Allison's comments are probably worth chasing a little further. ?Here's a summary of her points, and my comments:
1) ?Diode ring injection level should be higher for the 1n4148's
Likely true. ?An easy fix is to move from 1n4148's to a couple BAT54S's in that DBM. ?Perhaps even try a commercial DBM here if you really know what you want. ?And/or adjust the injection level. ?All easily done.
2) ?DBM wants ports at 50 ohms for better overload and IMD performance
Probably true. ?But 200 ohms works well for feeding the crystal filter, and does reduce power consumed. ?Somebody with the W7ZOI 50 ohm amps might patch them in and feed the filter appropriately, see if the improvement is worthwhile. ?Could be that much of this preference for 50 ohms is because it's a nice standard impedance to look at and test with, and not so much driven by performance. ?
3) ?Recommends diplexer between DBM and crystal filter
One of those things that adds parts and complexity, may well reduce spurious received signals. ?But we seem to have identified all the images and birdies thus far, perhaps this doesn't add much to performance. ?Would be worth trying, do an A/B comparison.
4) ?Too much RF gain
Don Cantrell has a nice hack for this, just shut down power to Q1. ?Though some gain at Q1 might prove useful in some cases. ?Alternately, put a pot on the rx signal between the relays. ? ?If we reduce RF gain we need to increase AF gain, and that will result in greater sensitivity to power supply noise. ?
5) ?Move VFO to high side of IF to reduce images.
Yup. ? My Si5351 VFO will be up around 19mhz, just a matter of Raduino code and moving the BFO a couple khz.
A few other things I'm keeping in mind:
6) ?Q13 can blow, either from a nearby transmitter or from too much power dissipation for a Sot23. ?Use a TO92 there? ?Add Raj's back-to-back diodes?
7) ?Q14 emitter resistor can burn out, find a bigger part?
8) ?Might want Raj's fractional ohm source resistor to protect the IRF510. ?My guess is it's not often seen in an RF power amp because on standard RF transistors the source is attached to the tab which gets grounded to the heat sink. ?Not true with the IRF510. ?Maybe a PolySwitch to limit current.
9) ?Add minor hardware bits needed to support CW, maybe standalone PSK31
10) ?Add features for easy testing.
Jerry, KE7ER
