EMRFD Message Archive 7225
Message Date From Subject 7225 2012-02-03 10:04:44 Ben Pharr Simple BPF for DC Receiver
I have built a few kits and done some digital electronics work, but
I'm getting ready to begin my exploration of RF design. I have decided
that my first project will be an incredibly simple direction
conversion receiver for the 80m SSB band. It'll use an NE602 and an
LM386. I'll expand from there.
First, I need a suitable bandpass filter to go in front of the NE602.
I've read the appropriate sections of EMRFD and I think a double-tuned
filter will do the trick, but I'm not sure how to modify the filter to
match the receiver's input impedance of 50 ohms to the 1500-3000 ohm
input impedance (depending on who you ask) of the NE602.
I saw the section in EMRFD on modifying the output impedance of an
LPF, but it's not immediately obvious to me how to apply that to a
BPF. I've also played with the version of ELSIE that comes with the
Handbook, but it doesn't look like that version allows you to change
the output impedance. I've looked at several existing designs, but
either they mentioned nothing about matching the input impedance of
the NE602 or they used some type of filter I wasn't familiar with and
didn't explain it well enough for me to be comfortable using it.
For what's it's worth, I have a large selection of capacitors and -2
and -6 toroids. I've also looked at the design of the Sudden receiver
and noticed that they used tunable inductors. I'm open to anything
that's simple and understandable.
I feel a little silly asking what seems like such a simple question,
but I want to do it right. Thanks in advance for any suggestions.
Thanks,
Ben7227 2012-02-03 12:17:39 ha5rxz Re: Simple BPF for DC Receiver
A quick thought occurs to me after reading your post and looking at the input stage of the Hybrid Cascode Amplifier where there is a similar problem. Let's assume that the input impedance of the NE602 is 2,000 ohms to make the math easier so just connect it to the output of the filter and put a 51 ohm resistor in parallel. 2,000 ohms in parallel with 51 ohms is 49.73 ohms, close enough.
Peter HA5RXZ
7228 2012-02-03 13:58:44 R Wall ML emails Re: Simple BPF for DC Receiver
An L matching circuit may help:
http://www.daycounter.com/Calculators/L-Matching-Network-Calculator.phtml
Roderick Wall, vk3yc.
7229 2012-02-03 14:31:24 Wes Re: Simple BPF for DC Receiver
If you take a look at the LADPAC software you will see that the program for a double tuned circuit will accomodate a wide range of resistive terminations. So you could match to the 3K input impedance of the NE602.
If you dig into the manual for the software, which is with the rest of the files, you will find that there is an applicati7230 2012-02-04 01:21:09 Roelof Bakker Re: Simple BPF for DC Receiver
There are several ways to solve your problem.
What I have done is make a tap on the coil about one third from the
grounded end.
This is about right.
You can also use a step up rf transformer from 50 ohm to 1500 ohm.
I hope this helps,
73,
Roelof Bakker, pa0rdt7231 2012-02-04 15:11:31 Ben Pharr Re: Simple BPF for DC Receiver 7232 2012-02-05 11:04:27 Wes Re: Simple BPF for DC Receiver
Glad that this was the info you needed.
Yea, it's OK to parallel caps, although you can get in trouble with interactions where the stray L of one cap resonates with the C of another. See the info in Ch2-decoupling for more detail. Ignore the Smith Charts if you find them frightening, as many readers do. An added low pass will fix the problems with resonances.
For your particular filter, you could get to slightly more practical C-t values by dropping the inductance to about 12 uH instead of the 18 that you used. Some of the programs for filter design point you toward a specific L value, but this is not really required. There is often more freedom available.
There is another problem. If you design a traditional filter that has the wide 13% fractional bandwidth you have specified, the stopband performance gets really poor above the passband. Try GPLA with a sweep up to 50 or 100 MHz. You can fix this again with an added low pass at the 50 Ohm end. An alternative would be to use a different filter topology. Take a look at the alternative that's in the program, although that might be rough with a high Z at one end. Same goes with the alternative bandpass that we used with the Lichen transceiver in Chapter 6.
Have fun.
73, Wes
w7zoi
7244 2012-02-13 07:47:46 Ben Pharr Re: Simple BPF for DC Receiver 7274 2012-02-20 05:50:28 Tim Re: Simple BPF for DC Receiver
10.7 MHz IF transformers are widely avaialble real cheap. e.g. Mouser 42IF222 and help match the 50 ohms to the higher impedance NE602 inputs.
In these transformers, it's the high impedance side that is tuned. By adding parallel capacitance to that high impedance side, it's pretty easy to make the high impedance side go to the 80M band (330 or so pF for the named Mouser part) or the 40M band (47pF if needed).
Now this single tuned circuit may not be everything needed to keep out of band signals from overloading the NE602, but it is a useful component in a larger solution. Or, in many NE602 based rigs used with modest antennas, it may be all that's needed.
7315 2012-03-08 08:45:23 Tim Re: Simple BPF for DC Receiver
The Small Wonder Labs transceivers use IF transformers extensively to both band pass filter and match to NE602.
http://smallwonderlabs.com/docs/SW40+_manual.pdf
e.g. in the SW40+ there's a single pole IF transformer filter/match at the input of NE602 U1 as well as two poles at the output of NE602 U5. They use the little mouser IF transformer parts extensively in their designs as does KD1JV.