Mid Side M-S Matrix Construction Information

Construction information for DIY projects, including the MS Mid Side Matrix, Elliptic Equalizer, Mastering Console, Phono Transfer System, Insert Switcher and the Dual Class-A Amplifier. You can post your baby pictures here.
Post Reply
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Construction Information

Post by mediatechnology »

sr1200 wrote:Im also still not decided on if i should integrate this with a analog sum box im designing. Im leaning toward keeping this a separate unit all together.
Well, with all the insert points it has you'll probably run out of back panel before you have room to put anything else in the box. :D

On those LEDs for the bargraphs you might consider the series stacked constant-current approach. But having said that, 1 mA/LED with parallel drive can be pretty bright these days.
sr1200
Posts: 6
Joined: Tue Feb 15, 2011 8:27 am

Re: Mid Side M-S Matrix Construction Information

Post by sr1200 »

I have an abundance of KA2284's was looking to use a few of.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Construction Information

Post by mediatechnology »

Are those still available?

With 20 modern LEDs at 1 mA/each you should be able to light the room at FS. I think on one of the LM3914 GR-10 builds we had a series resistor in series with the entire parallel LED array. The problem was dropping the brightness beyond the LM3914's minimum current so Roger decided to current limit the whole array as more lit up. Although it made the brightness of each LED drop due to current starvation as more LEDs lit, it worked to keep the overall brightness down and more nearly constant. The folks at Pr*digy didn't see Roger's brilliance and made fun of it. But it makes a lot of sense particularly if the engineer doesn't want to get blinded by LEDs.
sr1200
Posts: 6
Joined: Tue Feb 15, 2011 8:27 am

Re: Mid Side M-S Matrix Construction Information

Post by sr1200 »

Image
This is the schematic for the most part for the VU circuit.
EDIT: Disregard the bottom half circuit design its ALL sorts of wrong

What sucks is i just realized that the KA chip only handles up to 16V... doh. I guess i could put a voltage resistor in there for each of em. But I guess this is just another sign that i should make the 2 units separate. I think my entire summing box was 16V. (talk about light show... theres 104 LEDs on the front panel and 8 LED pushbutton switches... sunglasses required!)

And thats kinda a cool idea with the brightness issue. Why not make it a variable resistor (pot) so you could adjust the brightness like you would with an LCD? I should see about doing something similar in the sum box.... hmmmm......

Im not experienced enough to make fun of anyone for any idea... every idea has some kind of reason behind it and if it seems logical for the application... why not!?
sr1200
Posts: 6
Joined: Tue Feb 15, 2011 8:27 am

Re: Mid Side M-S Matrix Construction Information

Post by sr1200 »

I just about finished the build... gotta tie up some loose ends.
I tested it out last night after getting the jacks wired up. Is there any kind of dummy test I can do to make sure everything is working properly. I think i may have a phase issue somewhere (possibly a reversed set of wires). I took a drum kit bus sent it to the input. Sent the M send right back into the M return (unprocessed). Hooked the S send into a comp (a little light compression on there) and back into the return then sent the outs back to the console. The mid solo'd sounds about right (i think). But when the S gets added to it, im getting some weird drop outs on the right side of the stereo spread. If i phase reverse the S same dropouts just on the left. The S by itself also sounds about right. Any suggestions on a "test" of sorts?
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Construction Information

Post by mediatechnology »

sr1200: Thanks for the post.

First let's look at the physical polarity reversal on the encoder input and decoder output connectors:

Image

You'll notice on the encoder L and R input side "-" is on the top of the Phoenix connector and on the decoder L and R output "+" is on top. Check that first, because it could cause an overall polarity inversion in one or more of the paths.

Note also the physical polarity inversion between the "M Out" and "M In" connectors. Dittos also for the S Out and S In. For test, just use a twisted pair and let them naturally cross-over.

Are you using the 5532 (or whatever OPA) -6dB stage? It inverts. (Which is why the PCB physical orientation is different on the L and R inputs/outputs.) If there's no 5532 in the path, and you're jumping around it, that could be another source of inversion. (The Mid and Side I/O has a physical inversion because it simplified layout.)

Check also the 1246 and 1240 stuffing. From left to right it's: 1246, 1240, 1646, 1246, 1240, 5532, 1646. If the 1246 and 1240's are swapped, things are going to be very weird sounding indeed. :?

