Modified SCAF-1 Noise Filter For AM Applications

Modified SCAF-1 Noise Filter For AM Applications.

by Stewart (“Stu”) Personick, AB2EZ

Background

About a year ago, I built an Idiom Press SCAF-1 audio filter to use with my modern transceivers in CW applications.  Recently, I decided to modify a SCAF-1 for use in AM applications:  both transmitting and receiving.

My original objective was to demonstrate, to the AM community, the use of a switched capacitor filter for “brick-wall” bandlimiting of the output of an AM audio chain – in order to limit the bandwidth of the r.f. output signal produced by a vintage high-level-modulated vacuum tube transmitter, or a modern FET-based “Class E” transmitter.

Assuming that an AM transmitter is being operated properly (i.e., minimum negative peak limiting), and assuming that the transmitters modulation characteristic is reasonably linear – limiting the input audio bandwidth will do wonders to reduce interference produced by a strong transmitted signal into adjacent weaker signals.  Therefore, using an adjustable brick-wall audio low pass filter would address some of the complaints that operators of powerful AM stations (legal limit power + an efficiently-radiating antenna) receive about being “wide”.

A second very useful application has been in removing high frequency hiss from the output of my Drake R-4C receiver when using it on AM.

There are, of course, many alternative technologies one could use today to build a low pass audio filter with a sharp cutoff.  For home brewing, the use of switched capacitor filter integrated circuits, such as those used in the SCAF-1 seemed like a good approach to try.  After considering purchasing all of the necessary components, laying out a circuit board, and putting everything in a nice looking box, it made sense to me to just start with a SCAF-1, and change a few components, as needed.

Modifications

The required modifications fell into three main categories:

 – Remove or change the values of components that were limiting the low frequency response of the filter… so that the filter would be flat down to at least 50 Hz
– Remove or change the values of components (other than the switched capacitor filter IC chips) that were limiting the high frequency response of the filter, so that the filter would be flat up to at least 4.5 kHz when set to maximum bandwidth
– Modify the values of a few components associated with the 455 timer IC chip, so that the adjustable cutoff frequency of the filter would range from 1.5 kHz to 4.5 kHz

The needed component changes are shown in Table 1.

Note that the two 745 op-amp high-pass filter stages could have been bypassed, or could have been converted to low pass “roofing” filters (e.g. each with a 6 kHz cutoff frequency).  I chose to simply change a few resistor values, to move their existing high pass functionality to a lower cutoff frequency.

Since all of the high pass filter cutoff frequencies have been moved to below 50 Hz, one has to use the modified SCAF-1 with a well-filtered 12 volt power supply to avoid 60 Hz hum in the output.

Results

I was pleased with the results.

Using a simple home-brew white noise generator and a PC-based audio spectrum analyzer, I determined that the passband of the modified SCAF-1 is flat, and the high frequency cutoff is adjustable from 1.5 kHz – 4.5 kHz, with the expected brick wall characteristic.

When using it in conjunction with the audio chain of my AM station, and using a local off-air monitor, I was able to hear the desired, adjustable bandlimiting effect on my transmitted AM signal.  Listening to my local off-air monitor with my headphones, I could hear the loss of high frequencies (both room noise from things like fans and high frequency components in my voice) as I adjusted the cutoff frequency of the modified SCAF-1.  Naturally, with the cutoff set at 4.5 kHz, my audio is more “hi-fi”, but my signal occupies more bandwidth, and will produce more interference into “QSO”s on nearby frequencies – particularly when my signal is received at substantially higher signal strength than the signals I am interfering with.

The audio file SCAF-1.mp3 illustrates the effect of the modified SCAF-1 filter on what I sound like at the output of my off-air monitor when I vary the cutoff frequency between 4.5 kHz and 1.5 kHz.

Separately, I also found the modified SCAF-1 to be very helpful in removing high frequency noise from the output of my Drake R-4C receiver.  This noise is particularly noticeable and the filter is particularly helpful when using the Drake R-4C to listen to a weak signal, or to monitor the band when no signals are currently present.  Even though the R-4C has a good 1st i.f. filter, on AM it has no 2nd i.f. filter.  When a weak signal (or no signal) is received, the noise at the speaker output, from stages beyond the 1st i.f. filter, is quite high, and extends well beyond the range of frequencies I am trying to listen to.