Sansui QRX7001 repair

[SPXER]

I got the surround boards F-2088 and F-2087 recapped and lifted the blend resistors.
I did some listening last evening. After a bit of adjusting the front-back and left to right levels, I got the right combo. It just got better and better as my ears/brain adapted to the new sound.
Some recordings are more surround rich than others. Notable was Radiohead. Tool was intense, but Tool is intense in stereo as well.
I want to do the decoder alignment next, so I will have to find a signal generator app that will do all that is required. Any suggestion on an app?

 

[par4ken]

I got the surround boards F-2088 and F-2087 recapped and lifted the blend resistors.
I did some listening last evening. After a bit of adjusting the front-back and left to right levels, I got the right combo. It just got better and better as my ears/brain adapted to the new sound.
Some recordings are more surround rich than others. Notable was Radiohead. Tool was intense, but Tool is intense in stereo as well.
I want to do the decoder alignment next, so I will have to find a signal generator app that will do all that is required. Any suggestion on an app?

I use Adobe Audition 3. You can generate tones with most audio programs. Audacity is a free one.

Audacity | Free Audio editor, recorder, music making and more!

 

[SPXER]

So, I adjusted on the F-2088 board using the 4channelsound.com instructions. I could not get the front and rear speakers to be equal (using my vintage Realistic sound level meter). No way I could tell by listening. Turn your head slightly and the level and direction of the tone moves around. The speakers are not equal in reality in the least. I think the small rear speakers must be more efficient. I have VR4 and VR3 at the extremes and that is as close as it would get. I am not at all sure this is the correct adjustment. Is the adjustment to get the same sound level regardless of the speakers or the same wattage output?

 

[par4ken]

The alignment procedure is easiest if you have an oscilloscope, preferably dual trace. Without a scope you could use a multimeter or simply adjust by ear.

Ideally the outputs should be equal but I would simply adjust for minimum sound level.

So follow the procedure and simply adjust for minimum output. With an equal level in phase signal (0°) adjust VR2 for minimum from the back. With an equal out of phase signal (180°), adjust VR1 for minimum from the front. Left input only adjust VR4 for minimum on the right (F&B). Right input only adjust VR3 for minimum on the left (F&B).

Repeating the procedure over again might prove beneficial just in case there was any interaction and might also help to get the channel leakage levels better balanced.

 

[SPXER]

The alignment procedure is easiest if you have an oscilloscope, preferably dual trace. Without a scope you could use a multimeter or simply adjust by ear.

Ideally the outputs should be equal but I would simply adjust for minimum sound level.

So follow the procedure and simply adjust for minimum output. With an equal level in phase signal (0°) adjust VR2 for minimum from the back. With an equal out of phase signal (180°), adjust VR1 for minimum from the front. Left input only adjust VR4 for minimum on the right (F&B). Right input only adjust VR3 for minimum on the left (F&B).

Repeating the procedure over again might prove beneficial just in case there was any interaction and might also help to get the channel leakage levels better balanced.

I'm having difficulty with this. I am reading your instructions, 4channelsound's instructions and Varner's instructions. I am placing the meter leads into the speaker connectors on the receiver, red lead to right front +, and black lead to right back +. I turn VR4 and it does not go to 0. With volume at max it goes from 9-13v ac. Am I doing it wrong?

 

[Sonik Wiz]

I'm having difficulty with this. I am reading your instructions, 4channelsound's instructions and Varner's instructions. I am placing the meter leads into the speaker connectors on the receiver, red lead to right front +, and black lead to right back +. I turn VR4 and it does not go to 0. With volume at max it goes from 9-13v ac. Am I doing it wrong?

Firstly I would suggest NOT measuring from the speaker outputs but rather from the 4 ch tape output. This will bypass any level or balance inaccuracies downstream. If you get things tuned up ok using the line level tape outs but things still sound strange when listening over speaker then you know there's a problem other than decoding.

Also maybe I missed it... but what is your program source for test signals? Lets say for example you've made test signals burned to a CD or USB drive plugged into a disc spinner. Then play center front where Left = Right and measure the output. If they don't match say with in maybe .5 dB then you are not going to be able to tune up accurately no matter what. You would have to change/adjust your program source.

I would be glad to share test signals I made for QS if you want. Each one is 2 mins long allowing for a decent amount of time to adjust. Created in AA 3 these are white noise and usually sine waves would be used but I've used them in preference with good success. Just let me know if you want them.

 

[Sonik Wiz]

@SPXER Addendum: Maybe I'm stating the obvious but make sure you are adjusting in basic QS mode, not Hall or Surround Synthesizer.

Also the 4 ch tape out level will be quite lower than working with the speaker outs. Most gear of that vintage listed .5V as nominal rated output. This makes it very prone to hum pick up if you are just hooking DMM gator clips to the output. Instead just take an RCA phono cord, clip one end & hook up banana plugs to it. Of course the coax shield goes to black negative, core conductor goes to red positive. And try to keep it short length.

