Simple MIDI Line mixer with THAT VCAs.

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
Post Reply
SUBass
Posts: 10
Joined: Fri Mar 04, 2011 9:56 am

Simple MIDI Line mixer with THAT VCAs.

Post by SUBass »

THAT's the project I'm working at least. Since this place looks to be pretty THAT-centric I figured it's probably the best place to ask THAT questions (Besides THAT Corp of course)

So first up...I'm way out of my league based on the discussions I've read here so far so please bear with me if I say something crazy. Anyway...Here goes...

In building my bass rig I've come to the point where I need to clean up some things and make it all usable for live performance. To do so I need to make a magic whiz-bang box with a few pieces in it. Namely, a splitter and a mixer.

My signal first needs to split. One side will feed my amp, the other side will feed a tube preamp. The amp (Gallien Krueger 1001rb) has a series loop just before the power amp and here is where the mix happens. The loop send hits channel 1 of the mixer, the tube preamp output hits channel 2. These sum together and return to the GK loop return. The desire for MIDI comes in with being able to implement a volume pedal, tuner mute, channel selection etc.

I got a chance to play around with the 2180 last night. Essentially I built up the schematic on page 8 (figure 14) of the 2180 datasheet and gave it a whirl. Knowing that I've got unbalanced IN's and OUT's I didn't think much of line receiver/buffers and just tacked a 1/4" jack on the front of the 2180 with the C and R and let it rip. The result? Massive distortion. Tinkering around I found that the VCA was overloading. In hindsight...I grounded the negative Ec port but I didn't have anything connected to Ec+. Perhaps that's why I had an issue but it was an unexpected outcome.

Obviously I need to get a little more scientific and measure the levels that are floating around to be 100% sure but I can't imagine that the gain structure on the amp is really that goofy. Besides, inserting EQ's or compressors works just fine. I suspect that I just missed something with the Ec ports.

Shifting gears and looking at the overall signal for a moment....The input of the VCA appears to inverting which works out great using an inverting I-V opamp on the output. Everything flips back to normal and life is good. Is summing a couple of channels of these VCAs really as simple as just connecting their outputs together and hitting an inverting I-V opamp stage? If so...That's just incredibly easy. All you need to do is pick an appropriate opamp and you're done.

I noticed that in the OTB VCA mix bus thread there was a VCA stuck into the feedback loop of the output opamp. What was the the purpose of that? From what I got in the thread it was there to reduce distortion...But I guess I just failed to see how it worked.

Also...Any thoughts on the 2162 versus the 2180?

Thanks!

James
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Simple MIDI Line mixer with THAT VCAs.

Post by JR. »

SUBass wrote:THAT's the project I'm working at least. Since this place looks to be pretty THAT-centric I figured it's probably the best place to ask THAT questions (Besides THAT Corp of course)

So first up...I'm way out of my league based on the discussions I've read here so far so please bear with me if I say something crazy. Anyway...Here goes...
My first question is how do you plan to interface with MIDI? Program changes for preset scenes, midi controllers for sliders?
In building my bass rig I've come to the point where I need to clean up some things and make it all usable for live performance. To do so I need to make a magic whiz-bang box with a few pieces in it. Namely, a splitter and a mixer.

My signal first needs to split. One side will feed my amp, the other side will feed a tube preamp. The amp (Gallien Krueger 1001rb) has a series loop just before the power amp and here is where the mix happens. The loop send hits channel 1 of the mixer, the tube preamp output hits channel 2. These sum together and return to the GK loop return. The desire for MIDI comes in with being able to implement a volume pedal, tuner mute, channel selection etc.
I remain a little confused by the utility of MIDI for live performance... I have seen MIDI more in connection with recording (triggered by SMPTE time code or something similar.)

If you are manually controlling the MIDI, why not manually control your sub mixer?
I got a chance to play around with the 2180 last night. Essentially I built up the schematic on page 8 (figure 14) of the 2180 datasheet and gave it a whirl. Knowing that I've got unbalanced IN's and OUT's I didn't think much of line receiver/buffers and just tacked a 1/4" jack on the front of the 2180 with the C and R and let it rip. The result? Massive distortion. Tinkering around I found that the VCA was overloading. In hindsight...I grounded the negative Ec port but I didn't have anything connected to Ec+. Perhaps that's why I had an issue but it was an unexpected outcome.
?? Both control ports need to be close to 0V, and input current leeds to be within the dynamic range of the VCA.
Obviously I need to get a little more scientific and measure the levels that are floating around to be 100% sure but I can't imagine that the gain structure on the amp is really that goofy. Besides, inserting EQ's or compressors works just fine. I suspect that I just missed something with the Ec ports.

