Scandata faq q11

Data Slicer Modifications
By Bill Cheek

Thanks to Allan Dunn and Urban Larssen, I found out some neat stuff about these "data slicers"
commonly used in decoding MDT, trunked control, pagers and other NFM digital signals.

One of the designs, (I recently spoke of three), has a 10-k resistor from pin 3 of the LM-741 chip
to ground.  Turns out, this 10-k resistor may not be optimal.

I had a PRO-64 on my bench that just would not decode any trunk data.  Yet the same "data slicer"
moved over to my PRO-2004 did the job just fine using the same trunked control channel.

The output of the NFM Discriminator (baseband audio) on the PRO-2004 was about
.55 VAC (RMS), while the same signal on the PRO-64 was .33 VAC (RMS).  I found a 
shunt capacitor in the Discriminator output circuit of the PRO-64 (C-316 0.001-uF) 
that I didn't really think was siphoning off much of the signal, but just in case,
I removed it anyway.  No change. 

So, I went into the "data slicer" and soldered a 4.7-k resistor in parallel with the existing 10-k
resistor (which dropped the effective resistance to right at 3200 ohms: [R1*R2]/[R1+R2], and lo!
Suddenly valid data appeared on the screen.  Voile!

I think it was Mr. Larssen who suggested a trimmer potentiometer in place of the
10-k resistor might be a good idea, and this would seem to bear out. However, since the 
decoding of the PRO-2004's data wasn't affected by the lowering of that resistor value,
it might be acceptable to just use a standard 3.3-k resistor in lieu of the 10-k that
is specified. 

Some designs call for Pin 3 of the LM-741 to be grounded but Mr. Dunn suggested this
might make the circuit too critical or perhaps unstable.  Either he or Mr. Larssen
suggested a 1-k resistor might be a more suitable value.

Regardless, if your "data slicer" seems not to work, before you panic, take a second
look at the design of the circuit you used and measure the AC value of the signal
at Pin 2 of the LM-741.  You should have roughly 0.5v (AC) there from the Discriminator.
If not, then the value of that Pin 3 10-k resistor might be critical.  10-k seems not 
critical for stronger signals, but as I found in this PRO-64, it was weaker than expected.

Probably the weaker the signal, the more critical the design of the circuit.  The 100-k 
resistor from Pin 2 of the LM-741 is not critical and should be left as is.  Likewise,
the 100-k resistor from Pin 6 to Pin 3 should be left as is.  The one to focus on to "tweak"
is the 10-k at Pin 3 to ground.  Change it to a value of 1-k to 3.3-k if you have problems.

Also, I found problems with the choice of 741 chips. I pulled my hair out trying 
to make a uA741TC work, but the plain jane 741 op-amp from Radio Shack worked just 
fine.  Go figure.........

Return to SCANDATA.FAQ