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

[loudorren]

Hello Quadrophoners,

Kristian,

The bypass caps around the regulators on the ANRS pages were accidentally ommited. Thanks for observation. The final number of OpAmp bypass caps will be determined shortly by test. I will update the schematics then. By the way, the schematics that come with the units will be fully annotated.

Good guess on the ANRS chip, unfortunately incorrect. The CDC timer chip is a 74HCT4060 for better edge times. There is no 4000 or DG analog switches. Distortions on them are just not low enough. Unused inputs on these analog switches do not need to be terminated. A resistor is missing on the cart select pin that is diode isolated from the bias control. I will update that. Pull up and down level are supplied by the binary control switches that will be used so there will be no floating logic.

Plus and Minus 20 Volts DC is great! Filters are custom programmed. 565s are still made by 4 manufactures but these PLLs are not 565s.

To get the best sum and difference matrixing, I am using a .1% matching tolerance resistor network. There are a fair amount of voltage regulators in this system. They serve the purpose of reducing and isolating power suppy line noise.

Anyway, Kristian, Nice evaluation and thanks for pointing out schematic omissions. Drawing things this complex always leads to errors and I am using a completely new schematic capture program which is the best I have found, but I am still getting use to it.

Lou Dorren

 

[The Quadfather]

Hey Lou:
OK, I'm a little confused. In the preamp section, the complementary transistors at 1C and 2C appear to have their output connected to the power bus. If this was a gain riding devisement I could see it connected that way to maybe one or two stages, but it seems to be connected to the general busses. The transistors at 1E and 2E are connected likewise. Also, the input of the IC's at 1C and1D are connected to the power busses. So what's the deal?

The Quadfather

 

[The Quadfather]

OK, Looking at it a little longer, this appears to be some sort of hum bucker, regulator arrangement. If so, why include the complementary transistors? If the right two transistors failed at the same time, such as in a power surge, you could wind up with reverse polarity voltage on the op amps, burning up the whole thing. But then, we run that danger with direct coupled power amplifiers, but we like them because they are very clean. I guess I look at things from a repair tech's viewpoint.

The Quadfather

 

[quady-de]

Hello Lou,

please look at your private messages

thanks
Norman

 

[krkier]

OK, Looking at it a little longer, this appears to be some sort of hum bucker, regulator arrangement. If so, why include the complementary transistors? If the right two transistors failed at the same time, such as in a power surge, you could wind up with reverse polarity voltage on the op amps, burning up the whole thing. But then, we run that danger with direct coupled power amplifiers, but we like them because they are very clean. I guess I look at things from a repair tech's viewpoint.

The Quadfather

Generating supply voltages from the output of an OPA is probably too weak, since they have an output resistance from 100 Ohm upwards, and you have to source a bunch of components in this section. Not to speak from supply current for all these components. That's why those complimentary stages were included. - At least that's my guess.

-Kristian

 

[loudorren]

Hello to All QQs,

Quadfather,

One of the biggest problems with phono preamplifiers is the susceptibility to low frequency noise. This is mainly because of the RIAA low frequency emphasis which makes the situation more difficult. Even though most of today’s opamps have great power line rejection (the ability to reject noise on the power lines), this rejection assumes very low power supply impedance. Most regulators are pretty good at this, but when you are trying to amplify signals in the hundreds of microvolts and you have RIAA EQ, low frequency noise can leak in.

The two networks using the complimentary transistors are the actual power supply busses to the preamplifier op amps. As Kristian pointed out, the amplifiers driving the transistors only have a maximum output current of 40 milliamps, while the current required is 175 milliamps. The reason for the complimentary transistors instead of one transistor and one resistor is the maintenance of extremely low constant supply impedance across the entire supply range.

The failure mode you speak of is highly unlikely. First there are a set of high voltage regulators which have excellent input spike rejection and for overkill the complimentary pairs are 350 volt .5 amp transistors which should survive anything. If everything is as I planned the New QQ CD-4 demodulator repair tech should be as lonely as the Maytag repair person.

Lou Dorren

 

[The Quadfather]

Yeah, I know it's unlikely, but as a repair tech, you better believe I have seen just about everything! But I do recognize that absolute hum reduction is indeed a worthwhile goal. And if that's the best way to do it, then so be it. Most of my experience in the audio field is with older gear, and while I have seen "pass" transistors used primarily for hum reduction, they were usually not in complementary stages. But if it works, do it! I was more curious than anything.

The Quadfather

 

[loudorren]

Hello QQ CD-4ers,

Quadfather,

The circuit you write of is a capacitor multiplier. It uses the gain of the transistor to make a small capacitor, a large one. There still is a low level ripple and noise component that can get through since the cap multiplier is an open loop system. The system I am using in the new demodulator can bring the supply line noise down close to the theoretical noise limit (Johnson/Nyquist thermal noise floor of -174dBm in a 1 Hz bandwidth).

