DIY SQ Decoders?

Hi folks,

A bit of a quad noob here. While I have noticed that Full Logic SQ decoders seem a bit scarce/pricey, the Motorola chips for them seem pretty easy to find, for fairly reasonable money. Is there anyone that has done, or is doing, newer kits or products based on these?

Also, in a related question, I found one set of CBS/Motorola documentation that had three suggested SQ decoder circuits. The first one (L-1) used MC1312, MC1314, and MC1315 in a fairly plain implementation; L-2 added variable blend, with a J-FET between LB and RB, and L-3 was similar to L-2, but replaced the MC1312 with discrete components for the matrix decoder.

Anyone have experience with units that used any of these circuits? I'm thinking about building an L-3, maybe, if its performance seems worthwhile. Right now, I have nothing to decode SQ... except a Neural processor in my Pre/Pro. I tried this with a couple of SQ albums, and it sounded OK (pretty good, actually, on Caravanserai), but nothing like the original Quad mix off of the Six Wives of Henry the VIII DVD-A.

Cheers,
Paul

Hi Paul,

Welcome to the forum.

Have you considered software decoding? There's software and scripts that can take a two channel SQ source, process it, and leave you with four channels of decoded material.

Hi Mike,

Yes. I've seen those. From what I understand, though, they're basically the pre-logic decoding, with no steering or blend, and deliver only about 3dB of front/rear separation, as opposed to the 15dB from the CBS/Motorola circuits.

As you are new you may not have heard of the
Surround master
If you want the best SQ decoding and the best QS decoding
this is all you need it is all in 1 unit
From Australia Look up the tests

SQLGuy said:

Hi Mike,

Yes. I've seen those. From what I understand, though, they're basically the pre-logic decoding, with no steering or blend, and deliver only about 3dB of front/rear separation, as opposed to the 15dB from the CBS/Motorola circuits.

Click to expand...

Hey Paul.

That's something different, some audio software will do a non-real time processing with phase shifting and centre channel extraction to produce much better than 3dB separation. So basically you need the source, run the script, and then you end up with two WAV files for front and rear that can be burned to disc. It can take some time, but the results are good.

I have schematics for a full logic that I assembled from a kit that I purchased in the early 70's using those IC chips. I still have the decoder and it still works. It is comparable to the Sony SQD 2020 in performance.

If you are interested in the schematics let me know...

I would be interested in that. The L1, L2, and L3 circuits I was referring to are from this document: http://www.ka-electronics.com/images/pdf/CBS_Motorola_SQ_Decoders.pdf

Looks like the main advantages of L3 are lower distortion and lower noise by using the discrete matrix instead of the MC1312.

I would be interested.

The L1, L2, and L3 circuits I mentioned are from this document: http://www.ka-electronics.com/images/pdf/CBS_Motorola_SQ_Decoders.pdf

Looks like the performance of L3 is the same as L2, but with less noise and distortion due to the discrete matrix instead of MC1312.

Mike the Fish said:

Hey Paul.

That's something different, some audio software will do a non-real time processing with phase shifting and centre channel extraction to produce much better than 3dB separation. So basically you need the source, run the script, and then you end up with two WAV files for front and rear that can be burned to disc. It can take some time, but the results are good.

Click to expand...

Are there any of these that don't require Adobe Audition? The cost of Audition alone makes a lot of hardware solutions look pretty appealing.

rustyandi said:

As you are new you may not have heard of the
Surround master
If you want the best SQ decoding and the best QS decoding
this is all you need it is all in 1 unit
From Australia Look up the tests

Click to expand...

Looks interesting. Probably more than I'm interested in spending at this point, but nice to see that there are some new offerings in this area. I'm surprised that their web site says so little about SQ... didn't see it at all until I clicked on the drop-down to select the version.

SQLGuy said:

Are there any of these that don't require Adobe Audition? The cost of Audition alone makes a lot of hardware solutions look pretty appealing.

Click to expand...

Not that I know of. It requires Adobe Audition 3 (but no later) and was available to download from Adobe for free. I can't remember where the link is, but I think it was part of another package.

AA3 was discontinued years ago, but it seems still there:

general page
https://helpx.adobe.com/creative-suite/kb/cs2-product-downloads.html

aa3 specific links from the page above
http://download.adobe.com/pub/adobe/magic/creativesuite/CS2_EOL/AUDT/ADBEAudition_ESD1_WWEFGJIS.exe

http://download.adobe.com/pub/adobe/magic/creativesuite/CS2_EOL/AUDT/ADBEAudition_loopolgy.exe

http://download.adobe.com/pub/adobe/magic/creativesuite/CS2_EOL/AUDT/ADBEAudition_loopolgy.7z

key 1137-1004-8571-6848-7845-8029


3.01 patch

http://www.adobe.com/support/downloads/detail.jsp?ftpID=4237

SQLguy...The first schematic in that document is a direct copy of the Vista SQ decoder kit that I purchased in October of 1974 from Radio Electronics magazine.

The Bushmaster said:

SQLguy...The first schematic in that document is a direct copy of the Vista SQ decoder kit that I purchased in October of 1974 from Radio Electronics magazine.

Click to expand...

Do you still have it? How well did/does it work?

Yup..Still have it and it works quite well. Almost as good as the Sony SQD 2020. It's in my attic right now as I use the SQD 2020 in my main system.

Sounds like an L3 would be worth building, then. Probably even the L2...

OK. Another project for the list.

I started trying to decode the L3 schematic. Basically, what I found is that most of the parts called for are still available, but the high-tolerance capacitors can be difficult to find in reasonable voltages, and for reasonable prices. For instance, .012uF 2% necessitates a 1% 400V device from Mouser, for $4.69 a piece.

This is making me rethink the design a bit, as a lot of the high-tolerance parts are more easily and cheaply available for surface mount. A 2% 50V surface mount .012uF film-type cap is easily available for more like $1 a piece.

The JFET for the blend circuit is also only available in surface mount form factor currently. And going SMD saves all the drilling work when making the PCB... assuming it can all be routed on one side of a board.

Anyway, for through-hole, it looks like the whole thing could be done for about $100 or so in parts. Including the MC1314 and MC1315 ICs, but not including a case, knobs, and jacks.

One more update for now:

Next step is to lay out the circuit and board(s) in some software tools, to get a better idea of what will be needed to build this.

I also added some items to my parts list - case about $12 from China, RCA jacks about $13 from China (for a bunch of them, because I was also thinking that it would be good to incorporate input switching for an external Quad source, and an external 7.1 source, since I only have the one 7.1 input to the pre/pro, and I'd rather not switch cables or add yet another box for switching)... this also adds about $20 worth of relays to handle the switching. Power supply (external 28V 18W) is about $14.

The reason for 28V is so a separate 25V can be supplied for VCC to the MC1314. Per the data sheet, the MC1314's capabilities will be limited at the full extreme of the Dimension setting (20dB front/back separation) if the VCC isn't 25V. The rest of the circuit calls for 20V.

Here's where I am so far... laying out the circuit in Eagle so I can design a board for it.

The one problem I have is that the L3 schematic is pretty difficult to read. I am not 100% sure, for example, about the 16K and 390 Ohm resistors in the network. Anyone have a better copy, or know for sure what these values should be?

Thanks,
Paul

I've had a look at the schematic and I'd agree with your 'decoded' values for the resistors, but very difficult to read. I would go for Surface Mount components as the performance/quality exceeds the through hole devices you can buy now - though not that easy as an amateur to build though, I'd stick to 0805 or 1206/1210 sizes you can see them!