2164 VCA

[Gobotak]

I've been browsing the forum for a while but this is my first post so I'll start by saying hello to everyone.

I have a Coolaudio V2164D on my breadboard but am having some issues with it.  I'm using the schematic in figure 6.4 of the datasheet since ideally I would like to use a single 9V supply: https://www.coolaudio.com/docs/COOLAUDIO_V2164MD_DATASHEET.pdf

I am getting control of the volume but there is a lot of high frequency loss (like rolling the guitar tone knob all the way down) and a lo-fi output for lack of a better term.

The changes I have compared to the schematic example are 33K resistors instead of 30k, 470R resistor instead of 500R, and TL072 instead of the OP176 and OP482. 

Things I have tried are using a +/- 18V supply from a LT1054 using figure 6.1, buffering the input, using 20K resistors instead of 33k, and using a guitar with active pickups and one with passive pickups.

Has anyone used this chip with guitar signals?  If anyone has suggestions to point me in the right direction I would greatly appreciate it!

[Eb7+9]

For direct guitar use replace input 30k by 220k and Rb by 250k pot ...

otherwise you can leave things as is and stick a noisy buffer at the front

[idiot savant]

The Coolaudio V2164 is a clone of the old SSM2164 https://www.analog.com/media/en/technical-documentation/data-sheets/SSM2164.pdf

The new version from SSI is much nicer https://www.soundsemiconductor.com/downloads/ssi2164datasheet.pdf

I've used it pretty extensively for personal and commercial projects. It's used in synth projects, and you'll find a bunch of info over on the muffwiggler forums. I primarily use it with linearized CV inputs. And I usually use a bipolar supply, so I don't have much to suggest for single supply use.

A couple quick notes: Input current headroom is roughly 0.5mA on the Coolaudio, and old SSM chips, so you'll get the best noise performance with the smallest input(V to I) and output(I to V) resistors possible. For a 9V supply you can go as low as 10K just fine. Obviously you'll want a buffer in front for this case. Another place to experiment is using the mode pin. In my experience, class A isn't worth it. You'll get good performance with the mode pin open. Lastly, have a look through the new SSI datasheet, it's got more examples and useful info, though not all of it is relevant to the old chips.

[ElectricDruid]

I've been browsing the forum for a while but this is my first post so I'll start by saying hello to everyone.

Hi Gobotak!

I have a Coolaudio V2164D on my breadboard but am having some issues with it.  I'm using the schematic in figure 6.4 of the datasheet since ideally I would like to use a single 9V supply: https://www.coolaudio.com/docs/COOLAUDIO_V2164MD_DATASHEET.pdf

I am getting control of the volume but there is a lot of high frequency loss (like rolling the guitar tone knob all the way down) and a lo-fi output for lack of a better term.

The output I-to-V stage gives a rolloff at 48KHz with 33K/100p values, so that *should* be ok. Check you didn't use 100n instead!

The changes I have compared to the schematic example are 33K resistors instead of 30k, 470R resistor instead of 500R, and TL072 instead of the OP176 and OP482.

These are pretty standard tweaks. I often use 470R/560p for the network on the inputs too, and TL072 for basically everything.

Things I have tried are using a +/- 18V supply from a LT1054 using figure 6.1, buffering the input, using 20K resistors instead of 33k, and using a guitar with active pickups and one with passive pickups.

I can imagine using the LT1054 for the supply might add noise, but if you got the same problems with the single supply version, then that can't be the whole story.

Has anyone used this chip with guitar signals?  If anyone has suggestions to point me in the right direction I would greatly appreciate it!

The input impedance is quite low, so I'd put a good buffer in front of the VCA. The control input also has a low impedance (about 5K iirc) and any noise that goes in there modulates the signal passing through, so make sure that the *control* signal is quiet too.

Other than that, I've used the chip quite a lot and always found it pretty easy to get going.

[Gobotak]

Thanks for the help and suggestions!

I have it set up now with a buffer going before the chip still with 33k resistors but no luck yet.  If I probe the output pin of the vca I get a decent sound but still have high frequency loss.  When sent through the output op amp the signal is very distorted even though the overall volume is not louder compared to my input signal.  I've triple checked values and tested the op amps as simple buffers and everything seems fine on that end.

The input impedance is quite low, so I'd put a good buffer in front of the VCA. The control input also has a low impedance (about 5K iirc) and any noise that goes in there modulates the signal passing through, so make sure that the *control* signal is quiet too.

Currently my buffer is an inverting but I have tried a non inverting one also and from what I remember the results were the same between the two.  I'm not sure exactly what you mean by a quiet control signal.  Eventually I plan on using a microcontroller but right now I am using a potentiometer voltage divider ranging from 4.5V to Vcc.

In figure 6.4 and 6.1 it looks like the the 500R/560p is connected to ground. Since I am using a single supply should this be 0V or 4.5V?  It seems to be better at 4.5V but am wondering if this is the proper way.

[ElectricDruid]

I have it set up now with a buffer going before the chip still with 33k resistors but no luck yet.  If I probe the output pin of the vca I get a decent sound but still have high frequency loss.  When sent through the output op amp the signal is very distorted even though the overall volume is not louder compared to my input signal.  I've triple checked values and tested the op amps as simple buffers and everything seems fine on that end.

