EMRFD Message Archive 7495

Message Date From Subject
7495 2012-05-09 15:27:29 kc8sbv Audio Filtering
I am perplexed here...with the 4SQRP Club selling a 200hz audio filtar, would it filter out the opposite sideband in a DC RX, or do you need a Superhet, or IQ phasing DC RX?

7496 2012-05-09 15:42:56 Dan Mills Re: Audio Filtering
On Wed, 2012-05-09 at 22:27 +0000, kc8sbv wrote:
> I am perplexed here...with the 4SQRP Club selling a 200hz audio filtar, would it filter out the opposite sideband in a DC RX, or do you need a Superhet, or IQ phasing DC RX?

You need a superhet or phasing DC RX, by the time the signal hits the
output of your mixer in a conventional DCX rig, all information about
which sideband energy comes from has been lost.

73, Dan M0HCN.
7497 2012-05-10 06:12:07 Roelof Bakker Re: Audio Filtering
Hello Ed,

Though this filter does not remove the other sideband, it will help in
avoiding QRM from nearby stations.
I use a narrow filter in a DC RX and it really helps.
Changing the sideband you listen to, will often help as well.

Roelof Bakker, pa0rdt
7498 2012-05-10 09:10:32 kc8sbv Re: Audio Filtering
I am going to hazard a guess here...in a DC RX there is no room for one sideband to range from 0 Hz to 1000hz, and the other to be less than 0 hz, and thus the negative sideband mirrors over. In a superhet, there is room for both sidebands to mix to the IF center freq and be above and below the IF center freq, allowing for one of the sidebands to be filtered out.

Thats the best my head can conjure up as to why a 200hz filter works in a superhet, but not in a DC RX.

7499 2012-05-10 12:10:09 wa5bdu Re: Audio Filtering
Well, the sidebands in a CW signal are somewhat of an illusion rather than being actual sidebands like with an AM or SSB signal.

Consider a key-down continuous signal for simplicity. As you tune across the band, you hear a high note, then lower, then zero beat, then higher, and then gone. What you heard one either side of zero beat were not sidebands but were the result of the fact that the zero BW signal you were listening to can produce tones with the LO tuned N Hz above and N Hz below the transmit frequency. Say 7040.5kHz and 7039.5kHz for the 7040kHz transmit signal.

It's maybe easier to see with a DC receiver than with a superhet that doesn't give "single signal" CW reception. Say you're listening with a hallicrafters S-40 receiver with an IF at 455kHz and an IF bandwitdh of 15kHz at best. So you adjust your LO again for the two 500Hz signals 1kHz apart and produce in the IF signals at perhaps 455.5kHz and 454.5kHz. Both fit nicely in the 15kHz BW of the IF and so both are heard at equal loudness after detection with the BFO.

Now suppose you have an IF filter of 1kHz bandwidth centered
7502 2012-05-13 02:31:10 Joe Re: Audio Filtering
Extracted from: [emrfd] Digest Number 1611
> 1c. Re: Audio Filtering
> Posted by: "wa5bdu" kennnick@gmail.com wa5bdu
> Date: Thu May 10, 2012 12:10 pm ((PDT)
> You can't use a simple narrow audio filter to get rid of the extra "sideband" with a CW signal.
This is true. However the simple audio filter may provide great
improvement in the receiver's ability to remove interference.

Given an audio filter tuned to 800 Hz with a bandwidth of 200 Hz, what
signals relative to the zero/LO frequency are actually audible?

Well, it's the 200 Hz centered 800 Hz above the zero/LO frequency AND
the 200 Hz centered 800 Hz below the zero/LO frequency.

For interference to be heard, it must be within 100 Hz of the signal you
are trying to copy OR in a 200 Hz window located 1.6 KHZ either above or
below the desired signal in the RF spectrum.

The effectiveness of this approach is related to Rx frequency stability,
filter bandpass shape, and the like.

Taken to the extreme, the PA0LQ Gaussian audio filter is an example of
an analog design with a very narrow bandpass which implemented in a DC
Rx would provide a total bandwidth measured in the tens of Hz for both
the desired and image sideband.

This isn't very practical because the implementation requires the
remainder of the receiver's functional parts to be much better than what
is part of most every DC receiver.

Joe, K9HDE

PS Nick, I just used your statement as a place to redirect the
conversation towards a new view of the use of audio filters in DC
receivers. No offense was/is intended.

PPS If anyone can devise a revision to the PA0LQ filter to allow it to
be tunable over even a small range, say +/- 20 Hz, I'd love to try it
7503 2012-05-13 07:25:27 Roelof Bakker Re: Audio Filtering
Hello Joe,

I have done some work on the PA0LQ audio filter.

The original design has a center frequency of 1000 Hz, which is a bit too
high for my liking.
The design has been changed for a center frequency of 500 Hz.
When all four filter sections are tuned for maximum, the bandwidth is a
few Hz, way to narrow for practical use.
I have used stagger tuning and have used a bandwidth of 6 Hz and 12 Hz.
This is a nice value for slow morse, as used by non directional
aeronautical beacons at longwave.

To shift the center frequency of this filter over a small range, requires
4 resistors to be changed.
This is not very practical.
A better approach is to use what I call a pitch shifter.

In the past, I have used selective level meters as longwave receiver.
This works fine, except for the unusual high beat note.
A Wandel & Goltermann SPM-30 SLM has a beat note of 2000 Hz.
(And a narrow filter 24 Hz wide!)
Using a NE602 and a local oscillator at 2500 Hz, the beat note was changed
to 500 Hz.
When you use a tuneable LO, the filter can be moved over the passband.

I have written it up and can sent you the details if you like.

Roelof, pa0rdt