开云体育

Mike,

Very cool!

I'm trying to get dangerous with Kicad as well, it's a serious set of tools.
Being retired now, I no longer have access to the schematic editor and PADS PCB
board layout program that I had used for 30 years.
I've got a bunch of motor memory to retrain.

I believe you went with Jim's schematics from 2001:
? ??
That's a compromise, in that he was building to the arbitrary spec
of using only 2n2222's as the active device.

The Norcal schematics of 2008 made a bunch of improvements, and addressed any
issues that had come up between 2001 and 2008.
For example, looks like a much more serious bit of crystal filtering in the IF
? ??


Driving those mixers with 7dB of energy from an si5351 presents a conundrum.

The Si5351 datasheet claims it can drive 50 ohms, but then AN619 shows that
the CMOS outputs can be programmed to a max of 8ma.
The ap note is correct, the Si5351 does not do a very good job driving a 50 ohm load,
and loading it up like that will add to the crosstalk between Si5351 channels.

Those diode ring mixers want to be 50 ohms at all three ports.
With 8ma into 50 ohms, we get something on the order of 5 dBm of power.
The 7 dBm mixers will work at 5 dBm, but we're better off having a 6 dB resistive pad between the
oscillator and the mixer to absorb any unwanted spurious products coming out of the mixer and keep
them from bouncing back into the mixer to produce even more such junk.
And after adding that pad, we have lost roughly 8 dB of dynamic range off the top.
This is what the uBitx does, and it seems to work well enough.
But could do better with a buffer amp between the Si5351 and the resistive pad.

I agree about the audio transformer, there's a reason nobody sells them anymore.
Many of the audio amp IC's can drive 8 ohms directly, and might be cheaper than that transformer.

For a simple audio amp on a CW rig I'd take a look at the SW40+, includes an audio filter for CW.
If driving that amp from a diode ring mixer, you may need to add another audio gain stage up front.
? ??

The SW40+ uses an SA612 as the second mixer, we could patch that into the 2n2xx as well if so inclined.
We don't really need the dynamic range of a diode ring mixer after stripping off any QRM
with the very narrow IF crystal filter.? The SA612 would not need a buffer between it and the Si5351's output,
just a resistive attenuator to bring it down to between 200 and 300 mv into SA612 pin 6,
which has a roughly 10k input impedance.? The 14 dB of gain that the SA612 provides (vs 6 dB of loss
through a diode ring mixer)means either the audio amp or the IF amp can have 20 dB less gain,
not clear to me exactly how that extra gain should be distributed.

I'd leave the diode ring mixer up front of the IF alone, as that is what will make this better than all the SA612 rigs.

This old forgotten Philips ap note is a gold mine for those trying to understand the innards of the SA612:
? ??
Here's some history that conflicts slightly with that short piece on page one of the ap note:
? ??


Jerry, KE7ER