Lou Dorren: A new CD-4 Demodulator!!! [ARCHIVE]

[loudorren]

Hello Folks,

Quadzilla
Mike, Your cartridge will perform better then the AT440, but it is still not the best for optimum CD-4 reproduction. It has only 5 dB of separation at 40 KHz. The Ortofon MC Jubilee is a great CD-4 cartridge, but it is so expensive that you should take out insurance on it!

Quadro-action
Dietrich, Yes the new demodulator will work with any of the good CD-4 cartridges. It's performance with substandard CD-4 pickups is a bit better then the QSI5022, how much better will take some testing. The reason why I decided to include a pickup cartridge, is to give everybody a fighting chance at great CD-4 reproduction. This takes one of the unknowns out of the equation.

Quadfather, I am familiar with the clarion unit, uhg. The Doobie Bros and most every CD-4 record on Electra/Atlantic/Warner Bros were all very good. I helped them in advancing to the Q-540 record compound which has the least stylus static build up of all of the compounds tested. Stylus static build up is just like rubbing you feet on a rug or a cloth on a plastic rod. In the case of the phonograph record, the static attracts particles of dust in the air and settles them on the surface of the disk. As far as the test record is concerned, you may not have gotten the message. I put up on Ebay, 25 brand new CD-4 test LP records that Jim Gabbert and I made back in the 70's. As of this moment there a 10 left. The item number is 330199172179. There is one day left on the auction, but I will re-list.

Winopener, I say this without hesitation. This will be one of, if not the best phono preamplifiers you have every used. The reason for my confidence is because of the ultra low noise, ultra low distortion amplifiers, a high precision R.I.A.A. equalization network and ultra quiet power supplies.

kfbkfb
Kirk, The main channel to sub-channel delays are matched with great precision in the audio Low Pass Filters. This is a very critical match and is extremely difficult to make user adjustable.

To every one that got a test record from Ebay, they are on the way!

Lou Dorren

 

[winopener]

Hello Lou,
that's exactly why i've asked that... for sure we all here are intrested in high cd4 decoding but to use it also as a very high-end phono pre open up a bigger market, the audiophile ones.

 

[ress4278]

Lou, one of the set-up issues with the SH-400 was the CC function. The Techics record was/is the only source for properly setting this up, correct?
Will the new demodulator have a Crosstalk Cancellation option for setting-it up correctly?

 

[loudorren]

Hello Quad people,

Ress4278, Carrier crosstalk cancellation was an idea that I argued against with the Technics engineers. It is not needed and does run afoul the operations of the FM carrier limiters and the AM rejection of the QSI5022. The demodulators do not need this because the capture ratio of the QSI5022 and the new demodulator. Capture ratio is the ability of an FM system to capture the stronger of two signals on the same frequency and is measured as a dB ratio between the two signal strengths. The smaller the ratio the better the capture ratio. The QSI5022 had a capture ratio of 1.4dB, the new demodulator simulation comes in at .6 dB. How this works is that the separation of the cartridge in the sub-channel frequencies for good CD-4 cartridges is between 5 and 25 dB. With the low capture ratios of the QSI5022 and the new demodulator, they effectively reject the opposite groove wall 30KHz carrier which is 5 to 25 dB lower then the on groove wall carrier. The capture ratio of the QSI5022 is probably why some non CD-4 cartridges give some CD-4 performance.
So, no Carrier Cancellation Controls, but there will be 2 real time carrier level meters so you can see the actual carrier level of each channel.

Quadzilla, Does your Tate require phono level or line level? Does it have R.I.A.A eq built in?

Lou Dorren

ps The new ebay listing for the remaining 9 test records is 330202788081

 

[loudorren]

Hello Quadzilla,

Mike, you should be able to take the Left Front and Right Front main outputs and feed them directly into your Tate unit. Simple Y cables should do the trick. This is any demodulator. When the demodulator plays a record that has no carrier, the stereo output Left comes out the Left Front and Left Back outputs, and the stereo output Right comes out the Right Front and Right Back outputs.

Lou Dorren

 

[Cai Campbell]

Rather than rely on cable splitters, it might be more preferable to supply the demodulator/phono amp with two outputs: one four channel and one two channel. Even better would be a switch that allows selection of one or the other rather than have both be "live" simultaneously.

Anyway, just a thought...

 

[callmez]

Rather than rely on cable splitters, it might be more preferable to supply the demodulator/phono amp with two outputs: one four channel and one two channel. Even better would be a switch that allows selection of one or the other rather than have both be "live" simultaneously.

I'm for not adding too many bells and whistles so as to keep the cost down, but I'd really like a separate set of stereo outputs as Cai suggests. Switched or unswitched. And the carrier level meters would be cool, gives us something more to look at than just a box!

Mark Z

 

[aludra]

Hey Lou,

What a wonderful gift you have given us in 2008 with this new design!! I would personally LOVE preamp send jacks along with a return jacks back to the demodulator at line level that would allow us who do 1/2 speed CD-4 transfers a way to re-input the signal back to the demodulator circuits.
Also is the carrier level a dual manual controls or AGC??
Thanks again for your years of quadraphonic brilliance!!
Warmest Regards,
Mike

 

[quadtrade]

Rather than rely on cable splitters, it might be more preferable to supply the demodulator/phono amp with two outputs: one four channel and one two channel. Even better would be a switch that allows selection of one or the other rather than have both be "live" simultaneously.

