I don't own an RSPDuo and I have no commercial association with SDR Play
... although I do own an RSPdx and an RSP1B.
The RSPDuo does have separate front ends, but the two receivers are fed
from the same LO and they are stable. Which means that you don't
necessarily have a calibrated phase measurement, but you aren't going to
get that anyway unless you drive both antennas with an RF source the
same distance apart to calibrate your setup yourself. Even if you have
"equal" length feedlines they don't necessarily have exactly the same
velocity factor.
In any case, I'm not sure the objective is to actually know from which
direction the interference is coming from ... you just need to be able
to adjust the phase sufficiently to put the notch (or peak) in the
correct direction.
This is one of the demos I was referring to:
https://www.youtube.com/watch?v=B1InftO8IlY
Dave AB7E
On 8/17/2025 1:11 PM, Jim Lux wrote:
I’m not sure the RSPduo actually allows synchronized samples. You can use the same sampling clock, but
internally, the ADCs are fed from separate RF front ends, so there’s the phase ambiguity in the PLL synthesizer.
And, I’m not sure if their firmware does sample for sample paired acquisitions. Certainly not if you
have multiple RSPs (I have a stack of the SDRPlay receivers on my desk at work) - you can tie them all to a common clock,
but the actual acquisition isn’t synced.
Similarly for the plethora of “multichannel RTL-SDR” units out there (there’s a 5 input
version) - they basically use 5 of the RTL-SDR chip sets, and a USB hub, so there’s no synchronized
acquisition (it’s not something the backend sampler of the RTL-SDR does).
The USRP *does* have a way to do this (since they’re supplying devices to the MIMO crowd which
need 4x4 Tx/Rx) - but it’s pretty pricey, in ham terms. (I also have USRPs on my desk at
work - this is something I’m very interested in at work).
The N210 is about $3k, and it does 2 in and 2 out. You *might* be able to put 4 channels in with a pair of 2 Rx direct receiver
cards. I’ve not looked at the details yet. The B210 has the AD9361, and only goes down to 70MHz. I’ve looked
through the user guide for the 9361, and it doesn’t look like there’s a mixer bypass (I did think about
deliberately misprogramming the LO PLL, so there’s no LO, but I’m not sure that would work, or that there
aren’t other high pass blocks that cause problems.
One might also be able to gang up a set of Lime SDRs or ADALM Plutos - but those typically don’t go
down to HF, although they do expose enough digital lines to be able to sync across multiple card (with some
FPGA coding). The Lime mini is $400 (per channel) and the front end only goes down to 10 MHz, although
there might be a “bypass” to allow lower frequencies. The Pluto uses the AD9361.
And really if you’ve got a tuner on the front end, you’d like all the channels
to be fed by the same LO (so that the LO phase is the same for all converters).
On Sun, 17 Aug 2025 11:46:56 -0700, David Gilbert<ab7echo@gmail.com> wrote:
Before he retired, my oldest son used to write software for this sort of
thing. He told me that with eight synchronous antennas he could beam
form in any direction ... up/down and left/right. And of course for
receive they wouldn't have to be full size antennas, and depending upon
the resolution they wouldn't have to be spaced a large percentage of a
wavelength apart.
Units like the RSP Duo from SDRPlay (about $300) are synchronized
sufficiently to allow beam forming with software like SDR Uno, although
with just two antennas you don't get much gain. You can get a lot of
rejection in a very sharp notch that way, though. There are some pretty
impressive YouTube videos out there demonstrating the effect.
Dave
On 8/17/2025 11:11 AM, Jim Lux wrote:
with N antennas, one can theoretically null N-1 sources, so if the noise isn’t sort of
“generic atmospheric noise distributed in the direction of your desired signal” some
sort of multichannel coherent receiver could be a good thing.
And, one doesn’t need those antennas to be particularly “special” or spaced in any particular way - that is, if
you’re doing adaptive cancelling, it’s not like beamforming with a 4 square where some arrangements are better than
others. Obviously, “farther apart” is probably better than close together, but realistically, some smallish loops
oriented in different directions would probably work.
The trick is that there’s not a lot of inexpensive hardware out there that
provides synchronized acquisition of the signals - you can find RTL-SDRs (at the low
price end) and various other SDRs that provide synchronized clock, but they tend to
have separate USB interfaces and the sample streams are not synchronized.
For this application you don’t need huge dynamic range, assuming you have some
front end filtering to knock down things like AM BC stations.
Somewhere along the line, someone is going to finally build a N input RF front end for HF at a “consumer”
price point (e.g. not the $10k for a USRP with 4 inputs). Then, it’s a matter of implementing the combiner
algorithm, and then you can feed the “fixed up” RF into whatever receiver you are comfortable with
(including whatever AGC and filtering you like).
That’s going to change HF receive antenna concepts a lot - because, after all, we typically don’t
need “gain” on receive, what we look for is directivity. (if gain was what we need, then people
would spend some time building lower noise front ends for HF receivers, and they don’t)
On Wed, 13 Aug 2025 12:30:19 -0700, Jim Brown wrote:
On 8/13/2025 4:46 AM, Brian Beezley wrote:
Fading can be a problem on 160. When it is due to polarization rotation
of the incoming signal, a circularly polarized receive antenna can
eliminate it.
Very interesting ideas, Brian! However -- the overwhelming issue for
many of us is local noise. I have two reversible half-wave Beverages,
half-wave on 160M, that are effective as high as 20M, one to EU/VK, the
other to SA/JA. I also have a phased pair of VE3DO loops spaced 5/8-wave
on 160 that are also effective on 80M. All of these antennas are
vertically polarized. There's a noisy home with a solar system in the
direction of EU, another in the direction of SA, and a retreat center
with a large solar system in the direction of JA.
My point is that for most hams, receive noise is the dominant factor in
what we can hear. With the same Beverages and TX antenna, I could work
EU on 160 CW a few nights a year when I moved here in 2006. I haven't
heard EU on CW for five years. I do serious weak signal work on 6M, and
noise from most directions is limiting me by 12 dB or more.
BTW -- beginning with their K3, introduced in 2007, Elecraft has had the
option of a second synced RX, and I've been using it since 2008. I'm
phasing the two VE3DO loops with a DX Eng NCC-1 noise canceller, which
is a very nicely engineered unit. Measured responses are in this pdf.
http://k9yc.com/VE3DO.pdf
73, Jim K9YC
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|