At 3/19/2009 08:18, you wrote: >Hi Doug, > >Many years ago I had a similar situation. Instead of using the N06B filter >I used the built in hi-pass filter in a Comm Spec PL encoder/decoder >board. It is very similar to the NO6B filter.
Actually the reason I developed my own HPF was because of the poor performance of the filter in the CommSpec board. It's transient response is poor, causing a ringing "boxy" sound around 400 Hz. My design is an adaptation of the HPF used in the Sigtone C1116 CTCSS board. All I did was move the 3 dB point down from 300 Hz to ~210 Hz, since I don't use any of the higher CTCSS tones & didn't want to filter the audio any more than necessary, since some of the outgoing paths end up being HPF'd again downstream by 300 Hz cutoff HPFs. Even at 151.4 Hz the attenuation is about 20 dB - good enough for me. > What was happening was that the filter was introducing "twist" (subtle > distortion) to the DTMF waveform. IIRC both the CommSpec & Sigtone designs are flat once you're well within the voice band (600 Hz & higher). But if 2 or more CommSpec filters are cascaded in a system the high pass rolloff may be high enough to attenuate some of the low DTMF tones. But that's not your problem if you're using my design. I suggest looking at the output of the filter with a scope to see if there's any ultrasonic oscillation that could be confusing the DTMF decoder downstream. You do have that 27k or some value of resistor >= 1k on the output op amp (pin 8), right? An output isolation resistance is required to prevent external capacitive loading from causing that op amp to oscillate. Bob NO6B