 

[par4ken]

I'm having difficulty with this. I am reading your instructions, 4channelsound's instructions and Varner's instructions. I am placing the meter leads into the speaker connectors on the receiver, red lead to right front +, and black lead to right back +. I turn VR4 and it does not go to 0. With volume at max it goes from 9-13v ac. Am I doing it wrong?

I would suggest just going by ear. And make sure to do the adjustment with the QS mode setting, the procedure will not work correctly if Hall or Surround is selected. You would likely have to turn up the volume of the speakers that you are adjusting the null for while turning the other two completely off.

If Sonik's test signals include corner signals (Lf, Rf,Lb.Rb) you could use those to verify the alignment results.

If you can't get proper adjustment then one or more of the decoder chips could be bad. Replacements are available on eBay.

 

[SPXER]

Firstly I would suggest NOT measuring from the speaker outputs but rather from the 4 ch tape output. This will bypass any level or balance inaccuracies downstream. If you get things tuned up ok using the line level tape outs but things still sound strange when listening over speaker then you know there's a problem other than decoding.

Also maybe I missed it... but what is your program source for test signals? Lets say for example you've made test signals burned to a CD or USB drive plugged into a disc spinner. Then play center front where Left = Right and measure the output. If they don't match say with in maybe .5 dB then you are not going to be able to tune up accurately no matter what. You would have to change/adjust your program source.

I would be glad to share test signals I made for QS if you want. Each one is 2 mins long allowing for a decent amount of time to adjust. Created in AA 3 these are white noise and usually sine waves would be used but I've used them in preference with good success. Just let me know if you want them.

I am using a phone app called Keuwisoft Dual Channel Function Generator. It does left or right or both in or out of phase freq or noise.
Using the tape rec RCA jacks on the receiver, RF and RB, in QS with 1k hertz in left aux input. I get 158 mv ac. VR4 does not change the reading. I read 173mv at the phone.

 

[SPXER]

@SPXER Addendum: Maybe I'm stating the obvious but make sure you are adjusting in basic QS mode, not Hall or Surround Synthesizer.

Also the 4 ch tape out level will be quite lower than working with the speaker outs. Most gear of that vintage listed .5V as nominal rated output. This makes it very prone to hum pick up if you are just hooking DMM gator clips to the output. Instead just take an RCA phono cord, clip one end & hook up banana plugs to it. Of course the coax shield goes to black negative, core conductor goes to red positive. And try to keep it short length.

I am using an RCA patch cord, cut in two and hooked center conductors to the meter leads. Connected to RF and LF rec tape. Ground not connected.

 

[SPXER]

I would suggest just going by ear. And make sure to do the adjustment with the QS mode setting, the procedure will not work correctly if Hall or Surround is selected. You would likely have to turn up the volume of the speakers that you are adjusting the null for while turning the other two completely off.

If Sonik's test signals include corner signals (Lf, Rf,Lb.Rb) you could use those to verify the alignment results.

If you can't get proper adjustment then one or more of the decoder chips could be bad. Replacements are available on eBay.

At first, I tried by ear, but I found it confusing. I could hear something changing, but I could not tell what/where.

 

[par4ken]

If you can't adjust it usually means that you have a blown chip. Which one? If you delved into it further you might be able to determine that. I usually just replace them all. Use sockets for the replacements. With sockets installed you can always plug the old chips back in (one at a time) to determine which one is bad, if you want. The F-2088 board uses two HA1327 and one HD3103P. The F-2087 uses one HA1328.

Before doing that I would double check the manual to verify which pot to adjust and for what. Some manuals don't show details of the Vario-matrix boards, just showing them as blocks. I checked the QRX7001 schematic in the manual on HiFi Engine and it does show the board details.

I am using a phone app called Keuwisoft Dual Channel Function Generator. It does left or right or both in or out of phase freq or noise.
Using the tape rec RCA jacks on the receiver, RF and RB, in QS with 1k hertz in left aux input. I get 158 mv ac. VR4 does not change the reading. I read 173mv at the phone.

Did you try any of the other adjustments?

 

[SPXER]

Did you try any of the other adjustments?

I did not. Only VR4. I installed my spare 2088 board. Same result. The surround does work. I do not know if correctly or not. I will try adjusting the other VR's.
I suspect I do not have the hook-up correct, but I do not see what it could be.

I tried all 4 VR's. No change out of any.

 

[SPXER]

Picture of connections. Left AUX to phone at 1KHertz. Tape REC center connections to each meter lead.

 

[Sonik Wiz]

I am using an RCA patch cord, cut in two and hooked center conductors to the meter leads. Connected to RF and LF rec tape. Ground not connected.

Are you saying something like right front center conductor goes to black minus input on the meter? And back right center conductor goes to red pos input on the meter? Or vice versa? Unless you are using a DMM with a true differential input you aren't measuring anything worthwhile. And even then you would need a ground connection for a reference point.

Maybe I can catch @par4ken attention & have him double check my interpretation?

