Go to the Tech Index page Back to Home

Measuring PL Response Time By Robert W. Meister WA1MIK

The myth is that it's better to choose a higher PL frequency because the lower PL frequencies take longer to respond. If, for example, the PL decoder must recognize 20 full cycles to determine that the PL signal is correct and valid, a 67.0 Hz PL tone would take 298 msec and a 192.8 Hz PL tone would take 104 msec. This article will attempt to confirm or bust this myth, in true Mythbuster's fashion. No radios were hurt during these tests, and there will be no "big booms" at the end. Sorry, Jamie.

While this article started with a Motorola MaxTrac radio, it has expanded to include a Motorola Spectra radio and a GE MLS-II radio as well. I tested dash-mount mobile radios in proper working condition. To figure out if the computer-decoded radios are any better than the old mechanical reed decoders, I also ran similar tests using a Motorola MICOR receiver and a stock PL decoder. Where possible, I also tested DPL.

Rather than use the generic Continuous Tone-Coded Squelch System (CTCSS) and Digital Coded Squelch (DCS) terms, I'm using the Motorola trademarked PL (Private Line) and DPL (Digital Private Line) terms throughout this article. These are equivalent to the GE/Ericsson Channel Guard (CG) and Digital Channel Guard (DCG) terms.

Radios Tested:

I have a good assortment of MaxTrac, Spectra, and MLS-II radios, so I chose UHF 45 watt models. I programmed five modes on the same simplex frequency but with different PL and DPL codes. One of the accessory jack pins on the MaxTrac was programmed to output a COR+PL signal, which was monitored by the scope. One speaker line was monitored on the Spectra; about 6VDC is present on both speaker lines when the radio's Audio PA Enable signal is activated, and this only happens when the radio is decoding a properly encoded signal. I had to dig into the MLS-II radio and solder a wire to an IC pin carrying the RX MUTE signal, which is equivalent to the Audio PA Enable signal on the Motorola radios.

The only PL reed decoder I have is in a PURC5000 link receiver chassis. This unit currently has a UHF MICOR receiver board installed in it, although any band can be plugged in. It also has an audio/squelch board, PL decoder board, and an audio amplifier, so it can drive a loudspeaker directly. (See the "PURC5000 Link Receiver" articles in the MSF&PURC section of this web site.) Various signals are brought out to a terminal strip on the back. The PL board uses standard MICOR Vibrasponder decoding reeds, commonly called mini-reeds. These are slightly larger than the Vibrasender encoding reeds. A DPL decoder board can be plugged into the audio/squelch board, but the one I have doesn't work.

Setup:

I used an RF signal generator set for 750 Hz peak-to-peak deviation as confirmed with a spectrum analyzer with an FM demodulator (my modulation meter is unable to measure DPL deviation). This let me set any PL frequency. I could also feed a DPL signal generator into the signal generator, verify the same deviation, and thus test the same decode delay using DPL. The RF signal amplitude was 1mV.

I connected a dual-trace oscilloscope to trigger and view the modulating PL or DPL signal on the "A" channel and used the "B" channel to monitor an Audio Mute indication on the target receiver. I measured the time it takes for the microprocessor or PL decoder to recognize and decode the PL or DPL signal and un-mute the audio stage of the receiver. The diagram below shows the connections.

For each test, I turned the signal generator's modulation off, primed the scope for a single triggered trace, turned the modulation on, and used the scope's cursors to measure how long it took for the radio to recognize the PL or DPL. Each synthesized radio had receive-only channels programmed with the desired PL and DPL codes ahead of time.

Tests Performed:

I chose 67.0 Hz, 100.0 Hz, 141.3 Hz, and 192.8 Hz for the tests, primarily because I had PL decode reeds for those frequencies. The synthesized radios can be programmed to decode any PL frequency or DPL code. The RF signal generator can output any PL tone required, or use an external source; in this case the DPL Test Set. The DPL Test Set can be set to encode any DPL code but I chose 311. The table below indicates the time, in milliseconds, that the radios took to detect and decode the PL/DPL signal fed into it. Each entry is the average time in milliseconds for 10 test runs. The "Var" columns indicate the +/- span of the variation from the fastest to slowest time of all the runs.

