[Premium-Rx] Worlds First DSP-Receiver

[GandalfG8 at aol.com]

In a message dated 21/01/2008 12:34:20 GMT Standard Time,  
Rev-hwinter at msmisp.com writes:
Connoisseurs" of "high-class" Radio receivers,  to all of you

My name is Harry Winter and I am a retired  "Member-of-Staff" with Bell Labs. 
(Now Lucent Technology)  I am at the  present working on "upgrading" my 
CEI-type 232-2 receiver. --- (marketed first  in 1969, shortly before CEI became 
part of Watkins-Johnson)

This receiver  was called a "TUNABLE FILTER" by the designers and has a very 
low third IF of 15  or 25 KHz. No signal demodulation circuits or narrow 
band-pass filters are  provided. Final filtering and demodulation was obviously to 
be done by a very  powerful mini computer or even a Control-data or IBM 
Mainframe; --- in 1969  there was no PC, that came only in the late 70's and they 
were running only at a  few megahertz. Maybe even the Mainframes were not fast 
enough at that time and a  tape record of the 15 or 25 KHz had to be used 
instead.

--------------------------------
It seems quite a leap to infer that  demodulation was "obviously" to be done 
by a powerful mini computer etc, do you  have any evidence of this or is it 
just an  assumption?
-------------------------------

Never-the less, ---- THIS  WAS THE WORLDS FIRST DSP-RADIO! -----
-----------------------------

No, this was a "tunable filter", or to put it another way, a  conventional 
radio without a demodulator.
I agree adding DSP demodulation is  a good idea but whatever you choose to do 
with it doesn't alter what it was to  start with.
WJ also provided translated outputs to 15 KHz on some of their  other units 
and I've been thinking about adding a DSP demodulator to one of mine  but would 
never in a million years try to convince myself that was what it was  
designed for..
---------------------------- 

My upgrade consists of  replacing the IBM-mainframe with a DSP-chip from 
Analog Devices, the $12  ADSP-2184N, a TLV320AIC23 quad CODEC chip from TI and a 4 
Mbit Flash chip. All  this is running at 80 to 160 MIPS with lots of 
real-time to spare. This complete  circuit, called the DSPx, is available from the 
ARRL for about $100 and fits  into a tiny cast aluminum box of 100mm by 50 mm by 
25mm (Hammond Manufacturing,  Series 1590G) and has a power consumption of 250 
mille Watts.--- The software is  free and includes super-performance filters 
as well as USB, LSB, narrow FM and  CW and de-noiser and auto Notch.

(See:    http://www.kk7p.com/dspx.html  ) Also, take a look at how small the 
DSP  mainframe actually is:

http://www.msmisp.com/futuretest/CEI-232-2-C.jpg 

It's the black-box  beside the empty connector (Which is missing the AGC 
circuit board that needed  some very minor modification.)The front panel of the 
CEI-232 is only 200mm wide  and 80mm high, just large enough to have an IBM 
mainframe  included.



That's it Paul, just upgrading a 1960 Gull wing  Mercedes or E-Jaguar. 

---- Do I qualify as connoisseur? ----
 
-----------------------------------
Not if you need to find excuses to  redefine a CEI mechanically tuned RF/IF 
unit, however nicely made, as some form  of premium RX, which, at least in 
terms of the definition used  here, it certainly wasn't.
BTW....
I'd be interested to see your plans for fitting  the cold fusion power plant 
into the  Jag.
----------------------------------


PS, I would be very  interested to meet other owners of this unique "World 
first SD Receiver". (It's  like the first swept wing Jet fighter, the ME262, --- 
for which now 5 newly  build and upgraded versions are flying ---- cost is 
only $2 million each. 

--------------------------------
This is beginning to sound much  more like an Epay spiel than anything even 
remotely related to premium  receivers.
Whilst my first reaction was "what a load of old b******s", I  can't decide 
whether or not we're supposed to take any of it seriously or  whether April 1st 
just comes earlier for some than  others?
-----------------------------

I am also looking for a copy of  the schematic for the CEI-232, it would 
help, but I can do without. Once I have  the DSPx module fully integrated into the 
architecture of the CEI-2232, I will  certainly share the work and circuits 
with the other owners of this unique  receiver. Beginning in the 1990s all 
conventional analog receivers became  obsolete --- just like the area of the 
"Steam-engines" for trains.

---------------------------------------
Presumably by "analog"  receivers you mean those without DSP rather than, as 
more normally accepted,  those with analogue tuning.
Even so, overnight obsolence might seem like  another rather extreme leap of 
faith to most of  us.
---------------------------------------- 

As you might have  heard though, the front-end of radio-receivers will stay 
analog" for many more  decades, because of technology limitations. However, the 
DSP-part is influencing  the analog part of the "architecture" of 
dsp-receiver designs. ---- How does the  1960 design of the CEI-232 rate for these new 
requirements? ---- SUPRISINGLY  WELL! --- It is as if the engineers of CEI have 
been clairvoyant. One  requirement is a very high first IF, higher than twice 
the highest input  frequency, such a 30MHz input. First IF in the CEI-232 is 
65MHz! --- If you are  interested in the "receiver architecture" of 
dsp-receivers, read the second  installment in QE on this subject. It's somewhere in the 
list of DSP articles  below.

---------------------------------
Perhaps the high first IF might  have had more to do with image rejection, 
and perhaps most other manufacturers  were "clairvoyant" too as high first IFs 
were becoming very common at that  time.

>From your "higher than twice the highest input frequency" comment  It would 
seem you may be confusing IF frequency with Nyquist sampling  requirements.
There is not a requirement as such for a very high first IF  with DSP 
receivers and direct conversion to baseband is common.
However, ADC  limitations, as you suggest, has meant the retention of 
conventional hardware  front ends in many cases.

Aside from your rather optimistic  reclassification of the CEI unit, could 
you please enlighten us as to what other  receivers you are using that might be 
more in keeping with the "premium"  aspirations of the list?

regards

Nigel
GM8PZR