Lou Dorren

 

[The Quadfather]

OK, thanks for the information. I learn something new every now and then.
The Quadfather

 

[The Quadfather]

Hey Lou:
As the end of NTSC television broadcasting occurs, this thought occurs to me. hundreds of TV stations have Orban (and other brands) BTSC stereo generators in their transmitter sites that will no longer be needed. These devices do something very similar to CD-4 encoding. The main difference is that there are only two channels per unit. It would be required to have two of these units to make a complete modulator. The noise reduction is DBX rather than ANRS, but if it was ANRS, it would have to be reworked for half speed mastering anyway. I have secured the required two Orbans for myself. The only thing is that if I was to make something useful with them, I would have no way to cut the record. Others who have connections to TV stations might want to experiment with this idea also. These units should be in good condition having been in use constantly since the dawn of stereo NTSC television. Do you think that these units might make an effective CD-4 modulation system?

The Quadfather

 

[loudorren]

Hello QQs,

Quadfather,

Unfortunately, the only similarities between NTSC TV stereo and CD-4 is the input sum and difference matrix. As you know the CD-4 sub carrier system is FM-PM-FM modulation. NTSC TV is Pulse Amplitude Modulation (PAM) which is a double sideband suppressed carrier Amplitude Modulation system. The audio processing on the NTSC subchannel is single band limited DBX.

It would be easier to start from scratch.

Happy Holidays to every one. Things are moving along with the new demodulator. Urban eval is getting close.

Lou Dorren

 

[quady-de]

Hello Lou,

please read and anwser your private messages.

thanks
quady-de

 

[The Quadfather]

Maybe so, but it has a nice case and metering and pull out cards. It seems like it might be fairly easy to pull out the card responsible for the subcarrier generation and replace it with a homebrew card that would provide the necessary modulation type and NR and be set up for half speed operation. But I'll know more when I get the thing offline.

But that's what we really need, a new demodulator is nice, and I'll buy one when it's built, but we really need someone to build a modulator, and start cutting records. But we have always had methods of recording quad, but no access to big name music that we all want to hear in quad, or even just well performed music by lesser known acts. Now we have all these new multichannel systems and no one interested in serving this niche market. If our players break, the digital discs we bought will be useless. They won't be building any more DVD-A/SACD/DTS universal players. The new blu ray players won't be able to play these discs, except DTS. But as long as there are cartridge manufacturers that will still make microline stylii, we can play CD-4 records. The demodulators are simple and easy to repair, and quality turntables can still be bought new. CD-4 is the standard best suited to carry on the torch for quad. If we can just get someone to make the records.

 

[ndiamone]

CD-4 is the standard best suited to carry on the torch for quad. If we can just get someone to make the records.

I asked that question on the Lathe Trollers awhile back about the best `new' medium upon which to engrave CD-4 so it don't degrade, i.e. DMM, and all the guys said the mechanical act of the diamond DMM cutting stylus etching into the copper blank creates a chatter at around 30 KHz when cutting at the normal 33 RPM, which upon playback would be right in the middle of the CD-4 carrier band.

Only on a a Neumann. Nobody said anything about an Ortofon, other than if you were going to try it, to get one.

Guys also said if you could get the CD-4 Ortofon cutterhead and put a Japanese DMM diamond cutter stylus in it, then since the Ortofon cutting angle is supposed to be different from a Neumann, maybe you wouldn't have the mechanical chatter at all, and then whoever was cutting CD-4/DMM could go back to getting other audiophile work by cutting OMR's or Nautilus or CBS Half-Speeds, not to mention mastering a new generation of all the 45-RPM SuperSonic double. triple and quadruple audiophile disc sets that carry the same information as a single LP.

However, none of the mastering engineers apparently was able to tell on an Ortofon, once the mastering speed was dropped to 16 whether the mechanical chatter would still be there on a half-speed CD-4 DMM and still be in the 30 KHz range, or whether it would leave the CD-4 carrier alone and get moved up to 60 KHz, supposedly well above even the 50KHz top end of the CD-4.

The best setup (for CD-4) was the Ortofon DSS-731 cutter and matching GO-741 200 watt cutting amps.

The 731 cutter was specifically designed by Ortofon to cut CD-4 at half speed.
It had the widest frequency response of any audio disc recorder ever designed and could easily cut up to 30kHz in real time.

But that performance still wasn't good enough for real time cutting of CD-4 which required response up to 45kHz with minimal phase shift between the the two channels, so we used the 731 at half speed with wonderful results.

The stylus and collet assembly was very small compared with the clunkier Neumann mounts which went a long way toward providing the wonderful high frequency performance, but changing the stylus was much more of a challenge.
Anyway, there's much more to relate about that experience if anyone is interested in it.

The other issue was the shallow 40-60um size grooves most cutting houses used in the copper blank for
A) the chip to be able to be vacuumed up during cutting and
B) to maintain the low noise floor DMM is designed for.

