NanoVNA-F uses an ADF4350 already. Good point about register maps being
different, but the MAX and ADF are pin compatible last I checked
toggle quoted message
Show quoted text
On Tue, Mar 21, 2023, 11:22 AM Jim Lux <jimlux@...> wrote:
On 3/21/23 8:50 AM, Eitan Tsur wrote:
Does anyone know of any currently-on-the-market NanoVNA hardware
revisions with MAX2870/MAX2871 PLL's? Seems like they have better
performance at higher frequencies than the ADF4350/4351, etc...
Considering single quantities for a MAX chip are ~$14 USD from Mouser, I'd
be willing to pay the money for a better performing model if software is
available to support the part.
Changing PLL is a major redesign of both hardware and firmware (that is,
it's not a pin for pin drop in). The MAX2870 only goes down to 23.5 MHz
which is fairly important for a lot of users. For that matter the
ADF4350 only goes down to 137.5 MHz
The register map is quite different between the parts.
The NanoVNA uses a Si5351 part which is a clock generator that goes 2.5
kHz to 200 MHz (it will actually go higher), and costs $2.44 in single
quantity.
The design of a VNA using either the MAX or ADF parts would be quite
different.
So it's not the parts cost that's the driver - it's the design cost -
And I'd expect that the parts cost on something covering microwave
frequencies to be substantially higher than the parts going into the
NanoVNA.
For example, you're going to need a better reference oscillator to go to
4 GHz, since the phase noise starts to be a bigger part of the noise
budget, and that 20log(N) bites pretty hard. You'll also need to deal
with layout and substrate issues - at 300 MHz, regular old FR4 works
pretty well, and the small variations don't make much difference (or
will calibrate out). Start getting up to 3-4 GHz and FR4 works, but
it's lossy, and the piece to piece variability gets bigger. Some
calibrates out, some doesn't.
All this can be solved in one way or another, but it's a fair amount of
work.