EDIT. Been awhile since I had to think about this. :oops: After visiting the "thinking chair" it came back to me:

A quick check of the encoder can be done by feeding L, R or both.

1) Feeding both L and R produces 2X Mid.
2) Feeding L only produces Mid and Side.
3) Feeding R only produces Mid and -Side.

Once the encoder is verified, loop M Out to M In, and S Out to S In.

Then, L In routes to L Out and R In routes to R Out with overall unity gain.

Look also at the jumpers for points B, C, F and G:

Image

External Polarity Inversion

Is a device connected into an insert (Mid or Side) that inverts polarity? Worth asking 'cause you never know...

The Mid Side Encoder and Decoder are the Same Thing. They are interchangeable in their roles.

One more thing I'll remind anyone using these is that the encoder and decoder matrices are the same thing. The only difference is the -6dB stage in the decoder.

Some people fret about the naturally-occurring +6 dB buildup in the encoder. Using the decoder as an encoder and the decoder as an encoder allows the Mid build-up to be reduced by the OPA stage.

We put the -6dB stage in the decoder for a couple of reasons:

(1) The gain error that occurs from the 5532's external 1% resistors only affect L/R channel balance. If the decoder is used as an encoder, the tolerance of the discrete resistors reduce crosstalk performance.
(2) I'd rather use +/- 18V rails to pickup some headroom rather than reduce the natural Mid build-up or alter the ratio of Mid to Side. I suppose it's a matter of philosophy.

But you can have it both ways.

Let us know what you found.
sr1200
Posts: 6
Joined: Tue Feb 15, 2011 8:27 am

Re: Mid Side M-S Matrix Construction Information

Post by sr1200 »

Sure enough. The right input wasnt making full contact. Reseated the cable and its crystal clear! :? Great job on this btw!
Is there any benefit to rigging the jumpers to switches? From what i understand this would bypass the various stages (or would it just mute them?)

Thanks much for the reply. This thing is brilliant!

Edit: checking the polarity checking the l and r input individually that led me to it.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Construction Information

Post by mediatechnology »

Cool! Glad you got it working and thanks for the kind words.
sr1200 wrote: Is there any benefit to rigging the jumpers to switches? From what i understand this would bypass the various stages (or would it just mute them?)
Good question.

I'd put a bypass relay or switch between points C and F for the "What does it sound like without processing?" question. That's the most gain-accurate point for bypass because it eliminates the 1646, it's 25R output impedance, and the loading on the 25Rs caused by whatever is in the insert. It doesn't take much gain error at all in the M/S domain to blow x-talk. Going around the 1646 from a level and x-talk perspective is as close to "A" in an "A/B" comparison as you'll get.

Our friend Bob Katz has found a use for the board as a simple quad line driver/receiver using the various points:

viewtopic.php?f=6&t=112&start=193

You can also use the inserts for mutes. A number of people have done this.
isophase
Posts: 4
Joined: Mon Jul 18, 2011 10:06 am

Re: Mid Side M-S Matrix Construction Information

Post by isophase »

Hi everyone, this is my first post on the forum, i just started stuffing the board and was wondering if anyone experienced the same trouble as me: the tin is not "sticking" properly on all the ground pads? my soldering iron is set to 300degrees Celsius temperature, all the other pads i can solder no problem just the ground pads, i can clearly see from looking at the solder side of the PCB that the ground pads are not "isolated" like the other ones. i sort of have the impression that this has to do with the protection varnish messing with the flux and so the tin is not sticking properly. does anyone had the same problem? am i doing something wrong? i tried lowering the temp of my soldering iron but it is the same. i tested continuity between some ground pads i soldered in and i do get a contact there but the soldering looks really poor.
any hint greatly apreciated
and thanx to Wayne for the cool project !
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Mid Side M-S Matrix Construction Information

Post by mediatechnology »

I agree that the temp is too low. Have the same problem desoldering leaded radial caps on computer motherboards where the power planes are large and sometimes internal. It's a big heatsink for the iron.

I think my Weller's tips are 700 degree (PTA-7?) and I didn't have a problem getting it hot enough for the MS board. If it were me I'd re-hit some of the pads that were previously soldered at 600 degrees.
Post Reply