I've seen grooves in a DMM over 100um though, which are possible because the copper coating is upwards of 100um thick.
Guys say they can cut up to a 200um size groove though, but how you'd suck the chip up I dunno being that fat.

Nobody can pick me up being that fat.

If you could suck the chip up, then you'd get a nice deep groove, the Ortofon cutter would make it nice and quiet,
and it would give both the dynamic range of a normal LP and the more stable mastering for the CD-4 carrier you're looking for.

And I've also seen VAST differences between shallow and deep cuts for CD-4 on lacquer as well.
I have a Doors Greatest Hits on a Butterfly label that only takes up little more than half the disc,
and then I have a Dr. Teleny album that almost doesn't fit onto the record, the cut is so deep and the pitch is so wide by comparison.
Suffice it to say the CD-4 tone is CONSIDEABLY more solid and the sound CONSIDERABLY better on Dr. Teleny vs. my Doors.

Steven Berson, one of the other original DMM developers could probably tell you: http://www.totalsonicmastering.com/

Greg Bogantz says maybe he'll let the world's last still-functioning RCA Quadulator out for testing or redesigning off of for DMM CD-4.

So since there's supposedly a working Quadulator, Ortofon DMM lathes, heads and amps and Ortofon CD-4 heads, maybe some enterprising engineer can put all that together with a revamped and redesigned Dorrenized Quadulator, and get to cutting some test blanks. Then press up some test pressings in 180g grainless dye blue or red vinyl at RTI where a lot of audiophile records get pressed at, and get `em over to Lou Dorren for evaluation.

I call dibs on leftover test pressings and the leftover DMM CD-4 blank to frame `em once Lou is through evaluating `em.

Guys?

 

[The Quadfather]

That would b e really cool. But how do you get the master tapes? And of what bands? Unknowns or well knowns?

The Quadfather

 

[ndiamone]

Do it the same way everybody else does, a'la the multi-LP 45-RPM Super-Sonic or Half-Speed Mastered series:
Put the product out as a DMM Half-Speed Mastered (which it will have to be).
Then you'll have 2 selling points. CD-4 will be a bonus.
Show the cubicle warriors it can make them a little money.
Pay them their little bitty product licensing fee they want afterward.

Or, go about it the same way everybody who puts 24.99 for 40 Hits on TV.
Pay the ASCAP and BMI mechanical rights fees.
Go and do it.

 

[Quadro-Action]

Because the last message from Lou is weeks ago, I will take the interim period till his next message to show something about the correct justage of the pick up. A demodulator - new or old - will not give allone the best results of the wanted high channel separation. It is always a teamwork of both, which means demodulator and pick up. So we can already now have a working with our pick up(s) to get this best results from this part and together with the already own demodulators of (for example) Technics SH-400, JVC 10 (S) or JVC 50 or whenever with the annouced new.

Most of us knows the adjust procedure to eliminate the error angle. Therefore is needed the wellknown adjust stencil. By a correct position of the pick up body the tip of the needle will have the two positions on the record side of zero angle. Before and afterwards of this two points there are again error angels, which are smallest. Only tangential working tone arms may have always smaller or less error angels. But there are further on 2 other directions, which must have also a correct justage. Because you will need therefore a steresocopic microscope, it will be a lot more expensive as the stencils before. New it may cost like a new demodulator, but by ebay they are a lot more reasonable. Only with this set you can see the tip and his edges absolute clear.

Impotant is the look to the front of the pick up to see, how the needle is mounted - hopefully correct strictly vertical to the groove. If not, one must twist the body of the pick up (sometimes possible also the head-shell) a little from left to right or contrary till the needle shows a vertical position. Further important is the look to the diamond tip from above. Then one can see, if the small cutting edges are absolute horizontal, so that the small sides of the needle will tracking absolute straight accross both sides of the groove. If not, you must twist the pick up body a little horizontal in the head-shell from left or right, till you can see by looking through the microscope, that the diamond edges are in the correct position and the informations of both sides of the groove are scanning in exactly the same time.

You may wonder sometimes, when you are looking through such a microscope. There are often divergences in reality from the theoretic best position of the needle. Now I have hope, that I could give you for the meantime some diversion. Or there now to read other pointers for the exactly justage of the pick up or askings therefore.

 

[loudorren]

To all QuadrphonicQuad Members,

This is a very hard posting for me to make. As you know one of our members, Cai Cambell recently passed away. I knew Cai for quite a while, though we never physically met. He was a true Quadraphile and he knew the field quite well. Besides all that he was nice guy. He was to young when he passed and as a friend I will miss our phone calls.

My condolences go out to his family.

Keep those four channels going Cai, We will miss you.

Lou Dorren

 

[The Quadfather]

It's sad to hear about Cai. We're all getting older, the original quaddies. R.I.P.
Cai.

The Quadfather

 

[The Quadfather]

Hey Lou:
Any more news about the demodulator?
The Quadfather