Anyway, just a thought...

Sweet idea, and would not increase price point much!

 

[The Quadfather]

Dear Lou:
Apparently I missed the auction. If you have one left, I can just send you a check, and you can send it directly. If you like this idea, contact me privately. If not, let me know the item number when you repost. Oh, and for some of the guys here that like to fool with computers, it might be an interesting idea to offer a output-input loop between the preamp and the demod itself. If not in the base price, maybe as an option. It also would be cheap to do.

The Quadfather

 

[loudorren]

Hello to all,
Since it does not affect the design except for two holes and a couple of RCA jacks, I will include a direct output for left and right at line level. This will also give any that want a great phono preamplifier. You all contributed to this design addition, so I will call it the QQ Forum Direct Stereo Output! Thanks Cai, Tad,!!

Alaudra,
Mike, No carrier level controls, no AGC. All unnecessary! The problem with your other request is the R.I.A.A. eq. At half speed with the standard curve, severe overload at the low frequencies would occur. Adding switchable R.I.A.A. eq curves creates a mess with the preamplifier. Sorry.

Quadfather,

The new ebay listing for the remaining 9 test records is 330202788081.

Lou Dorren

 

[kfbkfb]

Regarding the Base band/Main channel (contains
LF+LB or RF+RB) and the Carrier/Sub-channel
(contains LF-LB or RF-RB) delay issue:

I'm getting my information about the delay from
the paper (48th AES 1974-05-07) entitled:
"The CD-4 Mark-II Modulation System"
in the section:
"Delay Relationship Between Carrier and
Base Band"

A High End JVC CD-4 Demodulator included a delay
adjustment, apparently (CD-4 approved) phono
cartridges differ in the amount of delay they
have in the 20kHz to 45kHz frequency range.
A special CD-4 test record had to be used to
optimize the delay adjustment.

Kirk Bayne

 

[aludra]

Hey Lou,
Thank you for your response...

"Mike, No carrier level controls, no AGC. All unnecessary! The problem with your other request is the R.I.A.A. eq. At half speed with the standard curve, severe overload at the low frequencies would occur. Adding switchable R.I.A.A. eq curves creates a mess with the preamplifier. Sorry"

I use the computer to double the speed of the 1/2 speed recording back into demodulator circuits..1/2 speed recorded flat without RIAA eq....software LO eq RIAA curve applied before return to demodulator...I have had excellent results using this technique...and have the added advantage of click removal before return which eliminates ANRS pumping on clicks and mute circuits false triggering...

Also how do you compensate for inner groove carrier level drop??...hence channel separation decreases??? I have always found in my CD-4 transfers if I gain ride the carriers as the record plays (Very easy with the SH-400) I can maintain the full channel separation till the end of the record although the noise increases somewhat...

And you are very correct about the CCC circuit in the SH-400...useless and I have removed that daughterboard in my SH-400 and bypassed it completely.
Big improvement!!!

I hope you will consider this simple loop for us 1/2 speeders... sure would LOVE to take advantage of your super preamp...at the minimum at least a line input to the demodulator. Thank you for your genius!!
Warmest Regards,
Mike

 

[aludra]

I also forgot to add I have rebuilt my SH-400 and inserted my own line inputs past the preamp stage to avoid the mediocre preamp
and LO RIAA eq stage....But would sure love to get an advanced 2008 demodulator for my collection...Count me 1st in line with my checkbook!!!

 

[loudorren]

Hello again,

It is nice to have a computer dedicated to the web close to my lab bench!

kfbkfb, Ah yes, the infamous JVC AES paper. Written when CD-4 was still a lab curiosity. About a year before it was presented. At the time of the paper, JVC was using a Foster/Seeley discriminator for the FM detector. No Limiting!! They wanted RCA to adopt the system but it suffered from several problems. I was asked to look at the problems and I developed the CD-4 FM Phase Lock Loop receiving system. If working correctly,the small diameter variation in carrier level will be taken care of by the limiter. Delay variations in early CD-4 pickups was pretty bad. This was because they were just figuring out how to make them. With the advent of the Shibata low tracking mass stylus, higher compliance cantilevers, and precision generator coil matching, hitting the CD-4 pickup specification became quite easy for the cartridge manufactures. The cartridges of today that are CD-4 compliant are an order of magnitude better then the 1970's. Main channel to Sub-channel cartridge delay is very accurate.

Aludra, The preamp in the SH400 is the QSI5022. By the way if you put the cartridge select in the semiconductor position and either capacitively couple the cartridge inputs or remove the DC bias from the input you will have a great flat preamp.

Lou Dorren

 

[kfbkfb]

CD-4 *Automatic* Channel Separation:

The product information about the Technics SA-7300X receiver
states that "Separation and carrier levels are automatically adjusted
to each cartridge characteristics, ensuring a very carefree operation."

