Hi everybody!
I read a lot of posts about XCVR2450 and I am still confused about its
performance.
I would like to set up a pulse doppler radar application around 2.4 GHz with
USRP2 + XCVR2450
My simple questions are:
Does XCVR2450 work in full duplex mode?
If yes: Does anybody know the isolation between tx and rx (Whole daughter
board: transceiver,amplifier and RF switch) ?
If not: Does anybody know the time that XCVR2450 takes to switch from tx to
rx state ?
Many thanks in advance,
Jorge.
On 10/11/2010 01:34 PM, Jorge M. wrote:
daughter board: transceiver,amplifier and RF switch) ?
If not: Does anybody know the time that XCVR2450 takes to switch from
tx to rx state ?
Many thanks in advance,
Jorge.
The XCVR2450 suports same-frequency full-duplex, which would be perfect
for radar applications. It uses a MAX2829 802.11 world-band
transceiver chip, which has only a single VFO in it, so the RX and TX
sides have to be on the same frequency.
The TX/RX isolation, based on reading the spec sheet for the switch, a
HWS383, gives 29.5dB isolation for the 2.4Ghz range, and about
20.5dB isolation for the 5GHz side.
I suspect that your application will be just fine on this card, using
full-duplex, same-frequency mode.
On Mon, Oct 11, 2010 at 10:34 AM, Jorge M. [email protected]
wrote:
My simple questions are:
Does XCVR2450 work in full duplex mode?
Yes, although with one synthesizer, it will be on the same frequency
for rx and tx
If yes: Does anybody know the isolation between tx and rx (Whole daughter
board: transceiver,amplifier and RF switch) ?
I do not have full characterization for you, but it will be dominated
by the antenna switch (U2), you can find the datasheet here:
http://www.hw.com.tw/PDF/Switch/en/HWS383_v1.pdf
Looks like ~30dB at 2.4GHz
If not: Does anybody know the time that XCVR2450 takes to switch from tx to
rx state ?
It should be approximately the round-trip time for a control packet,
since you will not be tuning to a new frequency. However, because you
can do full-duplex, I doubt you need to switch.
Jason
Jason,
You are right my question about switching between tx and rx only makes
sense
if it is no possible to tx and rx at the same time.
Since you told me it is possible to it at the same frequency (that means
small doppler frequency shifts) the system looks promissing. I was
afraid of
wasting most of the time of my unambiguous radar range waiting to the
system
to switch to reception mode.
Many thanks,
Jorge.
On 10/11/2010 02:24 PM, Jorge M. wrote:
Many thanks,
Jorge.
Given that you can process several MHz of bandwidth around your center
frequency, for doppler
radar, the “same frequency” limitation would seem to not be a
limitation at all.
Unless your closing velocities are very very high 
–
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
On 10/12/2010 06:05 PM, Jason A. wrote:
Ok, digging deeper into the max2829 datasheet, I stand corrected. The
XCVR can not do full-duplex operation. My apologies for the
confusion, our daughterboard app-notes will be updated shortly to
reflect this correction. The RFX2400 is full-duplex capable in the
2.4GHz band.
Jason
It’s true that table 5, page 28 doesn’t show “TXENA” and “RXENA” both
being “1” as a valid setting, but I wonder
if it’ll work anyway. If you look at the block diagram, it looks like
same-frequency full-duplex might actually work.
My recollection is that Matt posted about the XCVR2450 months ago, and
indicated that it would do same-frequency
full duplex.
–
Marcus L.
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
On Mon, Oct 11, 2010 at 11:05 AM, Jason A. [email protected] wrote:
On Mon, Oct 11, 2010 at 10:34 AM, Jorge M. [email protected] wrote:
My simple questions are:
Does XCVR2450 work in full duplex mode?
Yes, although with one synthesizer, it will be on the same frequency
for rx and tx
Ok, digging deeper into the max2829 datasheet, I stand corrected. The
XCVR can not do full-duplex operation. My apologies for the
confusion, our daughterboard app-notes will be updated shortly to
reflect this correction. The RFX2400 is full-duplex capable in the
2.4GHz band.
Jason
On 10/12/2010 03:33 PM, Marcus D. Leech wrote:
NO. The XCVR2450 is NOT full duplex. I have never said anything
different. The XCV2450 is NOT full duplex. Never was. Never will be.
Matt
On 10/12/2010 07:33 PM, Matt E. wrote:
NO. The XCVR2450 is NOT full duplex. I have never said anything
different. The XCV2450 is NOT full duplex. Never was. Never will be.
Matt
Well, then…
Jason and I then were both confused. I thought it was full-duplex but
with the proviso that owing to it having only the one VFO/PLL, the
full-duplex had to be “same frequency”, which would be perfect for
scenarios like radar.
You have too many daughtercards, Matt Makes it hard to keep track
of exactly which does what.
–
Marcus L.
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
Hi,
Now it is clear enough that the XCVR2450 is not full duplex.
The question I asked in my first post in the case it is not full duplex
is
still unknown for me.
How long does it take to the board to switch from transmission mode to
reception mode?
In the data sheet I can read a picture called “TX-RX turnaround
frequency
settling” but for me it is not very clear is the number I am looking
for.
Furthermore it scares me a little bit because if it takes several
microseconds to change state will mean that is going to affect very
seriously the minimum range of my radar detection system. It can take
even
the whole range (if I do not use an amplifier and a good antenna).
Does anybody have any idea?
Many thanks for your answers,
Jorge
–
Marcus L.
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org
Jason A. wrote in post #949262:
On Mon, Oct 11, 2010 at 10:34 AM, Jorge M. [email protected]
wrote:
My simple questions are:
Does XCVR2450 work in full duplex mode?
Yes, although with one synthesizer, it will be on the same frequency
for rx and tx
If yes: Does anybody know the isolation between tx and rx (Whole daughter
board: transceiver,amplifier and RF switch) ?
I do not have full characterization for you, but it will be dominated
by the antenna switch (U2), you can find the datasheet here:
http://www.hw.com.tw/PDF/Switch/en/HWS383_v1.pdf
Looks like ~30dB at 2.4GHz
If not: Does anybody know the time that XCVR2450 takes to switch from tx to
rx state ?
I assume you are implementing pulse doppler radar. You never need full
duplex in this case. Because, you transmit a pulse first and then start
listening to the receiver; NEVER both together.
Further on, the transceiver may not be very great help, it is too simple
for radar. The dynamic range of a pulse dopller radar receiver has to be
high +120dB. This will require logrithic amp or a lot of AGC. Imagine a
pulse was transmitted at time t=0 and start listening. Any pulse echoing
back close to t=0 must jave come from a small distance, thus, quite
large in magnitude and the one coming much later in time must be coming
from a large distance, thus, very weak. Therefore, you will need to
control the gain nearly logrithmically. Here it goes.
- transmit the oulse of duration tau.
- start AGC with least gain.
- progressively increase gain to the MAXGAIN in T seconds, pulse
repition interval.
- simulataneously to AGC activity, keep listening for echoes.
- stop listening at time T
goto 1
hope it kicks off radar-101
fshah
It should be approximately the round-trip time for a control packet,
since you will not be tuning to a new frequency. However, because you
can do full-duplex, I doubt you need to switch.
Jason