Steve I don't see the return loss plot, but 22db would be quite good. The 2.1khz BW is right on the money for SSB. To the best of my knowledge, the AADE software doesn't design QER filters. How did you do that? BTW, did you test your crystals for Rs? What method did you use? Also, keep in mind that grounding the xtal cases adds a little capacitance from each side to ground. This will also contribute to BW narrowing. Sent from my tablet -------- Original message -------- From Steve VK2SJA <vk2sja@xxxxxxxxxxx> Date: 09/28/2014 3:43 AM (GMT-05:00) To minima@xxxxxxxxxxxxx Subject [minima] 20Mhz Crystal Filter Hi All, Thought I'd share this with the group. I think you call it success? Yesterday taking a break from mixers I built a 20Mhz Crystal Filter. This is in fact my first home-brew ladder filter. And it seems to have worked a treat. But my passband appears to be a *lot* narrower than others are reporting. Which is my main reason for sharing. My crystals are full sized HC-49 and sourced cheap from eBay. Purchased about 50 off them from memory. I profiled a batch of 15 using a G3UUR test oscillator I built. Out of interest and a warning I suppose, I found three crystals during this process that would not "start" in the G3UUR test circuit. Not sure if they are truly bad or if they just don't like that oscillator circuit. Guess that's why they were cheap. After removing these from contention I managed to select 8 crystals with 66Hz of each-other. I plugged the required characterization information into the AADE Filter program and asked it for a filter 3Khz wide. I wanted something a bit narrower than 4 to 5Khz :-) The AADE Filter program told me to use 83pF caps from memory. At this point I was confused. As increasing the capacitance narrows bandwidth I thought and Farhan's Minima circuit uses 100pF caps for something like 4.5Khz BW, or so I thought. Anyway I went and built it with the 100pF caps (should have listened to the software !-) and this is what I got. -3dB BW 1.940 kHz -6dB BW 2.160 kHz -45dB BW 4.070 kHz Insertion loss 3.64dB As per the screen shots below. Also, today I measured the return loss of the filter. Another "first" for me. Trying to get a handle on the characteristic impedance of the filter. AADE Filter software was predicting something like 88 ohms from memory. I measured a return loss of 22.85dB. Which if I'm managed to understand things properly is closer to 60 ohms but at least we're in the same ballpark. So what I learnt from this is:- 1. Be careful of cheap bulk-buy eBay crystals. Some of them may not start. Test before using I'd suggest. 2. Not all crystals are the same and unless you like taking pot-luck with your filter BW it's probably a good idea to characterize them. 3. The software knows what it's talking about! :-) And my questions for the group are:- 1. Is this filter going to be alright for SSB phone? Or is it too narrow? 2. A return loss of 22.85dB seems pretty good to me. But I'll ask. In general terms what level of return loss is acceptable? And at what point should we start building matching circuits? 73, Steve.