All:
I'm not going to say much on this, because I want the discussion to flow, and then for John to make recommendations that we can peer reivew re the redesign.

I will say this:
Juan makes an interesting proposal, which will solve much of the EMI issue for weak signal receivers.  I still think we need to reduce, as much as reasonably possible, the noise generated, radiated, and conducted by the CAN-Do! widget under all postulated uses in Eagle.  There are many reasons why this is just a good idea.

73,
Jim
[email protected]

Juan Rivera wrote:

Jim and gang,

 

I have a bit of new information for you regarding the CAN-Do step-down converter:

 

The Maxim 1836 step-down converter is designed to deliver output currents of up to 125 milliamps, but in this circuit it is only supplying 11 mills.  The spec sheet shows that the efficiency and the switching frequency both drop off badly at that low output current.  It’s pretty clear that this IC is not the ideal choice for this very small load.  I can think of several alternatives:

 

1)       Replace the noisy switching downconverter with a linear regulator and remove all traces of generated EMI

2)      Find another more suitable step-down converter and design a new circuit and a new PCB

3)      At least for the 70 cm Receiver, disconnect the CAN-Do step-down converter and power the CAN-Do module from the +5 volt receiver supply. It shouldn’t even notice another 11 mills.

 

Let’s assume for the moment that we end up with a two-compartment enclosure.  What if we just disable the CAN-Do module’s step-down converter and create a small PCB that would attach to the 40-pin header and be the home for the power supplies for whatever was in the other side of the enclosure – in this case the receiver analog circuitry.  The CAN-Do module would get its power from that little board and so would the receiver.  In the case of this receiver we could put all the power supplies on a PCB about the size of a large postage stamp.

 

Don, KD6IRE amplifies on my idea and suggests that the digital compartment be placed off to the side – just wide enough to allow the existing CAN-Do module to fit.  The main compartment would then be an ‘L’ shape, but with access to the front panel.  All RF and IF I/O from the rear compartment would come up the leg on the “L”, with SMA connectors all located on one side of the CAN-Do D connector.

 

Again speaking only from the perspective of the 70 cm receiver, I think this would require the least amount of redesign and result in a pretty clean passband.  The benefits would be:

 

1)       No major modifications to the CAN-Do module required (same PCB and same connector)

2)      No connectors out the side or the top creating wiring headaches

3)      No more 5 kHz radiated or conducted CAN-Do EMI inside the enclosure or propagated back up the DC power lines (5 kHz spurs completely gone)

4)      Moves the 157 kHz receiver switching power supply to the separate compartment in the front on that little daughter board and gets rid of that spur in the passband

5)      Minor modifications to the enclosure – just another internal sheet metal piece

6)      The Receiver PCB form factor could remain the same or it could change to the “L” shape, depending on what works best.

 

This still leaves the issue of flex unresolved and this suggestion might not work for every project but it seems like it might be worth trying.

 

73,

 

Juan

 

P.S.  Maxim 1836 link à http://datasheets.maxim-ic.com/en/ds/MAX1836-MAX1837.pdf