Shifting gears and looking at the overall signal for a moment....The input of the VCA appears to inverting which works out great using an inverting I-V opamp on the output. Everything flips back to normal and life is good. Is summing a couple of channels of these VCAs really as simple as just connecting their outputs together and hitting an inverting I-V opamp stage? If so...That's just incredibly easy. All you need to do is pick an appropriate opamp and you're done.
Yes I have hypothesized just combining those VCA (current) outputs for summing without the build up in noise gain. For a modest number of VCAs this should work without too much concern fro output capacitance while the following opamp needs some feedback capacitance to swamp out the VCA capacitance.
I noticed that in the OTB VCA mix bus thread there was a VCA stuck into the feedback loop of the output opamp. What was the the purpose of that? From what I got in the thread it was there to reduce distortion...But I guess I just failed to see how it worked.

Also...Any thoughts on the 2162 versus the 2180?

Thanks!

James
Putting a VCA in the feedback path was something I have actually done in a console master section to globally reduce the master summer gain, but this only works with the old style VCA. The modern inverting VCA will not just drop in a feedback path and work.

JR
Cancel the "cancel culture", do not support mob hatred.
SUBass
Posts: 10
Joined: Fri Mar 04, 2011 9:56 am

Re: Simple MIDI Line mixer with THAT VCAs.

Post by SUBass »

Hey JR,

Thanks for replying. This is my live bass guitar rig and most important to me is presets and mute. Sometimes I'll want just the tube preamp...Sometimes I'll want just the SS preamp. Sometimes I'll want to blend the two and in between each preamp is an analog MIDI graphic EQ for tailoring different sounds for the preamps. For example...To change from a vintage-y tube sound to a funky slap sound would be mute the tube pre, unmute the SS pre, select the "slap" EQ curve on the SS EQ. Three control changes that with MIDI can be one button push on a foot controller. (preset recall on the graphic EQ is multi-button pushing affair)

There's some other stuff in the scope of the project...
A splitter so I can send the signals everywhere.
A couple of MIDI relays to select the channel of the tube pre.
And this piece...The VCA mixer to sum and switch the pres, and provide a muting facility for silent tuning.

There are products that do some of this stuff already. Splitters are readily available, MIDI controlled switchers for amps and FX loops are readily available. The simple 2-4 channel to 1 mixer....Yeah...Those are available....But no MIDI control.

There's some economics to it as well. I just have a hard time spending $400 for a Voodoo Labs GCX to work as a channel switcher for the tube pre. Especially considering I already need to develop some control for the VCA mixer. How hard could it be to add in some programmable relays?

Anyway....That's what I'm looking at.

If I'm just setting up the VCA to be unity gain right now to pass signal I should ground both EC inputs so that they're at 0V?

Thanks!
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Simple MIDI Line mixer with THAT VCAs.

Post by JR. »

I am still a little unclear about what you mean by MIDI..?

Are you using it to encompass all remote controlled functionality, or do you have some preference for using MIDI live?

Presets for a rig do not have to be done using MIDI control interfaces.

JR

PS: FWIW decades ago when I was designing stuff for bedroom recording I even designed a MIDI controlled insert-patch bay (MAP 8x4). This 1 rack space product had 8 inserts and 4 effects loops that could be electronically switched around and effects could be inserted in any combination, as programmed by MIDI program changes. That product is long obsolete but it was along the lines of moving a limited number of effects around, and if those effects had midi control, you could change them and move them around at the same time. With smpte to MIDI control you could perform some aspects of poor man's mixdown automation, making one reverb cover several different tasks within a single song. Of course this product made more sense back a few decades ago when a good reverb was $1000 not $100. Times change.
Cancel the "cancel culture", do not support mob hatred.
SUBass
Posts: 10
Joined: Fri Mar 04, 2011 9:56 am

Re: Simple MIDI Line mixer with THAT VCAs.

Post by SUBass »

I'm using MIDI for a couple of reasons.

I already have to use MIDI to change the presets on the EQ's. If I don't use MIDI for the mixer...I'll have to come up with some way to take whatever preset I select for the mixer and have it generate a MIDI message to the EQ's. Either way...I'm working with MIDI.
I already have a fully programmable MIDI foot controller that sends bank/patch change and continuous control data.

Right now I'm not too concerned about the control stuff as I need to get the analog electronics working first. If I were to make the VCA just be unity for now, I'd be happy.

Your MAP8x4 is actually still a relevant product. Check out the Voodoo Labs Ground Control/GCX combination or the RJM switching stuff. There's also a few smaller companies internationally that are making MIDI controlled FX loop/amp switchers. Even the most basic units are about $300. The GCX is a great piece at $400 street.

Still...Your MAP8x4, if it had the ability to shift loop order (like say insert 1 to loop 2 then loop 1 on one patch and then insert 1 to loop 1 then 4) that adds something I haven't seen in the regular products.

