is there a difference between having +/- control ports, and using a differential amp before the single sided control port of the 2164? i know the 2162 has the SYM pin, which can be used to do funky stuff: making positive and negative inputs attenuate differently (adds an offset as well), use it as an exponential converter, etc.
J3RK
guest wrote:
is there a difference between having +/- control ports, and using a differential amp before the single sided control port of the 2164? i know the 2162 has the SYM pin, which can be used to do funky stuff: making positive and negative inputs attenuate differently (adds an offset as well), use it as an exponential converter, etc.


I'm not sure if there's an actual behavior or performance difference, but it was interesting to try using both ports.

I did something that was roughly based on the following idea, but in the context of a two channel stereo mixer with VC panning (so more VCAs):

http://www.cordellaudio.com/preamplifiers/vca-based.shtml

(It's a little way down the page.)

It would have been MUCH easier to just use the 2164, but I was having fun messing with the 2180s that I had laying around for several years collecting dust. hihi The parts count ended up being pretty ridiculous though to get everything scaled and offset. The PCB was 4" x 6" for a simple VC mixer. I ended up just going back to my original output mixer, and adding LM13700 VCAs in front of it.

I'm actually planning a similar module using the SSI2164 (as I'm doing a complete refresh of all of my functions.) I'm pretty sure it's going to be the simplest of the bunch, and fewest parts needed.
guest
id be interested in hearing your thoughts on the new SSI2164 as an exponential converter (assuming your going to do that). i went back a few months ago to try and get better results from my initial tests on the SSM2164, and i kept getting too much crossover distortion. granted, it was totally usable, just not nearly as flat as a standard matched pair.
J3RK
guest wrote:
id be interested in hearing your thoughts on the new SSI2164 as an exponential converter (assuming your going to do that). i went back a few months ago to try and get better results from my initial tests on the SSM2164, and i kept getting too much crossover distortion. granted, it was totally usable, just not nearly as flat as a standard matched pair.


I'm using it on a few VCO designs currently. I haven't done any hugely scientific measurements though. I'm getting decent tracking, but I'm not sure how the tempco aspect is yet. I'm basically using the SSI datasheet 3 VCA method. I'm also using low drift op amps in the CV section as well as integrator, and using a decent (if a little slow) comparator. This may help a bit.

I'm not sure that it's perfect, or performs better than other methods, but it does remove the need for a PTC or 3046-style heater setup, is readily available, and doesn't take up too much board space. I did it more for these reasons (and that I'm working on a collaboration where these were THE most important reasons) than performance.

As you say, it's quite usable, and simplifies things quite a bit. I'm using this chip all over now, so it cuts the BOMs back for the new set of PCBs I've been working on.

I'll try to get some better measurements once the system I'm working on is complete. (next couple of weeks) I will say that the one VCO I do have built and running tracks well (around 6ish octaves I'd say) and stays pretty well in tune (against other gear that I've been controlling it with). Just from a subjective, in-use perspective.
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group