Bose 4401 CD-4 Demodulator?

[Asklia]

Well, the BOSE worked before and after enhancing. I cannot say "it helped". On the other hand, solid Ground lines are good practice in audio devices. I can only guess that it helps to enhance the SNR, THD or some other parameters. We must remember, that the boards are designed in the early 70's. Solid audio design knowledge was not really widespread. Other devices of that time are really full of cables. From this point of view, the BOSE is beautiful. And can be enhanced, in contrast to other old devices.

 

[Asklia]

Next step: Soldering for the first volume poti is done. This ALPS four-channel (fake) poti was build from two (fake) ALPS Stereo poti stepped PCB's and the axis of a (real) four-channel ALPS. Why this strange work? Real ALPS are not available with 4 channel including the loudness pin #4 (at least to me).

Now I am waiting for the chinese seller from Aliexpress to deliver the electronic volume board mentioned above.

 

[Asklia]

Using the stepped volume poti works quite well. Although the old poti had a value of 160 kOhm (log), this version with 100k Ohm works as well. A lower value (as long its log) would lower the thermal noise a little bit.

I received meanwhile two kind of boards of electronical volume poti replacements from China. The most promising with an TI/BB PGA 4311 worked, but there was a scratching noise in addition. Independent from volume setting. Not that much, but still audible. Not acceptable.

The other one, shown here, with little remote and even display, worked by far better. No noise, and same rotating element. This fits exactly in the given mounting hole and knob.

The remote board will get an external small box, and connected via the unused remote balance DIN connector on the backside.

 

[Asklia]

Here the link (might expire soon, the shops come and go):
https://www.aliexpress.com/item/100...o.order_list.order_list_main.4.12891802rphmSR
I can recommend it as a replacement poti; stereo or (of course) quad.

 

[Asklia]

Some proceedings.

The original headphone connection was really bad design: Unshielded cables going from the backside of the upper board to the front - and back again. Fully parallel with the unshielded 115/230 Volts power supply cable. And the headphone sockets are really "WW II army style".

I was not able to find a suitable replacement, since the Bose headphone sockets have two "normally closed" and one "normally open" contact. So I cleaned the "black" silver contacts carefully. New shielded audio cabling instead of hard wires. New sockets and plugs with gold plating of course. The old one - surprise - are tin coated as usual.

And I used the back side DIN socket to realize an external connection for the small display and remote control board. Now searching for a stylish case for this small board

 

[Asklia]

I forgot the picture of the external display PCB.

 

[Asklia]

If anybody knows about a source for a nice, small desktop case for the LED, please let me know.

 

[Asklia]

The CD-4 392 ICs on the TPM-19 demodulator board are quite warm (if not hot) during usage. So I decided to add a heatsink on these two ICs. So they live hopefully as long as me

 

[Asklia]

Dont worry: I got a spare decoder board from an old Fisher Receiver via Ebay. Although twice the size, the circuits are all the same. Including the CD4-392 ICs. Even the widespread Grundig CD-4 demods have the same circuits inside.

 

[Doug G.]

The QSI5022 chips in my Technics SH-400 have heat sinks mounted on them, too. As far as I know they are factory original.

Doug

 

[Asklia]

Interesting: They made an own Chipset? Do you have a picture from inside?

 

[par4ken]

Interesting: They made an own Chipset? Do you have a picture from inside?

Lou Doran, "Quadracast Systems". Where responsible for that chip. Lou was working on a new demodulator at the time of his death.

High Performance CD-4 Demodulator

 

[Asklia]

As time goes by...

My Bose projects comes to an end. Below a picture of the old components (the good, the bad and the ugly one's) from my Bose 4401. The old cover has been replaced by a new one, made from a perforated plate. The Bose stays now more cool than before.

Without the three decoders, a Bose 4401 does not use a lot of electrical energy. An unmodified device, but still with an analog SQ decoder inside, has a consumption of 6.5 Watts. But with the two digital signal processors and the hungry CD-4 demodulator, the power consumption has raised to 18.2 Watts.

It was a surprising journey to modify such an old device. In the end, its a very unique quad device: With the best SQ and QS decoders of all times; an internal CD-4 demodulator of the latest generation and with better electrical specs than ever before. By the way: The "ground noise" of about -120 dB measured above was produced by the Denon DVD-A11, not by the modifed Bose.

Some readers might ask, can such a modification be done again? It took a lot of working hours during the last 10 month. And I spend some money for several components, decoders, development, spare device, testing equipment, and mistakes. In the end, it was a budget of more than 4000 Swiss Franks, which is more or less the same in $ or €. I was lucky to have a Tinsmith on the other side of the street, manufacturing the metal parts. I am a professional chemist, so I had no problems getting and working with nitric acid to clean the internal connectors from tin before polishing and chemical gold plating.

Unfortunatly, Lumberg does no longer produce the vertical RCA connectors; neither plated with nickel nor gold. In my students days some 35 years ago, I already electroplated some phono RCA connectors with gold. So there is still a way to get golden RCA sockets.

What is connected to the Bose? A Thorens TD 524, equipped with a wonderful Audio Technica VM750SH including a Shibata style needle. It works perfectly with all of my CD-4 records. The Thorens TD 524 is the "home version" of an EMT "Rundfunk-Plattenspieler" 938.

