[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [linrad] A simple front end

Hi Roger andf all,

> I bought 2 R2PRO kits when they first came out at Dayton about 15 
> months ago,
> thinking I would try them with Linrad.  I have not built them as 
> I was not sure how I
> was going to get 90 KHz bandwidth out of them.  My simple 
> understanding was that it
> would be very difficult to get a 90 KHz-wide signal to feed into 
> the soundcard and
> maintain the proper phase relationships across the 90 KHz 
> bandwidth. 
I do not know what R2PRO is, but maintaining the phase relation
is not an issue. Linrad has a calibration routine to take care
of that in case it is a problem. It is hard for me to imagine,
the calibration is to take care of the phase shifts introduced
by the differences between the channels in case high Q anti-alias
filters are used (as is done in RX2500)

> It is the audio phase-shift network
> and diplexer that I thought would not have the necessary bandwidth.
????? These parts should be bypassed. Do not sum the signals in the
analog hardware - that is for Linrad to do.

> My recommendation would be, for 20 MHz CW;  
> either direct conversion, which I haven't tried, or dual conversion with:
> RF input to 20 MHz bandpass filter [3 coils and 2 capacitors] to
> RF Preamp.  Homebrew or MAR or ERA or something like a Cougar or 
> Q-Bit amplifier to
> TUF-1 or TUF-1H or TUF-3 or TAK-893 depending on how much LO 
> drive you can muster.
> LO could be either 24.7 or 3.3 MHz for 10.7 MHZ first IF.  to
> 10.7 MHZ, 50 KHz bandwidth surplus filter for 1st IF filter to
> homebrew discrete amp or MAR or ERA type device for 1st IF amp to
> TUF-1H 2nd IF mixer with
> 10.7 MHZ LO [exact LO depends upon your 1st IF filter width] to
> AD-797 Baseband amp to
> 96 KHZ sampling rate soundcard.
> I know this is very crude and simple but it worked very well for 
> me and the parts
> were available in my limited junkbox and at hamfests [ and eBay too].
This is a good way to go. You can easily extend to other bands.

> A direct conversion receiver has fewer parts, but I was more 
> confident that this
> design was relatively 'fool-proof' or at least 'fool-resistant', 
> which was what I needed ;)
The reason why it is so much easier is that you can place the desired
passband at audio frequencies from say 25 to 45 kHz. Then you can
have a lousy second order performance and still not be bothered
by audio overtones. The crystal filter will remove things below
22.5 that could have overtones within the passband. This is the
trick used in dsp-10 for example.


Leif / SM5BSZ