Unitrunker EDACS decode problems

[tbiggums]

I having trouble getting a radio to work well with Unitrunker Retro and Etrunk.

The discriminator waveform looks pretty good on a stand-alone oscilloscope, so I'm not sure why Unitrunker is only decoding around 50%. Below are some screeshots from Uniscope.

Is it possible there's some ground loop issue or something? I think the problems I'm having with Etrunk and the data slicer are due to a ground loop, as some 60 Hz signal appears once I connect the radio to the data slicer.

I've tried two different PC's for Unitrunker, and the decode rate never gets much above 50%. The one that the Uniscope screeshots were made from is a 3 GHz Pentium running XP Pro.

Initially, the waveform from the capacitively coupled discriminator tap was obviously clipping on Uniscope, so I made a voltage divider out of a couple of resistors to get it to the level shown in this screenshot.

 

[EJB]

Same thing here. I get 100% signal running Trunk 88 and barely 50% running a perfect signal for my local EDACS tower.

I am not impressed.

 

[SCPD]

Please post another screenshot with the symbol type to "impulse". The signal looks a little top heavy (positive pulses have more amplitude than the negative pulses).

 

[tbiggums]

Thanks for the reply! Here's a screenshot of the same EDACS signal with Uniscope set to "Pulse" mode.

These screenshots are from a GE FMD EDACS mobile. The discriminator on the FMD radio seems to usually sit around +2.5 VDC, with a 1V p-p signal on it when tuned to an EDACS CC.

I'm capacitively coupling to the discrimintor via a 0.1uF cap to remove the 2.5V DC component. I'm also using a resistive voltage divider to keep the signal from clipping at the soundcard input.

This FMD isn't working with a data slicer and Etrunk, either. I haven't tried using the voltage divider with Etrunk yet...

I have also been using a PRO-2006 and PRO-2035 with discriminator taps, and they seem to work better than this FMD with both Etrunk and Unitrunker. Unfortunately these two scanners suffer from adjacent channel interference, so that's why I'm trying to get Etrunk or Unitrunker working with the FMD.

 

[SCPD]

Yuck. Try increasing the amplitude slightly. The goal is to get an "eye" pattern where the center of the eye is clear of any traces. It should look more like this (image created with a different version of uniscope):

 

[SCPD]

You might try using the resistor divider only - without the DC blocking capacitor. I suspect your computer's audio input already includes DC blocking capacitor(s).

 

[tbiggums]

I played around with it a bit more last night, and no matter what amplitude I fed into the soundcard from the FMD, I couldn't get a good eye pattern on the "pulse" display. I didn't try removing the blocking capacitor from the radio yet, but I'll give that a shot...

I did hook up the PRO-2035 again, and made a few screeshots which are shown below. As with the FMD, I have to use a voltage divider to keep the signal from clipping at the soundcard input. Once I got the amplitude dialed in good, Unitrunker Retro was decoding 98%-100% with it (with no adjacent channel interference).

If the FMD is really producing a "top heavy" signal, is that something that going through a full alignment on the receiver should resolve? How common is this problem?

Just comparing the "Oscilloscope" screenshots between the FMD and PRO-2035, they seem pretty similar, but it does appear that some the more narrow FMD negative pulses just barely go below the "zero line". I'm guessing that's the real problem...

PRO-2035 screenshots:

 

[SCPD]

Another experiment worth trying is replace the resistor divider with a series resistor. Is the de-coupling capacitor located before or after the resistor divider?