A Vario-Matrix Survival Guide

Help Support QuadraphonicQuad:

par4ken

1K Club - QQ Shooting Star
QQ Supporter
Joined
May 28, 2011
Messages
1,143
Location
NW Ontario
There were many great articles published during the brief lifespan of "MCS Review" formerly "Quad Quarterly". Our own Sonik Wiz penned most of the technical articles. I was looking for the copy with his Pre-Synth circuit in it. but came upon his Vario-Matrix Survival Guide. This tells about all you need to know about Vario-Matrix decoding, how to maintain your equipment and even how to modify the various units.

This article is invaluable for me, as I'm about to do some maintenance on my QSD-1.

My required maintenance involves Deoxing the switch contacts. Sometimes I have to punch them in and out to get them to work. At times the sound is distorted until the push button is exercised (a high resistance contact can actually act as a diode which causes this type of distortion. Another problem is at times the bass drops out (another switch contact?) on one or more speakers, I thought that it was a bad input connection to my amplifier (as I Bi-amp), but switching over to the S&IC that problem is gone.

I plan to add the variable Hall and Surround pot to my Photolume decoder when I'm done with the QSD-1. For the QSD-1 I'll check that the hall and surround blend is correct, replacing resistors or using a trim pot if necessary.
 

Attachments

Sonik Wiz

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2005
Messages
2,266
Location
Kansas City
There were many great articles published during the brief lifespan of "MCS Review" formerly "Quad Quarterly". Our own Sonik Wiz penned most of the technical articles. I was looking for the copy with his Pre-Synth circuit in it. but came upon his Vario-Matrix Survival Guide. This tells about all you need to know about Vario-Matrix decoding, how to maintain your equipment and even how to modify the various units.

This article is invaluable for me, as I'm about to do some maintenance on my QSD-1.

My required maintenance involves Deoxing the switch contacts. Sometimes I have to punch them in and out to get them to work. At times the sound is distorted until the push button is exercised (a high resistance contact can actually act as a diode which causes this type of distortion. Another problem is at times the bass drops out (another switch contact?) on one or more speakers, I thought that it was a bad input connection to my amplifier (as I Bi-amp), but switching over to the S&IC that problem is gone.

I plan to add the variable Hall and Surround pot to my Photolume decoder when I'm done with the QSD-1. For the QSD-1 I'll check that the hall and surround blend is correct, replacing resistors or using a trim pot if necessary.
Thanks for the shout out & kind comments! When I last got the urge to do repairs on my Sansui QSD-1000 I went looking for my copy of that issue & could not find it. Anywhere. So thanks to you I have it back in my possession. That article & Uncle Sam's Department of Quad were the two contributions I was most proud of. And also I can not find the copy of the latter article either.

I like to think the Survival Guide was the launch pad for others that expanded on it such as Quad Bob (when his reputation was good) & maybe even QRX Restore.

When I did that article I remember tracing the signal path through the various IC's internal schematic to understand the details of Vario-Matrix operation. These are included in your posted scan but I just did one of the IC's from the original AES White Paper. I think this valuable stuff so here is another image that might be a bit higher quality:

SANSUI CHIPS.jpg


If you look close at HA1327 you can see my pencil notes on the signal mixing & polarity. Some of the features of the Sansui Vario-Matrix design that set it far apart from others was using a phase discriminator to determine direction, the fact that there are no 90 deg phase shifts needed until after decoding ( and then that is a debatable nicety), the use of FET's for the variable matrixing. In fact even this is unique because the FET's work as a variable shunt to ground rather gain controlling the signal as pass through. In other words, unlike SQ full logic decoders the signal that you hear is what didn't go through the gain control elements.

In the September '76 issue of Wireless World there was a QS Vario-Matrix kit article that described a different way to tune up & adjust the pots that control the front, back, left, right, parameters for optimum performance. Unlike mine it requires a simple 90 deg phase shift to check outputs and balancing. Either way works but mine is a bit simpler.


RE: Hall & Synthesis blend on the QSD-1: I've never checked the in phase blending on the Hall setting because that seems pretty tolerant considering the front chs are actually undecoded straight stereo. In the 4>5 units I've checked the dB blend for Synth it has always been a bit less than .414, closer to -10dB at .316. I think Sansui did this a bit conservatively accounting for the amount of already randomy rear ch signal that might be there on a stereo records. It would be good to replace the Synth fixed resistor with a multi-turn trim pot, set right where you think is the best, & leave it.

post back as the QSD-1 restoration project goes!
 

par4ken

1K Club - QQ Shooting Star
QQ Supporter
Joined
May 28, 2011
Messages
1,143
Location
NW Ontario
RE: Hall & Synthesis blend on the QSD-1: I've never checked the in phase blending on the Hall setting because that seems pretty tolerant considering the front chs are actually undecoded straight stereo. In the 4>5 units I've checked the dB blend for Synth it has always been a bit less than .414, closer to -10dB at .316. I think Sansui did this a bit conservatively accounting for the amount of already randomy rear ch signal that might be there on a stereo records. It would be good to replace the Synth fixed resistor with a multi-turn trim pot, set right where you think is the best, & leave it.
I downloaded the previous Wireless World article, and although the author states in both articles that the Hall mode passes the (Lt/Rt) stereo signal direct to the front outputs in Hall mode I believe that is incorrect. The Hall mode mixes the Lt and Rt together by .414 in phase, that makes the signal decode as stereo from the front. I don't know if the author was confused on that point or if it was just in the way he was trying to explain it. Interesting that the article is actually describing another Vario-matrix kit, available at the time. That one came on two boards that plugged together, I think that I would prefer the Photolume that uses only one board!
 