In future, an Oppo BD player with analog outputs will feed the Bose with music from CDs, SACDs and a few DVD-Audio DVDs. The Oppo BD is able to play digital files as well from my DLNA server, which hosts my digital music collection. A combined FM/DAB+/Internet radio tuner will feed music and news.

A JVC Digital Acoustics Prozessor XP-A 1010 TN can enhance stereo recordings with some room simulation programs. It is connected using the Bose external equalizer loop. Room correction is done by an MiniDSP 88A with Dirac, which follows the Bose. Power amplifiers are from Technics, SE-A3MK2 and A7. So I will never have cold in winter.

If anybody has questions in future about the Bose and its journey, don't hesitate to ask!
Happy Quad

 

[Owen Smith]

You do great work. I hope it sounds as good as it looks.

 

[Asklia]

"it sounds as good as it looks"

To be honest, it works very well. Except the decoders, there is not much electronics inside the audio path: Two classical NE5532 audio opamps at least, followed by a simple transistor stage. So not much chance to alter the audio signal itself. I guess the electronic volume poti will not produce any audible distortions. It could be changed to a classical mechanical poti as well.

All of the three quad decoders produce their own sound and distortions, of course. The two digital SQ and QS decoders from Involve Audio are much less digital than they claim: The signal is divided into three frequency bands, before it is AD converted and processed inside the TI signal processor. I guess the phase shifting is done in an analog way. The TI will do the matrix calculation. The converting AD and DA will produce some distortions as well.

The CD-4 decoder has its limitations as well: FM demodulation, limited frequency bands and heavy filtering stages will produce distortions for sure. Thats why Lou Doran mentioned above tried to build a better CD-4 decoder by enhancing the FM and filter stages. Nevertheless, the JVC TDM-19 decoder board used by Bose was one of the last CD-4 decoder developments 40 years ago. And nobody knows about the losses during the signal processing for the cutting process.

So for new recordings on SACD, the signal will pass quite well inside the Bose. And for the old recordings, we will never have any better technology to separate the encoded signals. Its a little bit sad, that Involve Audio does not show more about their way of decoding the matrix formats. There will be no commercial competitors in these days anymore.

My dream: Dutton Vocalion will get the tapes from A&M with the Rick Wakemans recordings: And give us a four channel SACD with King Arthur

 

[Asklia]

P.S.: I compared the two types of the Involve Audio decoders: The internal, made from the development boards, with the external sold as Surround Master V3. The external SM V3 was connected using the external equalizer connector at my Bose.

I could not hear any difference: Signal to noise seems to be the same, and channel separation from the Quadrafile Test record works in an identical way.

So anybody can have today the same type of decoder technology: Involve SM V3 for SQ and QS; and any of the later CD-4 decoders; built in or external ones. The Bose has the charm, that anything is "build in". It is simply a big four channel switch.

 

[kfbkfb]

https://www.quadraphonicquad.com/forums/threads/working-cd-4-software-demodulator.24900/
I don't know if this would be helpful to you, but there is a new all software CD-4 decoder.


Kirk Bayne

 

[Asklia]

Thank you very much for your hint! I know about this nice piece of software. It has two big disadvantages:

- it runs on an Apple computer; there are no other versions like Windows or Linux
- even worse: It needs a plain "needle rip" WITHOUT going through a phono preamp with RIAA equalisation

Very funny, I must admit. Whosoever has a LP record player with a preamp WITHOUT RIAA equalisation circuit? The signal of a normal pickup (especially MC type) is by far to low to be recorded with accceptable S/N by a sound card or other 96 kHz recording device like a USB recorder interface from TEAC or RME. And 96 kHz is necessary to preserve the 30 kHz CD-4 carrier signal, which goes up to 50 kHz. So no simple CD or DAT Recording.

To my opinion, usage of this really nice piece of software is limited to a very small circle of users. In addition, these funny guys at Spatial Audio do not give ANY hint how to setup such a recording chain. They even try to hide the fact, that their software is Apple-only.

Very sad.

 

[proufo]

Thank you very much for your hint! I know about this nice piece of software. It has two big disadvantages:

- it runs on an Apple computer; there are no other versions like Windows or Linux
- even worse: It needs a plain "needle rip" WITHOUT going through a phono preamp with RIAA equalisation

Very funny, I must admit. Whosoever has a LP record player with a preamp WITHOUT RIAA equalisation circuit? The signal of a normal pickup (especially MC type) is by far to low to be recorded with accceptable S/N by a sound card or other 96 kHz recording device like a USB recorder interface from TEAC or RME. And 96 kHz is necessary to preserve the 30 kHz CD-4 carrier signal, which goes up to 50 kHz. So no simple CD or DAT Recording.

To my opinion, usage of this really nice piece of software is limited to a very small circle of users. In addition, these funny guys at Spatial Audio do not give ANY hint how to setup such a recording chain. They even try to hide the fact, that their software is Apple-only.

Very sad.

Congratulations on the culmination of you beautiful project! Keep us posted on any improvement opportunities you discover in the future. Don't go Involve and believe for a minute that it cannot be improved.

 

[kfbkfb]

The software CD-4 decoder can be run under Windows:
https://www.quadraphonicquad.com/fo...software-demodulator.24900/page-3#post-392162

There's a simple circuit design for building a needledrop amplifier:
https://www.quadraphonicquad.com/fo...software-demodulator.24900/page-2#post-385180

Kirk Bayne