The VCA output is a current output, so what you'd hear there connecting an audio probe would probably depend on a lot of factors and I wouldn't be keen to draw too many conclusions from that.

Currently my buffer is an inverting but I have tried a non inverting one also and from what I remember the results were the same between the two.  I'm not sure exactly what you mean by a quiet control signal.  Eventually I plan on using a microcontroller but right now I am using a potentiometer voltage divider ranging from 4.5V to Vcc.

By "quiet control signal" I mean one that doesn't include any noise. For example, if you're using a DAC which is referenced to the same +5V supply as a microcontroller, there's quite likely to be a significant amount of digital noise on the DAC reference, and hence on the DAC output. That noise is then modulated with your audio as it passes through the VCA (Amplitude modulation - so we get sidebands!).

A pot should be quiet except when it's moving. I wouldn't expect silence when tweaking it, but it should be good in a fixed position. Microcontrollers and DACs controlling analog circuits is the delightful and frustrating area known as "mixed signal design"! In my view, it's where the really fantastic stuff is, the literal "best of both worlds", but it's hard to do without also getting the problems from both world too!

In figure 6.4 and 6.1 it looks like the the 500R/560p is connected to ground. Since I am using a single supply should this be 0V or 4.5V?  It seems to be better at 4.5V but am wondering if this is the proper way.

Since there's a capacitor which blocks the DC level, it doesn't really matter in theory. The V2164 datasheet actually shows it going to Ground rather than 4.5V.

[Gobotak]

I now have the chip working but have some questions regarding the control port.  Currently it works great with a pot varying the voltage from 9V to 4.5V with 4.5V being max volume and 9V attenuated.  Looking at the bottom of page 3 from this datasheet: https://www.soundsemiconductor.com/downloads/ssi2164datasheet.pdf  it looks like the control port should range from -660mV to 3V.  This would be ideal for me since I am hoping to eventually use a microcontroller.  Since I am not using a dual rail supply is there anything I need to do to properly scale the control port voltage?  Am I just misunderstanding the datasheet?

[ElectricDruid]

Yes, both the V2164 datasheet you posted originally and the SSI2164 datasheet give a figure of -33mV/dB for the control port.

This means that for the given control range of +20dB to -100dB, the voltage at the port goes from -660mV to 3.3V.
If you're not concerned about adding gain, you can ignore the -660mV/+20dB to 0V/0dB part and use just the 0V/0dB to 3.3V/-100dB.

If you're running from a 9V single supply as you describe with 4.5V playing the part of the "0V" midpoint voltage, you're feeding the control from 0V to 4.5V, so the upper 4.5-3.3=1.2V won't really do anything audible. The signal will already be hard off, so you won't hear any further reduction. You could add a resistor above the pot to limit the voltage to the useful 0V -> 3.3V  range.

[Gobotak]

Thanks for the explanation! It is making much more sense to me now I think.  I'm going to try to get it going with a microcontroller next.  The plan is to use a smooth pwm output.  With a dual rail supply I'm hoping I will be fine since the max output I get from my microcontroller is 3.3V.  I am also wondering how to go about using a single 9V supply.  Would there be a way to shift the pwm signal to range from 4.5V to 7.8V?  Or is there a different way to reference the pwm signal to the 4.5V virtual ground?

[Mark Hammer]

A recent post from the good folks over at Empress Pedals ( https://empresseffects.com/blogs/empress-blog/low-noise-pedals ) concerns how they achieve lowest possible noise in their pedals.  They note the use of triple-parallel VCAs as one of their strategies in the design of their Reverb pedal, using 2164 chips.  I have no idea how this is implemented.  I suppose for details, you'd need to speak to Steve or Jay there.

[ElectricDruid]

Thanks for the explanation! It is making much more sense to me now I think.  I'm going to try to get it going with a microcontroller next.  The plan is to use a smooth pwm output.  With a dual rail supply I'm hoping I will be fine since the max output I get from my microcontroller is 3.3V.  I am also wondering how to go about using a single 9V supply.  Would there be a way to shift the pwm signal to range from 4.5V to 7.8V?  Or is there a different way to reference the pwm signal to the 4.5V virtual ground?

One way might be to run the microcontroller on the "upper half" of the supply, so from 4.5V to 9V. Then the output is already in the right range. Otherwise, it's not too difficult to provide an offset voltage to shift it to the right range using an op-amp stage. It's a mixer application, essentially; you want to mix the PWM signal (either before or after you filter it) with a DC offset voltage.

[Gobotak]

I'm trying to use a summing amplifier to get my voltage in the correct range.   I currently have it set up like this:

There is a mistake in the schematic.  I'm trying to get 4.5-7.8V for my output but the image has it 3.3-7.8V.  On the breadboard my output is going from 2-3.5V.  I'm using a TL072 so my thought was it might not be getting the full swing but would that mean my range also gets decreased?  I would have thought if that is the case my lower voltage would still be around 4.5V.