You might want to dust off that "obsolete" design and bring it back into service. Especially if you can hit a lower price-point. Your R&D is already there. I would imagine back then you were probably using either a Z80 or 8031/51 MCU to run it. There are current MCU's that understand the instruction sets of those older chips...But re-writing the code for a modern PIC or AtMega should be pretty straight-forward.

For what it's worth...I was planning on using an Arduino based MCU for my project. I plan to use a I2C connected 8 bit DAC and have an I2C connected LCD panel. I'll obviously need to scale the 7 bits of MIDI CC data into 8 bits for the DAC which could be as simple as a x2 function. The low values should be enough down in the "dirt" that the "jump" from "2" to "0" wouldn't be audible. As pointed out in the THAT design note, using a RC network on the DAC output will create a smooth voltage output for the VCA.

Now...Out of MIDI and control issues....Some thoughts on the VCA bit.

One question on the THAT DN-02. THAT is using the Ec- port, they've got an inverting opamp with negative gain that scales the voltage down. Then to set the gain range, they tie a resistor from V- to the inverting input of the opamp. That leads me to a couple of thoughts. Using an adjustable regulator for +/- V and then tying to that to Ec ports is probably a bad idea. Also...Not that it's a big deal...But any noise from the PSU regulators is going to be right there on the Ec port. Granted...It will be at a significantly lower lever because of the resistor, but there's also the resistor noise. It just seems like a non-ideal way to set the offset voltage. Perhaps a precision voltage reference like this one http://www.national.com/pf/LM/LM4140.html#Overview would be better direction? You can get that national chip in a variety of voltages, and with the lower voltage you can use a smaller resistor into the opamp. Which leads into which opamp to use? Looks like I should be using a really low noise bi-polar input opamp. I say a bipolar opamp because I was thinking the input impedance would be lower which might be better for the current flows from the offset voltage and through the feedback resistor of the opamp.

I could be really over-thinking this noise on the EC port. I realize that my intended application is inherently noisy...But I have a desire for the mixer to preserve as much of the signal integrity as it can without introducing any artifacts of its own. Really though...The difference could be negligible.

James
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Simple MIDI Line mixer with THAT VCAs.

Post by JR. »

SUBass wrote:I'm using MIDI for a couple of reasons.

I already have to use MIDI to change the presets on the EQ's. If I don't use MIDI for the mixer...I'll have to come up with some way to take whatever preset I select for the mixer and have it generate a MIDI message to the EQ's. Either way...I'm working with MIDI.
I already have a fully programmable MIDI foot controller that sends bank/patch change and continuous control data.
sounds good...
Right now I'm not too concerned about the control stuff as I need to get the analog electronics working first. If I were to make the VCA just be unity for now, I'd be happy.
there should be plenty of cook book circuits from THAT app notes to copy,
Your MAP8x4 is actually still a relevant product. Check out the Voodoo Labs Ground Control/GCX combination or the RJM switching stuff. There's also a few smaller companies internationally that are making MIDI controlled FX loop/amp switchers. Even the most basic units are about $300. The GCX is a great piece at $400 street.

Still...Your MAP8x4, if it had the ability to shift loop order (like say insert 1 to loop 2 then loop 1 on one patch and then insert 1 to loop 1 then 4) that adds something I haven't seen in the regular products.
Yes it could do that, and I even added some clever tricks so you could switch on the fly without clicks...

But it was obsoleted long ago.. IIRC the processor board was hand taped (that tells you how long ago this was), and the digital group would have to re-engineer the processor board to make more... so we just obsoleted the whole product. Peavey had a number of MIDI products that were ahead of the market and did not enjoy much commercial success. A little too cool for typical Peavey dealers, not widely advertised or embraced by studio types who were too cool to use Peavey.
You might want to dust off that "obsolete" design and bring it back into service. Especially if you can hit a lower price-point. Your R&D is already there. I would imagine back then you were probably using either a Z80 or 8031/51 MCU to run it. There are current MCU's that understand the instruction sets of those older chips...But re-writing the code for a modern PIC or AtMega should be pretty straight-forward.
I was the analog guy on the project , but IIRC it may have been a 8038, I think that was a popular early 8 bitter.

I was involved with another digital project that used a 6511 and that twas supposed to be hot stuff way back then...
For what it's worth...I was planning on using an Arduino based MCU for my project. I plan to use a I2C connected 8 bit DAC and have an I2C connected LCD panel. I'll obviously need to scale the 7 bits of MIDI CC data into 8 bits for the DAC which could be as simple as a x2 function. The low values should be enough down in the "dirt" that the "jump" from "2" to "0" wouldn't be audible. As pointed out in the THAT design note, using a RC network on the DAC output will create a smooth voltage output for the VCA.
I've never looked very closely at Arduno,, I have done several designs recently using PICs and it is pretty remarkable how much you get for a few $.

