EMRFD Message Archive 8630

Message Date From Subject
8630 2013-05-06 09:45:30 wb9kzy diplexer frequency for an upconverter
Hello -

I've seen the formulas for diplexer designs in EMRFD, the first thing is to pick the frequency - my question: what should that frequency be?

The reason for asking is that I was looking at the LF Upconverter design in June 1977 QST:

<http://p1k.arrl.org/pubs_archive/70030>

which has a diode mixer with a hi-pass / low-pass type of diplexer. W1FB & WA1LNQ mention that the hi-pass half of the diplexer should be tuned to 3 times the IF frequency, but they don't say why. I've also seen this 3x figure mentioned in Solid State Design for the Radio Amateur as well as in the W1FB book Practical RF Design.

Is 3 times the IF frequency a good choice for the diplexer frequency and why ?

Chuck, WB9KZY
8631 2013-05-06 11:15:22 Graham / KE9H Re: diplexer frequency for an upconverter
It is a real good choice if you are driving it with a square wave,
or driving it hard enough for the L.O. signal to go non-linear.

There will be a bunch of odd harmonic stuff that has to go somewhere.

--- Graham / KE9H

==

8632 2013-05-06 13:47:55 wb9kzy Re: diplexer frequency for an upconverter
Graham -

Thanks for the reply, that makes sense but I guess it begs the question, why not target the 2nd harmonic, too ? (just in case the square wave duty cycle isn't exactly 50%)

Chuck, WB9KZY

8634 2013-05-06 21:24:24 Leonard Re: diplexer frequency for an upconverter
This doesn't directly address your question but it may have some pertinent information as it shows different frequency amplitudes on a dbm..

http://golddredgervideo.com/kitsandparts/doubleballancedmixer.htm

This was using a dbm board from kits and parts.

http://www.kitsandparts.com/drm.php

Leonard
http://kc0wox.com
8635 2013-05-06 22:27:07 Ashhar Farhan Re: diplexer frequency for an upconverter
Chuck,
We understand that the diode mixer as well as many other passive
mixers are bidirectional, one could, just as well injection a signal
at the if port and get the mixed output from the rf port.
With that in mind, when we imagine that there is a whole spectrum
being fed into the rf port which gets converted up or down and gets
out of the if port and fall up (let's say) the crystal filter. The
crystal filter allows only signals within its passband through and
reflects the rest back into the mixer. These signals, as they traverse
the diodes in the mixer, upsets their normal switching behaviour,
creating IMD.

Now, our hero, the diplexer appears amid this civic unrest. Its job is
simple, keep the signals coming out of the mixer from going back into
it.

Let's start with a simple one, you have a narrow band convertor RF
front end centered on 28 mhz, down converting to 14 MHz with a 42MHz
local oscillator. So, you are going to be bothered with the image at
70 MHz (28+42). You could simply use a low-pass filter to take out the
14 mhz energy and use a high pass to divert the 70 mhz energy to a 50
ohms resistor. What should the cut-off od these filter be? You can
take a shot at it, plug it into spice and see what happens.

You could, instead have a combination of bandpass / bandstop centered
around the IF, such that bandstop takes all the energy NOT at the IF
and harmlessly used it to warm up a dummy load.

Now, you can start looking at the complications of how the reject/pass
pieces work with each other inside the diplexer.

Trask wrote a very nice paper on diplexer. You can see it on his home
site. Or you could reach out for your IRFD (I hope it is on your
bookshelf, I find myself refering to it more than to emrfd, off late).

- farhan

On 5/6/13, wb9kzy <wb9kzy@arrl.net> wrote:
> Hello -
>
> I've seen the formulas for diplexer designs in EMRFD, the first thing is to
> pick the frequency - my question: what should that frequency be?
>
> The reason for asking is that I was looking at the LF Upconverter design in
> June 1977 QST:
>
> <http://p1k.arrl.org/pubs_archive/70030>
>
> which has a diode mixer with a hi-pass / low-pass type of diplexer. W1FB &
> WA1LNQ mention that the hi-pass half of the diplexer should be tuned to 3
> times the IF frequency, but they don't say why. I've also seen this 3x
> figure mentioned in Solid State Design for the Radio Amateur as well as in
> the W1FB book Practical RF Design.
>
> Is 3 times the IF frequency a good choice for the diplexer frequency and why
> ?
>
> Chuck, WB9KZY
>
>

--
Sent from my mobile device
8637 2013-05-07 05:43:36 wb9kzy Re: diplexer frequency for an upconverter
Leonard -

Thank you for the reference, the spectrum analyzer pictures make the situati
8638 2013-05-07 05:46:42 wb9kzy Re: diplexer frequency for an upconverter
Farhan -

Thank you for the illuminating reply - I will dig out "the red book" and start reading !

Chuck