At 08:45 AM 4/8/2007, Skip K3CC wrote:
>To flexradio Team,
>
>I was a potential buyer on a fixed income. However, I have mixed feelings
>about spending almost $5000 or more for a radio and computer system.
>
>I currently have a n Orion I and Yaesu FT2000 and FT 1000mp mk/V with mods.
>
>Question how does the Flexradio 5000 hamdle contest multi transmitter
>stations ?
I would imagine it will take some practical experience over a year or
so to really answer that question effectively.
1) The versions with the computer in the box potentially will have
less RFI susceptibility, which might reduce those sorts of hassles.
2) multi transmitter contest stations vary all the way from an adhoc
assembly on Field day to permanent installations with filtering and
such developed over years to adapt to the peculiarities of each installation.
3) In a multi Tx environment, I suspect that the rest of the "system"
(antennas, relays, filters, etc.) might have a bigger effect than the
radio itself.
4) A potential advantage of the software nature is that if you had a
stack of SDR1000s, SDR5000s, etc., and they're controlled in a
coordinated fashion, you might be able to be very clever about how
they're tuned to reduce spurs between radios. Example: If Tx A has a
harmonic that's close to where you want RxB to be tuned, rather than
tune RxB to where the desired signal is in the middle of the passband
(leaving the spur from TxA in the passband), you might tune RxB a
ways off (putting the TxA spur out of band) , and just adjust the IF
processing to match that tuning.
{We did something like this on the software radio (Electra) now
orbiting Mars (Mars Reconnaissance Orbiter), when an interfering
source from one of the other instruments had nasty inband spurs on
the UHF comm link from the rovers on the surface}
>I had asked about the cpu system in Dual core processors and was told that
>there was no problem, The reason I had asked was because of the AMD problems
>of timing. They are ok for gaming but I questioned them for this
>application.
That is certainly a huge potential trap.. And one where, to a certain
extent, you are subject to how the low level kernel and driver OS
software interacts with the processor (regardless of supplier). The
high performance computing community has forums filled will all sorts
of empirical stuff learned about various processor/motherboard/kernel
interactions. Mind you, these are people looking to eke out the
absolute best performance/dollar/reliability in cluster computers
with dozens, hundreds, or thousands of processors. However, their
experience goes to show that even among notionally similar
motherboards all using the same processor, there are substantial
performance differences.
This is probably a larger source of variability than other things..
motherboards change on a month to month basis, and have very short
market lifetimes. I suspect that Flex will have their hands full
dealing with this sort of thing, unless they've chosen a mobo with a
very long support tail. And even then, it's quite an ordeal to get an
identical motherboard 2 years later.
>Another problem is grounding and RF from close by transmitters . Have you
>found this to be a problem ???
Hah... *everyone* finds this to be a problem that needs attention,
regardless of radio. Even folks installing racks of $20K commercial
radios have this problem.
The real question is whether the design of the unit allows one to
apply straightforward corrections. The integrated PC will go a long
ways towards helping this, just because it has fewer exernal connections.
FlexLink (or whatever the new bus is called), on the other hand,
might prove to be a real chore to work with. This is especially true
if it's just vanilla I2C (as mentioned by an earlier poster), which
is single ended, multidrop, and has poorly defined electrical
interface properties. I2C was never intended to go anywhere other
than on the same circuit board or within the same chassis
enclosure... it was invented by Philips as a TV remote control
receiver/tuner/audio chip interface protocol for gosh sakes.
I've spent some amount of time over the past couple years trying
different approaches for distributed control of stuff in a high RF
environment (portable HF phased arrays), and, if nothing else,
optical or wireless interfaces are looking more and more attractive,
notwithstanding the higher cost. No common mode voltage problems, no
wires to carry RFI, etc. Cat 5 ethernet actually works fairly well
(good shielding from the twisted pairs, ethernet has galvanic
isolation, etc.) and there's a growing number of inexpensive Ethernet
interface modules (from Lantronix and the like)
Jim, W6RMK
_______________________________________________
FlexRadio mailing list
[email protected]
http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz
Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/
FlexRadio Knowledge Base: http://kb.flex-radio.com/
FlexRadio Homepage: http://www.flex-radio.com/