How did Technics do CD-4 automatic channel separation and could
the method be used in your new CD-4 Demodulator design?

Kirk Bayne

 

[The Quadfather]

Dear Lou:
Can we avoid surface mount components, especially the capacitors which have a short lifespan? Also, multilayer boards that are impossible to trace and repair? This stuff is used now-a-days even when it is not necessary, and it's a nightmare to work on it. Panasonic had a major problem with surface mount capacitors in their broadcast DVC Pro gear. Also, some of their early DVD players had the problems with what we came to call "capacitoritis". At least the quad era gear is easy to work on, which is good, because I've had to work on my stuff many times over the years. I hate to keep going back to the well on this, but we do need gear that is easily repaired. And we need the schematics and service info. It was a long time before we got prints on our Tate Audionics units which were regarded "Top Secret" for years. Still haven't seen them on the Fosgate units. It was all so unnecessary, because all the secret stuff was in the chips. But I guess if the industry has left the old construction methods too far behind, and surface mount is unavoidable, at least stick to single or double side board only with components on one side only and all conductors on the surface of the board, not inside, and all electrolytic capacitors lead mounted (better quality capacitors). Also, screw Rohs compliance. Use real leaded solder that does not require special super hot soldering gear, and doesn't grow conductive "fingers" that short out the gear over time. We don't have to worry about lead in the landfill, because this gear will never be thrown away! The danger of lead contamination is way over rated anyway, like most "dangers" we are warned about on products we buy. I guess that's all from me. Thank you for the opportunity to input on this project.

The Quadfather

 

[krkier]

Also, screw Rohs compliance. Use real leaded solder that does not require special super hot soldering gear, and doesn't grow conductive "fingers" that short out the gear over time. We don't have to worry about lead in the landfill, because this gear will never be thrown away! The danger of lead contamination is way over rated anyway, like most "dangers" we are warned about on products we buy.

If Lou likes to sell his decoder comercially, then he has to use RoHS leadfree components and solder!

Nowadays it is much harder to find components with lead than without. Nearly all manufacturers of components have changed their program to leadfree, only offering "green" products. It would be difficult for Lou to get these leaded components since he sure does not intent to built millions of his decoder.

And the change in soldering temperature is not that much, from 230°C to 260°C. Every professional soldering iron can be regulated up to 400°C or beyond.

The whisker problem (your "conductive fingers") is something that the industry is aware of. The problem is the surface of tin-plated materials, and their compounds. Think IC-pins, or component pins in general. Also printed cirquit boards with galvanic tin covering of solder pads, pins, and areas. But the industry has already methods to prevent this phenomenon.

-Kristian

 

[loudorren]

Hello Folks,

It's soap box time!!! As one in this industry for more years that I wish to quote, these are my opinions. You all know the saying "that opinions are like ..." Anyway, as to the construction of the new demodulator. I, unfortunately, have no control over the semiconductor manufactures packaging. 94 % of all new and even some legacy chips are in surface mount. I'll try to avoid these where ever possible, but it is a losing battle. Having said that, RoHS components are just fine so I don't mind using them. Lead free solder is another mater. It represents to the current industry the biggest failure mode of the "green" changes. It crystallizes, making diodes where diodes are not supposed to be. Parts do not re-flow well and tend to fall off in vibration tests. This industry is far from having solutions to all of the problems. Take Swatch for instance. They have gotten a waver to use leaded solder in there product because normal wristwatch abuse was causing massive product failure. Off the soap Box.

I am not going into large scale production on the demodulator. I intend to have built what this "club" of CD-4 fans needs. How many will be determined by all of you!

Double sided G-10 or FR4 printed circuit board with plated through holes, solder mask, and legends for easy servicing and components on only one side is the order of the day! Axial lead resistors, radial lead capacitors, and DIP chips with sockets where possible. I would not do it this way for large scale production, but for servicing purposes (I hope the users will never have to), I think this is the only way to go. I will also make available a package of critical replacement parts so if one needs to fix something you can. Schematics, definitely, plus a users and a service manual.

kfbkfb
Kirk, Automatic separation was a way of saying that they chose a fixed separation because everything would work OK, but not optimum. There is no good way to make a long term automated separation system, with out having to use wide dynamic range gain cell circuitry. The problem here is that the gain cell gets very expensive if you want ultra low distortion. You must have ultra low distortion to have great separation. Then you must have some way to store the control value for the gain cell, other wise you must set it up with a test record every time you turn the unit off and on. Having said all that, I like to adjust the separation controls manually, it gives me the sense of power (something I can do better than the machine)!

All for now,

Lou Dorren

 

[The Quadfather]

Hey Lou:
Ok, A parts kit is an excellent idea and I was going to suggest that. You beat me to it. I have a technical idea you might want to try. I don't know if it would work well, but you might want to test the idea. Instead of brick wall filtering the main audio, why couldn't you bandpass filter the carrier and then feed the carrier back into the main audio 180 degrees out of phase in order to cancel the carrier in the main signal. You would have to maintain cycle to cycle phase relationship to make sure that no modulation products got through the cancellation process. The main carrier could then be presented to the matrix with no bandpass filtering. Is it worth a shot?

The Quadfather