Mid Side M-S Matrix Uses No Precision Resistors

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
Post Reply
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by mediatechnology »

Here's a gain trim idea I'm kicking around which only requires a single-pole rotary switch. Something Paul said earlier made me think of this.

Image
https://www.ka-electronics.com/Images/j ... n_Trim.jpg

There may be a more efficient way to gain-structure this after I think about it some more but here we go. It has a couple of attributes. The first is that 0 dB is set by the gain of the wire with no resistor error for precision. Secondly, it only requires a single pole switch. A third lesser advantage is that the +3 dB gain stage is also 0.05 dB precise using a THAT1243 in a non-inverting +3 dB gain configuration. Thus the +3 dB setting is exactly 3 dB as a bonus.

The gain/attenuation settings on the front panel read: +3, +2, +1, 0, -1, -2 and -3 dB. The actual attenuation settings are 0, -1, -2, (-3 skipped), -4 -5 and - 6 dB. The step sizes and number of steps can be anything you want. For 6 dB gain a 1246 could be used.

The attenuator has to be buffered with ultra-precise 0 dB gain to prevent error in the middle, "straight wire" stop. This almost dictates the use of a voltage-follower. An inverting amp would have resistor error and double inversion, to maintain polarity, would be required. We need to chose the op amp carefully to avoid common-mode issues.

(To reduce common-mode error there is one more trick we could pull out of our sleeves: Use another THAT1240 which would attenuate the non-inverting input by 6 dB and have 6 dB of gain in the inverting feedback loop. The combined gain is an accurate 0 dB (due to on-chip resistor matching) but the CM range is about 1/2 that of a follower. The one disadvantage to this is that it loads the attenuator with an 18K +/-20% resistor at the +2, +1, -1, -2 and -3 dB steps.)
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by mediatechnology »

RF;

I think a 797 would be fine but we ought to try one for follower unity gain stability issues.

I think it will need a 100R Rfb and an optional 100R series resistor with the In+ input pin 3.

See http://www.analog.com/static/imported-f ... df#page=13

Figures 36 and 37. I used the R2 "balancing resistor" with the AD823/OPA2604 and it made some THD differences. Figure 36 just uses resistors; figure 37 adds compensation capacitors. I think I would use 37 but we really ought to build one.
Gold
Posts: 677
Joined: Sat Dec 01, 2007 5:20 pm

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by Gold »

mediatechnology wrote: Paul/Ruairi - If the +/- 3dB trim were on a pot would it be precise enough from a repeatability perspective? If not, one could build it as a stepped switch. It sounds like to me that you are leaning towards switches. Maybe we should make the board accept both types of controls.
I think +/- 3dB on a pot would be precise enough for me. I suspect other ME's would disagree. I suggest leaving provision for attenuator resistors or a relay attenuator off board. It would keep size/cost down and it's about the same amount of work to put the resistors on a switch as it is to put them on a board. This would leave the choice between using the toggle switch for var in/out or using a 2 pole rotary switch for the 0v connection. Relay attenuators sounds like a different project to me.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by mediatechnology »

Thanks Paul and Roger. Here's the variation of the circuit using a pot:

Image

https://www.ka-electronics.com/Images/j ... im_Pot.jpg

Not shown of course are the detailed 1243 or follower connections...

The layout looks great BTW. I see more meat around pin 3.
Gold
Posts: 677
Joined: Sat Dec 01, 2007 5:20 pm

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by Gold »

Looking at Wayne's gain trim circuit above the only downside I see is that the mono sum will be between 3dB and 6dB up already. Having 3dB of gain at the input might have headroom issues. It seems that to "fix" this you would give up the precise level matching by having makeup gain set by external resistors around an opamp. I think using a 1246 would be asking for trouble but a 1243 would probably be okay.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by mediatechnology »

Looking at Wayne's gain trim circuit above the only downside I see is that the mono sum will be between 3dB and 6dB up already. Having 3dB of gain at the input might have headroom issues. It seems that to "fix" this you would give up the precise level matching by having makeup gain set by external resistors around an opamp.
Paul - Your absolutely correct about that - the matrix M channel has, for fully-correlated inputs, +6 dB internal "gain" already due to M-S math. Adding another 3 dB in the M channel reduces headroom even more.

A fix for it as you say is to buffer the attenuator with a + 3dB (or +6) non-inverting stage after the wiper or switch stator. The top of the pot (I'll draw this) would just connect to the proceeding line receiver's output eliminating a stage. The +3 dB gain stage which followed the pot could be an op amp with only the bias resistor as attenuator load or it could be a 1243/1246 with an 18K ohm (+/-20%) load to eliminate external components.

The op amp, if it were an AD797, would require about 5 external components and the bias resistor would have to be almost as low, within the same decade, as the 1243/1246 load would be to prevent switching clicks. The 797 is a better overall performer though and it's bias resistor could be 1% and factored into the attenuation calculation.

In any case, there would still be that SPDT switch on the end to bypass the variable section.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by mediatechnology »

Paul G wrote:
What about having the decode line receiver 1240 feed a switch that is an attenuator with a maximum output of -6dB and have 6dB of makeup gain from an opamp afterward? The output of the opamp would have a low enough impedance to feed the decode circuit.
I think it's almost exactly what you mentioned about 5-6 pages ago Paul. Sorry I didn't get this sooner. Good eye BTW on spotting the headroom reduction issue. OK, here's that drawing.

Image

https://www.ka-electronics.com/Images/j ... _Rev_A.jpg

Roger - I just looked and I have two AD797s. I could proto this...
Mastertone
Posts: 11
Joined: Sat Jan 05, 2008 6:18 pm

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by Mastertone »

Isnt eaiser to have a switch to select between Cut/boost ? So when the pot is fccw there is 0 gain eihter way, in cut mode fcw this would mean 3dbcut and so on. I might have missed if you discussed this already.
Jonas Ekstrom
Mastertone
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by mediatechnology »

Isnt eaiser to have a switch to select between Cut/boost ? So when the pot is fccw there is 0 gain eihter way, in cut mode fcw this would mean 3dbcut and so on. I might have missed if you discussed this already.
That could be done too but it doesn't eliminate the need for a "fixed" gain switch (if that's what you meant) for precision decoding applications. As a precision decoder, the gain at 0 dB has to be within <<0.05 dB accurate.
Mastertone
Posts: 11
Joined: Sat Jan 05, 2008 6:18 pm

Re: Mid Side M-S Matrix Uses No Precision Resistors

Post by Mastertone »

No i understand the fixed gain switch. It would be 3 or 4 switches, fixed gain could have only 1 for both M/S, and 2 Cut / Boost switchs.

I just thought that it could be a little tweeky to have a non center detented pot with 0 gain at 50%.
Jonas Ekstrom
Mastertone
Post Reply