Sonik Wiz

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2005
Messages
2,266
Location
Kansas City
I downloaded the previous Wireless World article, and although the author states in both articles that the Hall mode passes the (Lt/Rt) stereo signal direct to the front outputs in Hall mode I believe that is incorrect. The Hall mode mixes the Lt and Rt together by .414 in phase, that makes the signal decode as stereo from the front. I don't know if the author was confused on that point or if it was just in the way he was trying to explain it. Interesting that the article is actually describing another Vario-matrix kit, available at the time. That one came on two boards that plugged together, I think that I would prefer the Photolume that uses only one board!
Ah Sansui says the front chs are un-decoded stereo as well. Taken from a 4 page promo piece about the QSD-1 titled The Ultimate Decoder/Synthesizer:

" Hall Mode: The 2-channel signals are reproduced through the front channels as ordinary stereo, while the QSD-1 sends ambience signals to the back channels to result in the same kind of rich musical sound you would hear in a concert hall."

You can also trace the signal in hall mode on the Photolume or QSD-2 schematic to verify this. Actually the QSD-1 schematic is easier to see this as it has separate switches for the decoding modes. Take a look at F-2467 illustration. You'll notice that the switch for synthesizer has only 2 poles to it as it just switches in the out of phase (rather opposite polarity) mixing resistor into the circuit. But the Hall mode is 4 pole as it switches in the in phase blend resistor while also routing the 2 ch input to Front L/R.

Are we not geeks? No. We are Quadra-Geeks!
 

Sonik Wiz

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2005
Messages
2,266
Location
Kansas City
More of Uncle Sam
Thank you Ken!
I considered the Patent & Trademark Office a gold mine of worth while information. There is so much hype in advertising & only so much you can take as true in a electronics review in a magazine. But there is relatively little BS in a patent application. In the halcyon days of quad, if you wanted to find out how something worked there were patents & AES papers.

You also found out about cool ideas that never made it to market. Such as Sony having their own variable matrix decoder for both SQ & QS. II think it was Mitsubishi that had a highly advanced RM (not QS) decoder with direction sensing on 8 points where as Sansui used only 4. There was a myriad of schemes for synthing stereo to surround some very intriguing, some down right laughable.

All of this was in the days you had to some how without use of the internet find the patent numbers you wanted, get a USPS money order cuz each patent copy cost 10 cents, and do all this via snail mail. But most of the patents I have came from that time period. I have maybe 12"+ of printed out patents all about audio & surround sound.
 

par4ken

1K Club - QQ Shooting Star
QQ Supporter
Joined
May 28, 2011
Messages
1,143
Location
NW Ontario
Ah Sansui says the front chs are un-decoded stereo as well. Taken from a 4 page promo piece about the QSD-1 titled The Ultimate Decoder/Synthesizer:

" Hall Mode: The 2-channel signals are reproduced through the front channels as ordinary stereo, while the QSD-1 sends ambience signals to the back channels to result in the same kind of rich musical sound you would hear in a concert hall."

You can also trace the signal in hall mode on the Photolume or QSD-2 schematic to verify this. Actually the QSD-1 schematic is easier to see this as it has separate switches for the decoding modes. Take a look at F-2467 illustration. You'll notice that the switch for synthesizer has only 2 poles to it as it just switches in the out of phase (rather opposite polarity) mixing resistor into the circuit. But the Hall mode is 4 pole as it switches in the in phase blend resistor while also routing the 2 ch input to Front L/R.

Are we not geeks? No. We are Quadra-Geeks!
I'm still scratching my head about that one. I see the front bypass now on the Photolume but still not in the Wireless World article, it's hard to see with diagrams that extend across multiple pages.

That amounts to half decoding and why use in-phase blend which will actually cancel out some of the ambience that we are trying to reproduce? Hall mode is one that I rarely use, I would rather use SQ or the stereo enhance setting of the S&IC with the pot all the way down. I don't mean to second guess the designers but WTF!

Your pre-synth gave Hall and Surround by simply adjusting a pot, but the fronts were decoded too! I'm still looking for that article, I saw it not long ago. It's just so easy to misplace those small little magazines!
 

