nanovna-H v3.4 is out, and I don’t yet see significant problem reports.
When I compare the circuit with v3.3 (which I have), apart from new battery charger IC etc, the changes are in three areas:
- decoupling power to the mixers;
- increasing the drive to the mixers; and
- higher attenuation of input on the rx port.
I have already done #1 (nanovna-H – a ferrite cored test inductor – drilling down on a glitch – the fix), and #2 and #3 are resistor changes, but there may be significant track routing changes – I have not checked.
I doubt #2 and #3 make much difference, I already use it with an external 10dB attenuator on rx port, and would probably need 5dB after those resistor changes, so no real benefit. Efforts to improve performance above 900MHz is a little like trying to make a silk purse out of a sow’s ear.
For the moment, mine will remain a V3.3 PCB save the addition of the power supply decoupling capacitors to the mixers.
The component changes are:
- replace R13 with 0R;
- replace R22,23 with 51R;
- replace R24 with 91R
- replace R25 with 11R;
- remove R29,30,31;
- piggyback 100nf on C40, C42, C44.
R29,30,31 are 0402 parts with other parts nearby, care is needed. The rest of the resistors are 0603 parts and fairly easy access. The caps are 0402.
All resistors are 1%.
There may be associated track changes that are relevant.
Reinforce SMA connectors
One of the weaknesses of the nanovna-H v3.3 is that the SMA connectors are end launch on 1.6mm FR4, and the board flexes if you attempt to tighten the SMA connectors to near 1Nm (the spec). This flexing is likely to damage tracks over time.
So improve to the rigidity, I have soldered a strip of 1.6×6.4mm brass between the connectors. The brass is a tight fit, and the capillary joint is very thin to achieve maximum strength. The tight fit also helps counter torque.
Above, the modified connectors. It is a VERY tricky soldering operation, there is GREAT risk of getting too much solder on things, especially the SMA threads. BE WARNED.
But, it works. When torquing nuts to 1Nm, the springiness has gone.
Quite early in the piece I modified the nanovna-H so that it could be start in bootloader mode without removing the case. The mod calls for replacing R5 with a 1k (1402) and running a short jumper from the T terminal of the jog switch to the un-grounded end of R6.
To use it, hold the jog switch in and turn the nanovna on.
Above a pic of the mod. It is a simple mod, but very fine soldering so it might be within everyone’s capability.
Having done the mod, I don’t think in weeks of playing with the thing and doing literally scores of firmware loads, that I have used it. It is more convenient than the menu option, but I find myself clicking into the menu before I remember it is there!