Radio	67.0	100.0	141.3	192.8	Var	DPL	Var
MaxTrac	190.2	194.8	218.0	210.6	33-51	161.8	39
Spectra	248.6	245.6	247.8	245.0	13-17	139.8	41
MLS-II	266.8	189.6	138.6	113.8	7-33	176.2	30
MICOR	180.0	182.0	208.0	200.0	0	279.4	28

Observations:

Since the radios were receiving an uninterrupted carrier during the entire test, the front panel BUSY indicator was either blinking (MaxTrac) or always lit (Spectra and MLS-II). I did notice that the decode times seemed to fluctuate wildly on the MaxTrac, however I couldn't determine whether it was longer if the BUSY indicator was on or off. The results were just as erratic when I switched the RF carrier on and off rather than just the modulation.

PL Decoding Results:

The MaxTrac PL decode times had a lower average than the Spectra decode times, but the variation in decode times for the MaxTrac spanned over +/- 50 msec, whereas the Spectra variation time was much more consistent at +/- 15 msec. The MLS-II fared much better, with very consistent decode times, and unlike the Motorola radios, its decode time DID go lower as the PL frequency increased, sampling between 16 and 20 cycles of the PL tone for each frequency. The Vibrasponder reed had identical timings for each of the 10 trials, however the decode time appeared to be longer for the higher frequencies. I did not expect this. Also, the decode time lengthened slightly with greatly reduced input deviation: at 100 Hz deviation with the 67.0 Hz reed, the decode time increased from 180 to 186 msec for each run. At 200 Hz and higher deviations, the decode time was as shown above. The PL decoder board provides a lot of filtering, gain, and limiting, and it drives the reed with a lot of clean signal.

DPL Decoding Results:

The MaxTrac DPL decode time had a variation of +/- 39 msec. The Spectra DPL decode time had a variation of +/- 41 msec. I'm guessing they use similar algorithms. The MLS-II was in the same ballpark and reasonably consistent with the lowest variation of all the radios tested. I could not get the link receiver's DPL decoder to function so I could not test DPL on that unit. It uses a custom integrated circuit for encoding and decoding, so it should be relatively fast. The DPL Test Set uses two of the same ICs, so I tested that unit's decoder instead. I could not find a spec on the decode time in the manual; the times I got seem way too long.

Conclusions:

The original myth was that lower PL tones take longer to decode than higher PL tones. This may have indeed been the case in the 1960s and 1970s when the big copper PL reeds were in use, but these days the small mini-reeds and newer synthesized radios don't seem to have enough frequency-dependency to make a noticeable difference in decode time. There was too much variation in the test results to make an absolute conclusion, however for the radios and frequencies I tested, this myth is busted for some radios (Motorola MaxTrac, Spectra, and mechanical reed decoders) yet confirmed for others (GE/Ericsson MLS-II).

Test Equipment Used:

• Agilent E4430B RF Signal Generator
• Agilent E4401B Spectrum Analyzer
• Fluke 199C Portable Oscilloscope
• Motorola SLN-6413A DPL Test Set

Credits and Acknowledgements:

MaxTrac, Spectra, MICOR, PL, DPL, Vibrasponder, Vibrasender, and a bunch of other terms are trademarks of Motorola, Inc.

MLS-II, Channel Guard, and Digital Channel Guard are trademarks of GE/Ericsson.

Contact Information:

The author can be contacted at: his-callsign [ at ] comcast [ dot ] net.

Back to the top of the page
Go to the Tech Index page
Back to Home

This page created Sunday 25-Dec-2011

Diagrams, article text, and hand-coded HTML © Copyright 2011 by Robert W. Meister WA1MIK.

This web page, this web site, the information presented in and on its pages and in these modifications and conversions is © Copyrighted 1995 and (date of last update) by Kevin Custer W3KKC and multiple originating authors. All Rights Reserved, including that of paper and web publication elsewhere.