MS circuit

Im trying to put together a small board for MS encoding and decoding.

I found one using the THAT1240 chips and put an existing debalancing and balancing stage on it.

Before I start putting it on a vero board and rey ro add a hard bypass and mute function, I would like to see if you can help me out.

1. Any "best" ways to compensate for the 6dB gain in the 1646 chip?
2. Is is possible to add a level control pot, and if so, how would I do that? (Would like to know how it works, not just how to do it if possible - if possible)

Gustav

http://groupdiy.com/index.php?topic=37821.msg465860#msg465860

Found this quite interesting. The last part about how to implement the compensation. It seems there are a few ways how it can be done.

Been searching the web about this matter for quite some time too. I find it difficult to understand how to get unity because of the unpredictability of the correlation between sources.

I was thinking about the level control. I want let's say +/- 3dB. Can we compensate -3dB instead of -6dB put a pot after it and buffer it again? That way we can leave one gain stage out of the chain?

Just found out the THAT chip comes in 1203 and 1206 variants.

Still would like something to control the insert send level and the insert return level.

Gustav

There is another way though no insert in this schematic:

Pip said:

There is another way though no insert in this schematic:

Click to expand...

Thanks for sharing.

Id prefer to avoid using a transformer for this particular circuit.

Gustav

Isn't that the Wayne Kirkwood circuit? Wouldn't just buying his PCB's be easier and cheaper?
http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=7&t=346

Gold said:

Isn't that the Wayne Kirkwood circuit?

Click to expand...

Found this pic for the MS stuff, so it seems you are right (at least partially)

http://www.ka-electronics.com/images/jpg/M-S_Simple.jpg

Added the rest from the balancing/debalancing boards I have already done earlier.

Is this a no-no to use? I am used to THAT circuits I find being common, so I didn't think of this after finding it - actually selected it so quickly without looking further because it was THAT based.

Gold said:

Wouldn't just buying his PCB's be easier and cheaper?

Click to expand...

Might be, but I would like to do a board, if nothing else, just to do it.

I would also still like to add an attenuator (maybe pre and post the insert point), but I am not sure if its as simple as adding a voltage divider where needed. Will look into it when the chips are in
Gustav

Read the Kirkwood threads.  EVERY aspect has been documented. 

http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=7&t=262

Gold said:

Isn't that the Wayne Kirkwood circuit? Wouldn't just buying his PCB's be easier and cheaper?
http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=7&t=346

Click to expand...

Just to second Paul's recommendation.

Also to point out that if you are worried about crosstalk, LAYOUT & WIRING becomes far more important than 0.1% resistors.  Buy Wayne's PCB.

If you get layout & wiring right, you can just use 1% resistors from the same batch from a reputable manufacturer.  The actual value won't be always within 0.1% of nominal but the standard deviation will be well under 0.1% for a good maker.

And furthermore, if you include twiddle pots & caps, you'll be able to get good crosstalk at 2 frequencies but it will be poorer everywhere else cos the asymmetry introduced in the layout.  Hence I'm not surprised Igor got poor crosstalk on his breadboard.

The consistant "Dont DIY" answers are making me realize I may be stepping on some toes here.

I certainly didn't mean to. I found the MS portion of the circuit and applied the other building blocks from schematics I had already put into Eagle to do balancing and debalancing boards a while back.

The THAT applications I come across are usually standard and even described in their own application notes, so I didn't think too much about this one. I will seek a different solution that does not violate someone else's work, but I do want to do a layout to learn and create another building block for my own library of tested layouts and applications to go into larger context.

My apologies, and cheers.

Gustav

Gustav said:

The consistant "Dont DIY" answers are making me realize I may be stepping on some toes here.

Click to expand...

I don't think anyone is trying to dissuade you.  But Wayne's PCB has all the bells & whistles and you can use as much or as little as you want.

My own comments are only about crosstalk.  If you are actually using MS, crosstalk is a moot point. 

I did the Stereo Width module for Calrec M-series .. basically MS with a pot to change relative gains.  I just specified that the 'zero crosstalk' setting had to be within the centre detent on the pot.

ricardo said:

Gustav said:

The consistant "Dont DIY" answers are making me realize I may be stepping on some toes here.

Click to expand...

I don't think anyone is trying to dissuade you.  But Wayne's PCB has all the bells & whistles and you can use as much or as little as you want.

My own comments are only about crosstalk.  If you are actually using MS, crosstalk is a moot point. 

I did the Stereo Width module for Calrec M-series .. basically MS with a pot to change relative gains.  I just specified that the 'zero crosstalk' setting had to be within the centre detent on the pot.

Click to expand...

I dont see it as a negative thing if I was simply mistaken in seeing the MS portion as a standard application.

Doing this would not be a "make-it-super-high-spec" project for me, I am simply not smart enough to aim for that. I do enjoy it as a learning experience, and the risk of violating someone else's work would surely not be fun, which is why I am addressing the sense I got in that direction. I have done a few things based on other THAT chips, which were all part of THATs own application notes, and I thought the encoding/decoding here was too.

I see it like painting a tree before doing a huge landscape painting, and if I buy ready-made tress, it won't be a step towards something more

I will Vero the MS portion (already have small PCBs for the balancing and debalancing portion), and I will play around with voltage dividers for MS blend controls

Gustav

If I'm not mistaken, Wayne wrote the app notes on THAT parts for MS.  Point being to sell THAT parts, as well as his PCB kit.  The board, laid out by Roger Foote, is as much a one stop "from the horses mouth" solution as one could want.  So just don't do unnecessary work unless you just want to. 

Gustav said:

Doing this would not be a "make-it-super-high-spec" project for me, I am simply not smart enough to aim for that.

Click to expand...

You meant to write "I simply don't care enough" or something like that, right?... Tsk tsk Dr. Jantelov

VictorQ said:

Gustav said:

Doing this would not be a "make-it-super-high-spec" project for me, I am simply not smart enough to aim for that.

Click to expand...

You meant to write "I simply don't care enough" or something like that, right?... Tsk tsk Dr. Jantelov

Click to expand...

I am not sure what you mean, or why you felt a need to post this, but what I meant was exactly what I wrote. I am trying to learn how it even works and get some MS processing happening in a piece of hardware, so fine tuning the last .01% is not within my scope. Its a pretty common curve to learn and apply things of gradually increasing complexity, and I am not sure how my place and level of ambition on this particular curve relates to the term "Jantelov".

Gustav

I like "not within my scope" much better.
I am sorry if I was projecting, I certainly didn't mean to put you down. Quite the contrary. I think you have demonstrated here and elsewhere that you are "smart enough" to do stuff...

VictorQ said:

you are "smart enough" to do stuff...

Click to expand...

I much prefer creating something concrete and moving on rather than contemplating grand schemes that never amount to anything. To some it may be a "Jante"-mentality, but I prefer to see it as a realist approach.

I spent 6 years obtaining a degree in philosophy and political theory and a few years after that working in public administration. It really burned me out on the whole "grand scheming, no action" approach to learning, working and living

So this is step one. I hope to turn it into a rats nest with levels and bypass before the week is over.

Gustav