Bose 4401 CD-4 Demodulator?

[DuncanS]

Some Impressions from the Alchemy and the results. Digikey delivered the Molex connectors for upside, downside and edge version. Since a 15-pin edge version in gold was not available, I re-used the old see-through plastic frame with the new golden spring plates of another pinout, as can be seen in the picture. Amazing, after half a century, there are still compatible or identical products available! Nevertheless, I did not find the vertical connecting bars. Therefore the "golden chemistry".

Samtec do similar gold-plated connectors https://suddendocs.samtec.com/catal...W*MTY4Mjg4NjE3NC4xLjEuMTY4Mjg4NjQ0MS41OC4wLjA.

 

[Asklia]

Thanks! But as far as I can see in the document, the connectors have a pitch of 0.100" (0.254 mm) instead of 0.156 (0.396 mm) as necessary for the Bose. Old style

 

[THM]

Some Impressions from the Alchemy and the results. Digikey delivered the Molex connectors for upside, downside and edge version. Since a 15-pin edge version in gold was not available, I re-used the old see-through plastic frame with the new golden spring plates of another pinout, as can be seen in the picture. Amazing, after half a century, there are still compatible or identical products available! Nevertheless, I did not find the vertical connecting bars. Therefore the "golden chemistry".

Thanks for sharing.
Molex 4030 series vertical headers are available in gold coating, but unfortunately they're too short for this application : https://fr.farnell.com/w/c/connecte...-4030&plaquage-du-contact=contacts-plaques-or

 

[barfle]

I’ve fought the connector war a few times myself. It’s remarkable how difficult connecting a wire can be sometimes. And doing your own gold plating? Wow. Just wow. Outside my skill set for sure.

 

[Asklia]

Thank you very much for your compliments. I am a professional chemist, so some things are easier for me.
The solution for chemical gold plating is commercially available at Amazon. There is indeed no knowledge necessary. But <cleaning> was the harder work. Getting rid of the tin and polishing the surface needed a lot of finest grinding paper (again from Amazon) as well as ellbow grease

 

[Asklia]

For those who might be interested, some pictures:
- renovated lower board
- the Involve SQ and QS decoders
- the gold plated connectors and new sockets
- power section with new components like rectifier diodes
- new and bigger capacitors in the power section
- new shielding between the upper and lower board like in the first series
- new toroidal transformer and shielding for the 230 Volts power section

 

[furui_suterioo]

For those who might be interested, some pictures:
- renovated lower board
- the Involve SQ and QS decoders
- the gold plated connectors and new sockets
- power section with new components like rectifier diodes
- new and bigger capacitors in the power section
- new shielding between the upper and lower board like in the first series
- new core transformer and shielding for the 230 Volts power section

This is like the quad equivalent of hot rodding! Great job

 

[Asklia]

In minimum, I have one nice effect, although I cannot explain it. The BOSE had always some power-on "plop", although there is this power-on delay circuit in the power supply and the FET for muting. The FET circuit itself suppresses the signal by only 6 db, not more.

Now, after all of this mechanical and electronic modifications, the "plop" disappeared. Maybe, but only maybe, the main capacitor loading takes now more time. And this helps for a more clean power up.

 

[Sonik Wiz]

In minimum, I have one nice effect, although I cannot explain it. The BOSE had always some power-on "plop", although there is this power-on delay circuit in the power supply and the FET for muting. The FET circuit itself suppresses the signal by only 6 db, not more.

Now, after all of this mechanical and electronic modifications, the "plop" disappeared. Maybe, but only maybe, the main capacitor loading takes now more time. And this helps for a more clean power up.

Although I'm not much of a CD-4 fan, I am thoroughly impressed with your skill and passion on this project. Bravo! And thanks for sharing.

 

[MidiMagic]

For those who might be interested, some pictures:
- renovated lower board
- the Involve SQ and QS decoders
- the gold plated connectors and new sockets
- power section with new components like rectifier diodes
- new and bigger capacitors in the power section
- new shielding between the upper and lower board like in the first series
- new toroidal transformer and shielding for the 230 Volts power section

How does it know what kind of record is playing to choose the correct demodulator/decoder?

 

[Asklia]

In the old days, you have to choose the correct mode: CD-4, Stereo, SQ, QS...
Its written on the record sleeve ;-)

By the way:
CD-4 = Demodulator
SQ & QS = Decoder

 

[Asklia]

I did some preliminary measurements with the revised BOSE 4401, using an RME Babyface Pro FS Analog Audio to USB Interface and AudioTester V3.0 software.

