multi-channel receiver with web interface

[Mike O'Dell]

i believe it's the same FPGA or possibly its bigger brother.
the current single pipe in the QS-1R takes about 18% of the FPGA.
how many more will fit is an interesting question. there is
a version of the QS microcode that is running two pipes with
SDRMAX3 host software.

i think the most interesting feature of the WebSDR board is using GigE
as the pipe to the host. I don't know what chip WebSDR is using,
which I'd like to understand. I know the new commercial/military
DSP systems are using full-duplex switched GigE for the same job
and the packet formats are getting standardized quickly for easy
interoperability (for a huge change!). The advantage USB2.0 has
is that a laptop usually only has one Etherspigot and only sometimes
is it GigE. on the other hand, GigE Express Cards are now available,
so maybe that ain't an issue, really. doing USB3.0 will be quite
a bit harder.

if you haven't snooped around the microTCA world, it's the new platform
being used for "high-speed signal processing" by lots of interesting 
people, even civilians doing multi-protocol cellular base stations.
(see Vanu.com - they have a base station that runs CDMA + GSM in the
same white-box PCs) microTCA is essentially CompactPCI-style boards
with an extra set of connectors fully-meshing the card slots plus
another set going to a switch module (optional - the full mesh always
works). they are impedance-controlled differential traces and are
used with GigE, Infiniband, and I think PCIe.

the kit is too expensive for us, but watching the commercial
players foreshadows what we can expect to be sliding down the cost
curves pretty quickly.

the "tinkertoy" approach of the HPSDR project is notionally correct,
but circumstances have overtaken the specific design and the 
partitioning decisions are no longer good choices. but having
some tinkertoys is clearly a good idea - the interesting question
becomes what should the partitioning look like today and for the
next 5 years?  a companion board the size of the QS-1R could easily
run a Linux system with plenty of go-fast for much amusement.
maybe we need an SDR equivalent of the PC-104 form factor!
actually, there is a successor to PC-104 called SUMIT which
does the "embrace and extend" including PCIe lanes. and there
are now PCIe switch chips, too.  maybe thinking along the EPIC/PCIe
and SUMIT/PCIe avenue is an interesting notion. Hmmm - a CPU board
with a 64-bit MIPS/Broadcom part that has a decent Linux port
(the part i'm thinking of is in most of the small NAS crates and
has lots of on-the-chip I/O), etc, etc.  just musing....

	-mo


On 1/26/10 1:24 PM, Tom Azlin N4ZPT wrote:
> Thanks!  Does sound very much like the QS1R where some of the discussion
> is that up to 7 or 8 receivers can run simultaneously in the gate array.
>
> 73, Tom n4zpt
>
> On 1/26/2010 6:16 AM, Mike O'Dell wrote:
>> http://websdr.ewi.utwente.nl:8901/
>>
>> very much like the SQ-1R but with an Ethernet interface.
>>
>>      -mo
>>
>>
>
> _______________________________________________
> Tacos mailing list
> Tacos at amrad.org
> http://www.amrad.org/mailman/listinfo/tacos

-- 
"Of course it's hard!
If it was easy, we'd be buying it from somebody else!"