I took the time to go through Lou Dorrens latest - revision 5 - published schematic set of his "High Performance 2007 CD-4 Demodulator", starting
here .
(
https://www.quadraphonicquad.com/fo...Demodulator!!!&p=164167&viewfull=1#post164167)
Reading through the thread, the five PDFs and some of Kristian's post (forum user "krkier") I extracted most of the B.O.M. (Bill Of Material):
- U4/U18/U24/U31/U44/U50 - (analogue mux) - Analog Devices DG 5209 or Intersil DG409
- U17/U43 - (14 stage binary counter) - 74HCT4060 or On Semi MC14060 (or equivalent)
- U16/U42 - (ANRS - compandor) - On Semi NE570 available at DigiKey (NE570DGOS-ND) or Youwang Electronics UTC571
- U10/U36 - (FM-PM FSK Demodulator/Tone Decoder) - Exar XR2211A obsolete, but NJM2211 from "New Japan Radio" seems like a drop-in replacement
- U8/U34 - (limiter) - NXP SA606 (with ability to drive the RSSI meters directly
- U55/U56 - (NOR gate) - e.g. two single gates 1G02
A good choice for opamps is the Linear technology
LT1115 which can run from +/-22V rails and is well suited for moving coil and moving magnet cartridges. Most like also strain gauge cartridges.
Also remember this CD-4 demodulator has a"wide bandwidth" input suited for external RIAA or strain gauge amplifiers.
Regulators and transistors are known and availlable from many sources.
Which leaves the analogue filters. Not much has been said, but it seems not to be a DSP so I think these are built around switched capacitor filters, e.g. Linear Technology
LTC1064 or "Active RC" filters like again Linear technology
LT1562. These have the advantage not needing a clock located around 800kHz.
- FDB1/FDB4 - 30kHz BP filter ~40dB out of band attenuation, ref Installment III figure 23.
- FDB2/FDB3/FDB5/FDB6 - 15.667kHz 8th order LP filter
The filter symbols in the revision 5 schematics are 10 pins and I speculate that Lou Dorren made modules with switched capacitor filter ICs and supporting circuitry like buffer opamps. I.e. small PCBs. Ref to post
986 where the picture of Lou Dorren's PCB shows six daugther boards. Speculation again, two are larger, probably they are the BP-filters and the four smaller PCBs the LP-filters.
What I do not like particularly is the use of analogue switches (the muxes) and switched capacitor filters. The switched capacitor filters require good layout of the PCB, there is a possibility for unacceptable clock bleed through, but they also provide for steep filters without a dedicated DSP. Analogue opamp based filters with resistors and capacitors are also a complex task.
What is possible is to renounce on the filter performance and use e.g. 4th order filters. These will probably be better than the old demodulator designs anyway.
One note on the strain gauge (SG) cartridge input: The design differs between left and right channel design as the opamp U5b for the left channel is inverting while the corresponding U29b is non-inverting. This has to do with the biasing of the SG cartridge, as explained by Lou Dorren in the original thread.
This is my analyzis of the design and project so far.
Regds
Rolv-Karsten