1. Noise Floor of the Equipment itself: Analog Input, short circuit to Ground including cables
Comment: Overall Noise is very low; average below -140db except 20-50 Hz

2. BOSE Noise at Tape1 Out with Aux In 0db 100 Hz
Comment: Average Noise above 200 Hz around-120db – excellent for an 50 year old revised preamp.

3. BOSE Noise at Tape1 Out with Aux In 0db 440 Hz
Comment: Here you can see the noise at 50 and 100 Hz of the simple constructed Bose Power Supply: Below 95 db is not bad, but clearly visible and worse than the average -120 db. Needs obviously some further development and tests.

4. BOSE Noise at Tape1 Out with Aux In 0db 1000 Hz
Comment: Same as with 440Hz.

5. BOSE Noise at EQ Out with Aux In 0db 1000 Hz
Comment: EQ Out is nearly identical to Tape1 Out. Playing with Sample Rate.

6. BOSE Noise at Tape1 Out with Phono In from LP Record 0db 1000 Hz
Comment: Preliminary, must be repeated. High THD peaks visible from some overload. Since the signal from the record itself is 0db, scale must be 20db shifted. So average noise at 1 kHz is around 70 db. From 20Hz to 100Hz, you see the ground noise of the record itself: CBS SQ Test LP, Band 1.

 

[proufo]

At the very, very end of you endeavor, won't it be the cherry on top to add a regulated power supply (outboard more likely, although it wouldn´t surprise me if you manage to make it fit inside):

SSR03 Sjöström Super Regulator Power Supply Built

 

[Asklia]

Thank you very much!

I know the page and the products of Mr. Sjöström. For the DC filter and Power on delay, I preferred a combined soft start controller and DC blocker product from NewClassD (NewClassD SoftStart 230V). This filter is build in into the Timer Master Switch, not the Bose.

I have currently several different power supplies here on my desk, some based on LM317 or servo regulated LM317/337 as well as a copy of a STUDER 900 line ultra-low noise regulated linear power (https://www.aliexpress.com/item/100....order_list.order_list_main.29.3b0b1802K9GgjT).

I will test these power supplies in the next weeks and measure the effect on the 50 and 100 Hz noise.

 

[Asklia]

I made a mechanical and electrical "proof-of-concept", feeding the internal Bose power supply circuit directly with +/- 25 Volt DC from a LM317/LM337 based Servo Rectification Filter Power Supply Board. I used my spare Bose board for this test. It fits nicely.

Necessary changes: I unsoldered the two internal capacitors, since they are now obsolete when feeding directly with servo regulated DC 25 Volts instead of 20 Volts AC. In a later step, the rectifying diodes can be unsoldered. Currently, they allow blind connecting the + and - feed from the power supply without thinking about polarity. More changes are not necessary. These power supplies are available in good quality at Ali Express for less than 10 currency units.





P.S.: The red glow in the picture is from two red power supply LEDs, not from some burning and melting components...

 

[proufo]

It fits nicely.

Oh, what a joy.

 

[Asklia]

Some news.

BOSE did not care very much about power distribution. The Ground lines are sometimes very thin; and spread over the board like spaghetti. Same for the +/- 15 Volt lines. I enhanced the main lines with isolated copper wires to make them somewhat more strong. Especially on the Phono Board, the lines are very thin. And the JVC CD-4 board, connected via the Phono board, takes a lot of power!

 

[Asklia]

The "well known weak point" of the BOSE 4401 is now also under construction: The volume poti. I received suitable 7-pin connectors from Digikey and "reduced" them to 4-pin connectors. With these Gold plated connectors, I can exchange the different solutions for testing quite easily.

Beside the different ALPS (Fake) potis and stepped solutions, I have ordered some Texas Instrument PGA4311 based "High End 4-channel Volume Control Panels" from Ali. Inside the ALPS fake poti are - surprise - SMD resistors!

 

[Asklia]

Since the BOSE is a real "4.0" device, a miniDSP DDRC88AB will do the conversion from 4.0 to 4.1 in future. The first Quad preamp with LFE channel and DIRAC Live room correction; between BOSE and Power amp.

 

[Owen Smith]

Some news.

BOSE did not care very much about power distribution. The Ground lines are sometimes very thin; and spread over the board like spaghetti. Same for the +/- 15 Volt lines. I enhanced the main lines with isolated copper wires to make them somewhat more strong. Especially on the Phono Board, the lines are very thin. And the JVC CD-4 board, connected via the Phono board, takes a lot of power!

I've encountered that problem before in digital logic boards in the 1990s. One soldered on extra ground wire made them work reliably. Did it help anything in your case?