Sansui Vario-Matrix Blend Resistors

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

QuadBob

400 Club - QQ All-Star
QQ Supporter
Since 2002/2003
Joined
Jul 9, 2003
Messages
430
Location
Las Vegas, NV
I've been asked the question of "Why did Sansui put blend resistors into the VM decoder?" so many times, I thought I'd post an answer here for everyone:
As to why Sansui put them in there.......two reasons;
1) the blend resistor between the left (total) input and the right (total) input helps to insure that the decoder gets a balanced input. Decoding ability is very dependent on the input being balanced......the better the balance, the better the decoding ability.
This was an issue in the '70s......turntables, 8track decks, cheap RCA cables, etc.....all could have 2 channel outputs that weren't as "balanced" as they could be. The blend resistor insured that the decoder would never see just a left or right input......always both channels.
However, the decoder operates best the higher the separation is between the inputs, as this gives it more information to deal with. And, with todays digital sources (CD, etc,).....they have "infinite" separation on their
output.......and a perfectly balanced output too!
So, get rid of the blend resistor and performance (and output separation increases dramatically! From the stock rating of 20db between the channels......to over 40db and technically only limited by the source material.
2) There are two more blend resistors between the LF and LR, and RF and RR signals. These were put in to insure that there was always "something" in the rear channels. Even if the decoder didn't pick up any "out of phase" information for the rear channels......you'd still have sound there at a
lower level dependent on the value of the blend resistor.
In other words.......no matter what, all four channels would have output. This is true even if there is NO out of phase information in the source signal. You can easily check this yourself........play something that is "mono" into the decoder.
Based upon the accuracy and capability of the Sansui QS
chips.....you should have nothing in the rear channels at all (there is no out of phase info available). But, you will notice that you still have sound in the rear channels.......basically the same as the front channels only at a lower level......... you are hearing the "bleeding" that the blend resistors introduce.
Remove them and the decoder is totally controlling the
rear channel output.........you will ONLY hear out of phase information and/or the "horseshoe" output of the "surround" mode where the out of phase info is combined with the Hard Left/Hard Right signals moved to Rear Left and Rear Right as designed. Again, you've got dramatically increased
separation in the rear channels........depending on the source material, you may have little or no output in the rear channels now.........but the rear channels are much more discrete, and much more exciting when they are getting the signals from the chips!
At the time Sansui designed the QSD-1/QSD-2 and 001 receivers VM pcbs.......the above reasons were why they put the blend resistors in. And, they thought limiting the separation to 20db was still better than any other
decoder on the market!
It wasn't until the Audionics S&IC and TATE II that they realized that we wanted ALL the separation we could get!! <G> It is notable that they did away with the LT/RT blend resistor in the QSD-1000 that was produced in 1982. By then they knew that the best decoding was the most discrete decoding that you could obtain! Although they still put blend resistors between LF/LR and RF/RR to still keep all four channels active. But, just the removal of the LT/RT blend resistor raised its separation rating to 25db in the QSD-1000.
So, if you really want to compare how QS decoded through any of the VM decoders compares to SQ through a TATE or S&IC, you should remove the blend resistors, recalibrate the boards......then you've got an Apples to Apples comparison!

QuadBob:cool:

 
I'm very interested in this, since I've pretty much completed a total rebuild of my QRX-8001. The blend resistors are the final thing. I have the schematics so really just need to know the board locations/numbers of the resistors.

can you email me at dave(at)madgroup.co.uk

replacing (at) with @ for anti-spam reasons,

Thanks.:D
 
Hey Dave,

I seriously thought about just including the resistor numbers in my previous post. But, I thought it better not to because:

1) The two blend resistors for LF/LR and RF/RR are really crammed into a very crowded location on the pcb, and are poorly marked on the board itself.......it is very easy to unknowingly cut the "wrong" resistor. Which of course would really screw up the decoder.......and of course they would be flaming me for their mistake.

2) The decoder board calibration should at the least be verified, and preferably a complete calibration should be done to insure proper response and calibration afterwards.

