Cambridge Audio 640P

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Cambridge Audio 640P - Update

Well, schematic or no, I took the plunge and installed gold plated sockets and OPA2134 OpAmps in the MM section yesterday.

Because of the type of solder used, getting the NE5532's out was a total pain but, the improvement in sound was more than worth it. Less congestion, greater space and better imaging yet, retains smoothness the 640P is known for.

Next step is to enjoy it for while as I research an input and output cap change - get rid of the electrolytics in the signal path

I think it's pretty much done to be honest. I am considering splitting the power supply off the PCB and putting it in all a full width case.

Maybe replace the regs with discreet units, not sure i need to probe their output noise first.

Has anyone looked underneath the metal enclosure labeled "discrete low noise MC stage"?


  • 640pInside.jpg
    574.3 KB · Views: 727
The copper cover over the "Discrete Input Stage(s)" serves two purposes:
RFI shielding and Temperature Stabilization of the PNP transistors underneath.

The stage itself is simply a Red LED biased CCS (constant current source) driving a
differential amplifier using two sets of three matched PNP epitaxial transistors; in tube-parlance - a long tailed pair. One side of the diff amp is grounded, the other side
receives its signal from the MC or MM cartridge through a coupling capacitor. The
design principle here is pretty simple:

1. Parallel together three transistors, already closely matched by the OEM to further
reduce noise factor.
2. Use a CCS driving the differential amplifier to achieve very high PSRR/CMRR.

There is little you could do to improve what Cambridge has done there... It's an excellent
input stage for the price point. The only way to possible improve this would be to
entirely gut the Discrete Stage and replace it with a similar circuit based on the
Linear Systems LSK389 and cascode current source. The only advantage of resorting
that that option would be to permit direct coupling from the cartridge to the JFET gate
and eliminating the input capacitors and furthermore, that treatment would only really
benefit someone using MC cartridges and not high-level MM that most people have.

You might make marginal improvements of the tail resistors in the existing ckt...
These are R54,R71 and R62, R79. You could replace these with 0.1% MF resistors. The more closely matched the tail resistors - the better the CMRR of that diff pair.

-- Jim
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.