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.
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 »

I did a 6 dB gain trim here. It's the second stage. In this thingy it could be a NE5534, AD797 or whatever rather than the 823 Bifet. I had to use the 823 due to the first stage which we don't need. This one had ~ +/- 6dB adjustment.

Image

It inverts. This is not a problem however as the 1246's inputs can be redefined (flipped) causing a second effective polarity inversion. We just call the plus input "minus" on the return input's 1246.
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 »

I'm thinking at F, though C would allow you to hear noise floor/errors/etc of processing with no source signal. Paul may have a strong argument for C.
I'm thinkin' F. If you mute at C to check noise floor then you can't isolate in L or R output which external processor channel, M or S, is producing the noise. Do I have that right? The options may weigh evenly.
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: The options may weigh evenly.
Seems that way to me. I have no preference.
emrr
Posts: 570
Joined: Sun Mar 02, 2008 10:27 pm
Location: NC, USA
Contact:

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

Post by emrr »

mediatechnology wrote:If you mute at C to check noise floor then you can't isolate in L or R output which external processor channel, M or S, is producing the noise. Do I have that right?

Correct; would really require mutes at both places to check noise floor I was imagining. Seems overkill.
Best,

Doug Williams
Electromagnetic Radiation Recorders
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 »

I think putting it at point F then reduces ambiguity. I'll try and draw up a schematic (or parts of one) for Roger to work from. I think RF is right though: We probably should use relays not only for the soft bypass but also for the mutes at F. That makes thre DPDT relays.
Mastertone
Posts: 11
Joined: Sat Jan 05, 2008 6:18 pm

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

Post by Mastertone »

woow, you guys have been busy here.

Dont know if its worth anything but i totally agree with Pauls idea about GainTrims and the possibility to mute mid/side.

For me, one major thing would be to try to keep the Pcbs down to a minimum size.
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 »

Thanks for your input Mastertone.

I think we will be building in the gain trims with the bypass and mutes right after them.

When bypassed, gain trim will not be in effect so maximum separation is maintained. I think it would be useful to have the mutes active regardless of the bypass so one can hear the source (mid or side) in isolation. If I'm wrong about that now would be a good time to let me know.

We have all the various blocks documented, balanced in, balanced out and the matrix - the only thing I want to prototype is the +/- 6dB gain stage using a bipolar op amp. The one I showed earlier is DC-coupled and with the increased bias current of a 5534-class op amp it may need a coupling cap to keep rotational noise down.
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 »

There may be a problem having a potentiometer-based variable gain stage in the insert return that did not occur to me until I floated this idea to someone off-forum.

A solution using a pot would require a very accurate center tap to maintain precise 0 dB attenuation. Otherwise, the decoder could not be used frestanding unless the pot were bypassed to a fixed gain stage or a stepped switch is used. Separate gain-accurate inputs could also be provided. The accuracy of the matrix would be blown for pure decoding use if this isn't done.

A pot-based gain stage increases noise, and bypassing it to maintain precise 0 dB gain also requires maintaining polarity and additional switches.

A stepped switch would be repeatable and have precise 0 dB gain. Buffering is required using a switch though if it feeds the decoder matrix because the source impedance has to be zero, or nearly so. This also adds complexity and noise.

Another option, also adding a layer, would be separate gain-accurate decode inputs. (Two more 1240s and jacks.) This would have the gain trim circuitry plus more stuff.

The trim might also be needed in the L and R domain to restablish balance rather than the M-S domain controlling ratios.

It seems an outboard Gold Point attenuator (buffered?) would be perfect for someone wanting gain trim. With it patchable to various points in the chain it could be at the M-S insert send, M-S insert return or L and R outputs.

In the interest of simplicity, we should do a fixed gain version and figure out what places level trim would be needed. This keeps the matrix simple and returns us to K-I-S-S. Here's the block diagram I propose:

Image

https://www.ka-electronics.com/Images/j ... gram_1.jpg

The above has three relays: Bypass, Mute Mid, Mute Side. The mute connections to the relays would normally be grounded but I suppose that it wouldn't hurt to have a link and a tiepoint there so they could be used for something else. There are also tiepoints (extra pads) at the SE outputs of the encoder and the output of the bypass relay. A break in the trace with link and added tie points between the insert return's 1246 outputs before feeding the relay (break not shown) might also be useful.

Any ideas?
emrr
Posts: 570
Joined: Sun Mar 02, 2008 10:27 pm
Location: NC, USA
Contact:

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

Post by emrr »

I'm happy leaving it out in the interest of honest stand-alone decoding. Honestly, most of the insert processing equipment I would use has variable gain already, or it's easy enough to patch something else in when needed. Hmm, side chain processing with gain means an alignment procedure will have to be followed when attempting to know what unity really sounds like.

If attempting to include gain variance, separate gain-accurate decode inputs sounds like an acceptable addition to me. Then we have purist path and a variable path when needed. Why not switch the same jacks to different 1240's? Constraints of layout recommendations, or additional relays?

I am already struggling with the reality of having a workstation option for this already. There are times when a hardware solution would be most handy; Paul clearly needs hardware. I guess simplicity and economy are what I'm thinking about. At some point the finished price comes into competition with passive transformer solutions, and becomes less appealing to some of the potential market.
Best,

Doug Williams
Electromagnetic Radiation Recorders
Gold
Posts: 677
Joined: Sat Dec 01, 2007 5:20 pm

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

Post by Gold »

If it becomes too much of a problem to have variable gain so be it.

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.
Post Reply