Sonik Wiz

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2005
Messages
2,266
Location
Kansas City
I'm still scratching my head about that one. I see the front bypass now on the Photolume but still not in the Wireless World article, it's hard to see with diagrams that extend across multiple pages.
I agree; I do not see the Hall front ch bypass on the WW article either. This article was one of three that also included an SQ & CD-4 kits from Compcor. Maybe the WW QS article just had a schematic error but was done properly on the kits. Who knows?

That amounts to half decoding and why use in-phase blend which will actually cancel out some of the ambience that we are trying to reproduce? Hall mode is one that I rarely use, I would rather use SQ or the stereo enhance setting of the S&IC with the pot all the way down. I don't mean to second guess the designers but WTF!
In another article that would take me hours to find (my closet has a really poor search function) I remember that the in phase blending was to keep any direct sounds from coming out of the rears. Not decoding the front chs added to the over all integration of the sound field. There's an abundance of classical/jazz made with close micing & multi-micing that gave a superficially impressive wide stereo image. In surround, well, I gues some folks don't like the idea of a cymbal crash coming out of left back scaring them.

I didn't use it much unless it was a poorly recorded live concert or something clean & pristine like a Sheffield Direct to Disc. In the former noise & distortion can emphasized in the rear chs & Hall helped with that. In the latter I just liked the least amount of processing to honor the high quality source.

In some ways the Hall decode mode can be seen as effectively close to an EV 4 decoder with stereo. But the Sansui method still had L/R enhancement in the rears so you could still have good depth along with good width between the rear speakers.

Your pre-synth gave Hall and Surround by simply adjusting a pot, but the fronts were decoded too! I'm still looking for that article, I saw it not long ago. It's just so easy to misplace those small little magazines!
Roger, that!
I am in pre-construction on a new front end for my SMv2. It will include L/R balance & Phase balance & be remote controlled. I gave some extended thought to using just out of phase blending? Or did I want in phase blending as well? I decided on the latter as it just seem more complete to me. I then started to look at various circuits most of which were centered on the M-S matrix encode/decode. That seemed a little bit more complex than it needed to be. So I've settled on the circuit shown in Fig 2 or 3:


The only parts I've recieved so far is a nice Avel torroidal Xfmr and 2, motorized ALPS 10k linear pots for the balance controls.
 

par4ken

1K Club - QQ Shooting Star
QQ Supporter
Joined
May 28, 2011
Messages
1,143
Location
NW Ontario
Roger, that!
I am in pre-construction on a new front end for my SMv2. It will include L/R balance & Phase balance & be remote controlled. I gave some extended thought to using just out of phase blending? Or did I want in phase blending as well? I decided on the latter as it just seem more complete to me. I then started to look at various circuits most of which were centered on the M-S matrix encode/decode. That seemed a little bit more complex than it needed to be. So I've settled on the circuit shown in Fig 2 or 3:
I'm still looking for your original circuit. I remember you placing a capacitor in the circuit so that the bass wouldn't cancel with the out of phase blend. I was thinking of trying that with my simple decoder, although it does seem fine as is. Maybe you could add a bypass for the front channels, to see if that sounds better in Hall Mode, I can't really see it myself though.

Too bad there are no circuit boards available from the Eliot site for the "Width Controller" project, It would of made construction a breeze. I bought boards from them in the past for an active crossover.

While searching I came upon my latter SQ encoder design, with the stereo to SQ synth feature. I forget if it was published in the MCS Review or not but I remember building several of them. Now today I can't find it looking in the same place but did I find the instructions for my Audionics 106A Decoder. That one takes me back to almost the beginning of my Quad journey!
 

Sonik Wiz

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2005
Messages
2,266
Location
Kansas City
I'm still looking for your original circuit. I remember you placing a capacitor in the circuit so that the bass wouldn't cancel with the out of phase blend. I was thinking of trying that with my simple decoder, although it does seem fine as is. Maybe you could add a bypass for the front channels, to see if that sounds better in Hall Mode, I can't really see it myself though.

Too bad there are no circuit boards available from the Eliot site for the "Width Controller" project, It would of made construction a breeze. I bought boards from them in the past for an active crossover.

While searching I came upon my latter SQ encoder design, with the stereo to SQ synth feature. I forget if it was published in the MCS Review or not but I remember building several of them. Now today I can't find it looking in the same place but did I find the instructions for my Audionics 106A Decoder. That one takes me back to almost the beginning of my Quad journey!
Hey Ken
This continued exchange has motivated me to go where I've never gone before. Well, close enough in a very, very long time. To the bottom & the very back of my quad lit closet. There I found all the Quad Quarterly & MCS Reviews that I thought were lost forever. I found my pre-synth article you've mentioned from Summer 1982 (the same year I was married. Oh my poor wife...). It also includes your fine contribution Building a Quad Panning Circuit. I've also scanned that. Your schematic looks much better than mine. I had not yet discovered Letraset press type for electronics. But I think you can easily copy/transpose the circuit to something better. I haven't compared wire to wire yet but it looks almost identical to the ESP width controller, just adding a cap to for low pass roll off.
 

Attachments

Top