Hook up one RCA cable the way I described above. Apply a i KHz signal to left only input. Adjust VR 4 for lowest output from right front. Now switch cable to right back & compare. Be nice if they match but probably wont. If one has greater separation (lower output) than the other, then adjust the lower output ch with VR 4 for a little higher output until left front & left back are as low as they can go but still matching. Yes it's tedious & maybe time consuming but this is really the only way unless you've built some custom DIY test gear. Or use a dual trace scope as mentioned before.

And front/back adjustments will effect the left/right to a degree because it is all interactive. It takes a lot of toggling back & forth to get it all uniformly tweaked.

 

[SPXER]

Are you saying something like right front center conductor goes to black minus input on the meter? And back right center conductor goes to red pos input on the meter? Or vice versa? Unless you are using a DMM with a true differential input you aren't measuring anything worthwhile. And even then you would need a ground connection for a reference point.

Maybe I can catch @par4ken attention & have him double check my interpretation?

Hook up one RCA cable the way I described above. Apply a i KHz signal to left only input. Adjust VR 4 for lowest output from right front. Now switch cable to right back & compare. Be nice if they match but probably wont. If one has greater separation (lower output) than the other, then adjust the lower output ch with VR 4 for a little higher output until left front & left back are as low as they can go but still matching. Yes it's tedious & maybe time consuming but this is really the only way unless you've built some custom DIY test gear. Or use a dual trace scope as mentioned before.

And front/back adjustments will effect the left/right to a degree because it is all interactive. It takes a lot of toggling back & forth to get it all uniformly tweaked.

I thought I was to measure the differential of front to back. Like a Hafler circuit works from left to right channels. My misunderstanding. I will start over. Thanks.

 

[SPXER]

I was able to hear the speaker pair in test better by turning the other pair down to zero. I did get the speakers to minimize/balance that way, except for VR1. I could not hear any difference in adjusting VR1.
The readings from the tape outs never changed, even with the rewiring as per Sonik Wiz. I will have to test the tape outs for functionality. I do get numbers. Way off from each and not changing with VR adjustment.

 

[par4ken]

The readings from the tape outs never changed, even with the rewiring as per Sonik Wiz. I will have to test the tape outs for functionality. I do get numbers. Way off from each and not changing with VR adjustment.

Are you sure that the tape output jacks connect to the QS decoder? AFAIK very often the matrix decoder does not feed to the Tape Recorder inputs. It is normally thought that you would record stereo and matrix as stereo (2CH) and then play back through the decoder. It doesn't make a lot of sense to record from the decoder outputs using up twice the tape making a discrete tape that really isn't discrete.

If you had preamplifier output/amplifier input jacks you could use those by removing the jumper plugs but it doesn't appear that that unit has that feature.

The signals are not in phase on the sides so a differential reading would be meaningless there. Even across the front and back adjusting for a null there would just mean that any crosstalk is evenly balanced. Best to simply adjust for minimum output from each in turn.

I could not hear any difference in adjusting VR1.

It sounds like you are mostly in business, VR1 should do something if everything is working properly. Did you say that you have two 2088 boards? Do both adjust the same? There might be a problem with the 2087 board (HA1328). If the leakage is low enough you might want to just call it a day and leave it as is for now.

 

[SPXER]

Are you sure that the tape output jacks connect to the QS decoder? AFAIK very often the matrix decoder does not feed to the Tape Recorder inputs. It is normally thought that you would record stereo and matrix as stereo (2CH) and then play back through the decoder. It doesn't make a lot of sense to record from the decoder outputs using up twice the tape making a discrete tape that really isn't discrete.

This makes sense.

If you had preamplifier output/amplifier input jacks you could use those by removing the jumper plugs but it doesn't appear that that unit has that feature.

The signals are not in phase on the sides so a differential reading would be meaningless there. Even across the front and back adjusting for a null there would just mean that any crosstalk is evenly balanced. Best to simply adjust for minimum output from each in turn.

So, take the readings from the speaker connectors? How am I assured the amp is at a balance? Wouldn't it be better to take the readings from the pin-outs 2-5 on the 2087 board?

It sounds like you are mostly in business, VR1 should do something if everything is working properly. Did you say that you have two 2088 boards? Do both adjust the same? There might be a problem with the 2087 board (HA1328). If the leakage is low enough you might want to just call it a day and leave it as is for now.

I have a spare 2087 as well. The ones I am working with have been recapped.

 

[Sonik Wiz]

This makes sense.

So, take the readings from the speaker connectors? How am I assured the amp is at a balance? Wouldn't it be better to take the readings from the pin-outs 2-5 on the 2087 board?

I have a spare 2087 as well. The ones I am working with have been recapped.

I don't have any additional insights to offer at this point. I mainly just want to say how impressed I am at your passion & involvement on this project. If any home DIY restore project deserves to succeed it's this one. Best of luck as time goes on.

Yes I think it is best to measure just the decoded outputs prior to level/balance controls. And FWIW the only quad receiver I ever owned, a Kenwood 9940, did pass the SQ/CD-4 decoded outputs through the 4 ch tape out. The Sansui might be different. I certainly think the layout & labeling on the back panel is rather confusing.