EMRFD Message Archive 13182

Message Date From Subject
13182 2016-09-27 17:08:05 IEN LPF BPF
Could I use a band-pass filter in a power amplifier output instead of LPF?
That way, the filter could be shared as output filter in Tx and input filter
in Rx mode.

San Nicolás
13183 2016-09-27 18:23:27 Tayloe, Dan (Noki... Re: LPF BPF

You can do this.  However, a BPF typically has more loss than the TX LPF.  Not many folks want to lose 2 or 3 dB of TX power in the output filter. That is why you normally see the arrangement we have today, TX LPF followed by a RX BPF. 


-          Dan, N7VE


13184 2016-09-27 18:35:24 Tayloe, Dan (Noki... Re: LPF BPF

I think also if a BPF were used on the TX side, the voltages could get rather high.  The BPF caps would need a voltage rating to match the expected voltages.  If you don’t know what the voltages might look like, you can always simulate the circuit or try to run the math manually.

-          Dan, N7VE


13186 2016-09-28 04:34:39 Steve Dick Re: LPF BPF
I just made a QRP bandpass filter which has the advantages of being shareable between transmit and receive.  It is based on the work that was done by W3NQN in May-June 1998 issue of QST magazine (“Clean up your signals with bandpass filters”).  These are basically order 3 Checychev filters but with a twist. They used trifilar wound or quadrafilar wound inductors for the input and output inductors. Read the article to get a better understanding of what their advantages are.  My QRP filter used bifilar wound inductors for simplicity for the input and output inductors.
I used T50-6 toroids and 500 volt silver mica capacitors for the caps.  The filter has a measured loss of less than 0.6dB and measured attenuation greater than 50dB at 3.5 MHz and at 14 MHz.  I would have so say though that it is impossible to align the filter without proper test equipment – a return loss bridge, signal generator, and power meter.  It also is quite component sensitive. Very small changes in inductor windings lave large effects on alignment. But it can be aligned in an orderly fashion, and if you are patient and take your tile with alignment, you can get great results.  I will add a small writeup in the “files” section within the week on this filter.  If you want to get an idea of filter performance, use the ELSIE program with the following parameters:

Topology: Shunt input bandpass
Family: Checychev
Bandwidth: 908.6K (The oddball value allowed use of standard capacitors)
Center Frequency: 7.15M
Order: 3
Input termination: 50
Passband Ripple 0.018 (Note: This results in return loss of –23.8dB. You can control the return loss by adjusting the passband ripple parameter a nice advantage over Butterworth filters)
Plot it from 3.5 to 14 MHz.
Note again that L1 and L3 are bifilar wound, where L2-C2 sit on the halfway tap of bifilar wound L1 and L3. That means the actual inductance values for L1 and L3 are 4 times what is shown in ELSIE (twice the number of turns on L1 and L3 for a standard inductor) and C1 and C3 are one quarter of what is shown in ELSIE.  The 50 ohm tap point on the bifilar winding allows this impedance transformation.
“Digital Steve”, K1RF
13187 2016-09-28 09:44:18 IEN Re: LPF BPF

Its a good idea, but, I wanted to use only one filter to ease switcing bands.
----- Original Message -----
13188 2016-09-28 09:45:08 IEN Re: LPF BPF
Thanks Dan.
----- Original Message -----
13189 2016-09-28 14:11:23 IEN Re: LPF BPF

Just  the filter I had in mind.-
----- Original Message -----