Interesting.
Thanks for bring this up!
Looking at the schematic on page 10 of the pdf pointed to by the reddit post,
I'd guess the first resistor to burn due to DC injected into the TX SMA port would be R14.
Current there is roughly Vtx/(49+56), and power at R14 is I**2 * R = (Vtx/105)**2 * 56 Watts
Assuming R14 is good for 1/10 Watts before it burns out:
0.10 = (Vtx/105)**2 * 56
sqrt(0.10/56) = Vtx/105
Vtx = 105 * sqrt(0.10/56) = 4.43 Volts DC max into the TX SMA.
A thorough analysis of that resistor network might bring the max DC voltage even lower.
On the RX SMA port, first to go is the 100 ohms at R24, placed directly across the port.
0.10 = Vrx**2/100
0.10*100 = Vrx**2
Vrx = sqrt(10) = 3.16
So if DC exceeds 3 volts into either port, a blocking cap should be used
to prevent burning out a resistor in the nanovna.
Given all the blocking caps further back at the SA612A mixers and Si5351 clock generator,
I believe the nanovna should still give accurate results with a DC bias on the TX and/or RX port.
My nanovna has RFI covers over those resistors,
the figure of 0.1 Watts max per resistor assumes 0805 resistors.
If it's 0603 resistors, then it's 0.063 Watts max per resistor,
and you'd best keep DC voltage below sqrt(0.063watts*100ohms) = 2.5 Volts.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Mon, Oct 5, 2020 at 09:07 AM, Carol KP4MD wrote:
A schematic diagram posted a similar thread on Reddit
shows there is a DC current path through resistors on the NanoVNA ports.
You can possibly damage the NanoVNA ports if you apply a DC voltage from the
bias T to them. You need to attach a DC blocking capacitor in series with any
connection to the NanoVNA on which a DC voltage is present.
Carol