The key thing about interfacing to VCA is that the control voltages are in X mV=YdB so you output your DAC say 0-5V or 0-3.3V, and divide it down and offset it for +.-whatever dB you want. Putting the final resistor divider right at the VCA will help attenuate digital noise.

One tradeoff is that VCAs want to see low impedance drive impedance at the control voltage ports (for low distortion) so you may want to buffer the control voltage with opamp.
Now...Out of MIDI and control issues....Some thoughts on the VCA bit.

One question on the THAT DN-02. THAT is using the Ec- port, they've got an inverting opamp with negative gain that scales the voltage down. Then to set the gain range, they tie a resistor from V- to the inverting input of the opamp. That leads me to a couple of thoughts. Using an adjustable regulator for +/- V and then tying to that to Ec ports is probably a bad idea. Also...Not that it's a big deal...But any noise from the PSU regulators is going to be right there on the Ec port. Granted...It will be at a significantly lower lever because of the resistor, but there's also the resistor noise. It just seems like a non-ideal way to set the offset voltage. Perhaps a precision voltage reference like this one http://www.national.com/pf/LM/LM4140.html#Overview would be better direction? You can get that national chip in a variety of voltages, and with the lower voltage you can use a smaller resistor into the opamp. Which leads into which opamp to use? Looks like I should be using a really low noise bi-polar input opamp. I say a bipolar opamp because I was thinking the input impedance would be lower which might be better for the current flows from the offset voltage and through the feedback resistor of the opamp.

I could be really over-thinking this noise on the EC port. I realize that my intended application is inherently noisy...But I have a desire for the mixer to preserve as much of the signal integrity as it can without introducing any artifacts of its own. Really though...The difference could be negligible.

James
Noise on the EC port is important... you don't need precision voltage references because the VCA = 0db at 0v so generally any error because of supply voltage will be at dB full scale end of range, not at unity gain.

re: the THAt design note.. they generally have their act together. You will observe that the resistor from -V is 178k feeding a 2k feedback R so power supply noise will be attenuated buy ratio 2/178. But if concerned you can split it into two 90k resistors with a cap in the middle to ground (like the filter on the DAC output).

it seems there are other threads around here with some discussion about interfacing micros with VCA control, but perhaps not as rigorous or vetted like the THAT app note.

JR
Cancel the "cancel culture", do not support mob hatred.
SUBass
Posts: 10
Joined: Fri Mar 04, 2011 9:56 am

Re: Simple MIDI Line mixer with THAT VCAs.

Post by SUBass »

Thanks for that John.

I got my Arduino and LCD kits in yesterday and put those together. I had it pushing messages to the I2C LCD and turning relays on and off last night. I still need to work out some sort of crude user interface then I can start looking at the MIDI I/O. A rotary encoder and a couple of buttons would probably work.

Tonight I'll get back to the VCA and see if I can make that fly. I think the problem I had with distortion before was leaving the Ec pin unterminated. If I ground both Ec pins that should set the VCA in "unity" and signal should pass through without issue (since both pins will be at "zero" and equal from one another.)

I'm waiting on a MSOP to DIP adapter for the DAC I'm using to be able to make the control voltage circuitry.

Good times!
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Simple MIDI Line mixer with THAT VCAs.

Post by JR. »

Yes, if one EC pin is left floating, it will surely misbehave.

Both EC pins need to be near 0V and within mV of each other. For very large dB commands (like mute) you may want to drive both control ports in opposite directions (gain is the voltage difference between them), so neither one it too far from 0V, but most applications only drive one port or the other.

JR

PS: it's been a while since I messed with LCD.. the old ones I used took crude 4 bit parallel com (IIRC). it seems logical to go serial there too. Does the Arduino have firmware for reading rotary encoders easily?
Cancel the "cancel culture", do not support mob hatred.
juniorhifikit
Posts: 139
Joined: Sun Oct 04, 2009 8:37 am
Location: Oakland/Paris

Re: Simple MIDI Line mixer with THAT VCAs.

Post by juniorhifikit »

SUBass wrote: I could be really over-thinking this noise on the EC port. I realize that my intended application is inherently noisy...But I have a desire for the mixer to preserve as much of the signal integrity as it can without introducing any artifacts of its own. Really though...The difference could be negligible.

James
I've been working on a similar project, and from my experience I can say that you're not over thinking the noise on the EC port issue. Everything in the CV shows up in the audio to some extent. I found that a precision regulator was overkill, but a well filtered bias supply is indeed important.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Simple MIDI Line mixer with THAT VCAs.

Post by JR. »

Yup, driving the EC pins is a mix of needing to be both low noise and low impedance. High noise will cause a gain modulation by that noise, high impedance a distortion mechanism.

JR
Cancel the "cancel culture", do not support mob hatred.
Post Reply