3) Anyone with the technical ability to read schematics should be easily capable of identifying the resistors themselves, as I pretty much did tell you exactly where they are located in the signal path. If you have the service literature.....schematic and pcb board layout prints you can then physically locate the resistors and desolder one side. (all you need to do is lift one side to break the path).

So, I hope you understand that I wanted to explain both the reasoning behind their inclusion, and the results once removed...without having everyone out there jump into some fairly complex circuitry with a soldering iron or side-cutters!
I'll send you an email Dave to discuss your particular needs.

QuadBob:cool:
 
Thanks Bob. I see where you're coming from. Look forward to hearing from you soon.



Dave. :D
 
If some one has a schematic of the QSD-2 I would really appreciate getting a copy so I can remove the blens resistors. this sounds great!

Norm
 
OK Gentlemen,

I know a number of you have now experimented with removing the blend resistors on your Vario-matrix decoders.

Comments????:smokin:

Or are you too busy re-listening to all your music to write now?:rollin:

QuadBob:cool:
 
OK Gentlemen,

I know a number of you have now experimented with removing the blend resistors on your Vario-matrix decoders.

Comments????:smokin:

Or are you too busy re-listening to all your music to write now?:rollin:

QuadBob:cool:


I would have liked to hear replies on this too.

I'm googling the web and chasing links tonight and a few end up right back here at QQ so I'm replying to really old posts tonight.

Why, I don't know :)

I just want to say that regarding blend resistors, if a given matrix relies on out-of-phase information for it's rear channels, then blending the front channels cancels them out in the front, sacrificing front left to front right separation in exchange for greater front to rear separation. That's the whole game right there, I just never knew Sansui did it.

I'm very interested in replacing the blend resistors on non-logic SQ decoders with variable potentiometers, and I can't believe no one else already did it. It should be possible to vary SQ from anything between it's original form to a symmetrical distribution pattern like QS, to a movie setting that favors front to rear effects. I'm very interested in hearing an SQ non-logic decoder without the blend resistors.
 
Hey Dave,

I seriously thought about just including the resistor numbers in my previous post. But, I thought it better not to because:
........

So, I hope you understand that I wanted to explain both the reasoning behind their inclusion, and the results once removed...without having everyone out there jump into some fairly complex circuitry with a soldering iron or side-cutters!
I'll send you an email Dave to discuss your particular needs.:

Any chance I could get you in to a discussion about the location and removal of the blending resistors in a Sansui QRX-7001 (F-2087 and F-2088)?

Thanks.
 
In Autumn of 1984 Scott Varner wrote an article in MCS Review about the Vario Matrix System and talked about the blind resistors and the effects that takes place when some of the blind resistor are replaced with Pots.

:sun
I Love QSD-1
 
supid question, are these blend resistors only in the QSD's and the X001 models, or are the in all the QRX models? would this procedure benefit a X500 model? thanks
 
On board 2088 lift one leg (or snip with cutters) R-601 which is the Left/Right Total blend resistor. On board 2087 lift or snip R-39 and R-40. (I know this is an old thread, but I do not feel QB has a right to withhold this info. So, I am making it available to anyone doing a search for this. QB is already "withholding" too many peoples' cash and equipment.)
 
On board 2088 lift one leg (or snip with cutters) R-601 which is the Left/Right Total blend resistor. On board 2087 lift or snip R-39 and R-40. (I know this is an old thread, but I do not feel QB has a right to withhold this info. So, I am making it available to anyone doing a search for this. QB is already "withholding" too many peoples' cash and equipment.)

QB is one of those scummy collectors who want to be the sole source of information for the gratification of his own ego.
 
I would like to know this also.
I have a repair manual with circuit diagram. I have a guess but I know next to nothing about electronics so probably a bad guess.
 
QSD-1 has F-2466 phase shifter board, F-2462 for filter board. Also F-2463(low), F-2464(mid), and F-2465(high) vario matrix boards.
Anybody know where the blend resistors are on these boards?
 
Back
Top