March 2012 Archives

Less Syntax Highlighting

Motorola Mitrek and TinyTrak3

I can't believe this hasn't come up, but a little searching seems to indicate that not very many folks interface a Motorola Mitrek to a TinyTrak3.

I ran into an issue with a Motorola Mitrek radio and a TinyTrak3. I was trying to determine what the radio was expecting as far as a modulating signal was concerned. Motorola, in the lowband Mitrek manual, indicates that a 1KHz tone should be generated at 1V peak-to-peak and feed into the microphone input through a .33uF capacitor. The DEV adjustment should be turned until deviation reaches 4.8KHz, and PL deviation should be between 0.5 and 1.0 KHz.

Short story since I'm light on time. Remove/don't install R8 into the TinyTrak3. This is used to couple PTT to TX Audio or vice versa. It simulates how HT external microphones used to be wired (2.2Kohms to ground to transmit). Short R5. I tried halving the resistance with another 220Kohm resistor for 105Kohm... Just nix that altogether and short R5. This can be done with a solder bridge, leaving you the option of going back to 220Kohm later if you need to.

I repeat: don't install R8 or remove R8 and short R5.

Short version again: Short R5 on the TinyTrak3 if you need more drive. See page 15 of the TinyTrak3Plus manual: "No audio is heard on a receiver." See page 14 of the TinyTrak3 manual under the same heading. It's the same paragraph verbatim.

Then set the  R6 to maximum. Tell the TinyTrak3 to generate both tones and adjust the IDC pot on the channel element for -/+5.0KHz deviation. Then adjust R6 for 4.0KHz deviation.

(The reason for doing this is that when the IDC (limiter) kicks in, it introduces clipping, distortion, and harmonics to the signal. By setting the channel element for 5KHz, the limiter is high enough to prevent introducing distortion through normal peak clipping. The stock Mitrek microphone circuits contain a brick-wall splatter filter to prevent excessive modulation/sidebands/harmonics.)

This message is my basis for that assumption of 4.0KHz deviation:

Reproduced here:

From: WA8LMF

Michael Crowder wrote:
> Rudy, Thanks for the assistance.
> I've tried re-loading a new config, but I still have the problem that
> the 1200Hz tone is twice the amplitude of the 2200Hz tone. I'm using
> the TinyTrak3 configuration program to cause the MicroTrak8000 to send
> the separate tones, and then looking at the received signal on my PC
> using AGWPacket engine. Anyone have any ideas what may be causing the
> different levels in tones?
> --
> M

This is perfectly normal if you are taking the audio off the speaker
terminal of the receiver.

You are seeing the normal de-emphasis curve of a voice FM receiver. At
the transmit end, audio fed into the mic jack has the high frequencies
boosted relative to the low. Ideally, this boost or pre-emphasis should
be 6dB/octave over the range from 300 to 3000 Hz. The over-the-air
deviation or level of the tone is directly proportional to it's
frequency; i.e. a tone at 2400 Hz should have twice the deviation as one
at 1200 Hz. At the receiver, a corresponding de-emphasis network
(high-frequency cut) cancels out the boost applied at the sending end
for a net flat (equal levels for all audio frequencies) response at the

However, if you apply the TX audio to the 6-pin mini-DIN "data" or
"packet jack" (on radios that have it) instead of the mic jack, no
transmit pre-emphasis is applied -- both tones go out at exactly the
same level. [Note that the MicroTrak integrated TT3 and transmitter has
no mic input with pre-emphasis - the TT is coupled directly to the TX
modulator with net flat response.] At the RX end, you continue to
apply the high-freq roll-off (de-emphasis) [assuming you are taking the
RX audio off the speaker or equivalent] with the result that the
attenuated 2200 Hz high tone is now about HALF the amplitude of the 1200
Hz low tone.

Historically, 1200 baud packet has been a "jam it in the mic jack of any
radio" (with the corresponding pre-emphasis applied) convention, while
9600 baud packet has been coupled directly into the transmit modulator
(with the resulting flat response). At the receive end, 1200 baud has
customarily been taken off the (de-emphasized) speaker output while 9600
receive has to be taken directly off the receiver discriminator (flat
response) before any high-freq de-emphasis is applied.

Traditionally, 1200 baud packet devices have been set to yield about
3.5-4.0 KHz deviation on the high tone which yields about 2.0-2.5 KHz
deviation on the low tone. Devices such as the Kenwood APRS radios and
the Microtraks that transmit "flat" response at 1200 baud are in
conflict with decades of packet convention that says 1200 baud
transmissions should be pre-emphasized.

Many hardware-based TNCs are very intolerant of unequal tone levels,
especially when the high tone is LOWER in level than the low tone. The
AGW Packet Engine and MixW software modems, on the other hand, are quite
tolerant of mis-matched tone levels. The result is that you may be able
to successfully decode your own transmissions with skewed tone levels in
your soft TNC setup, while a digipeater equippped with a TNC2 or KPC3
hardware TNC connected to the (de-emphasiszed) speaker out of it's radio
will not.

The solution is to apply pre-emphasis to the transmitted signal. Place
a small capacitor (probably somewhere between .01 and .005 uf -- you
will have to experiment) in series between the TinyTrak TX audio out and
transmitter audio-in. If the cap is small enough, it will attenuate the
low frequency relative to the high frequency enough to create a net
pre-emphasis effect. In your RX setup with de-emphasis, you should then
observe that the two tone levels are now nearly the same.


Stephen H. Smith wa8lmf (at)
EchoLink Node: 14400 [Think bottom of the 2M band]
Home Page: --OR--

The service monitor verifies these levels, total deviation around 4.0KHz (below the IDC limit) for both tones or the high tone, and around 2.5KHz deviation for the low tone alone.

Another suggestion from Bruce Lane via The Batboard is to set the 40W radio to 15W. This appears to limit current draw to about 6A at 13.8V. Seeing as how Motorola said not less than 55W on the 100W Syntors, I'm guessing it's not a good idea to go any lower on the Mitrek lest the PA become unstable and transmit spurs:

Youbetcha. I've had excellent results from a modified Mitrek. The mods involved were changing two capacitors in the transmitter section (so the multipliers would tune up -- I had the 150.8-174 split radio), and having International Crystal re-do the channel elements for 144.3900.

The only other change I made was to turn down the output power to about 15 watts (40-watt radio). Not only will this make the PA last just about forever, it also cuts down the amount of current needed on transmit.

If you go with a Mitrek, let me know. I can help with a step-by-step.

73 de KC7GR

Repeater-Builder website is a great central resource:

SEITS has some info as well:

SRGClub has a great write up with some images you might need if you're like me and don't have the service manual in front of you.

R909 is TX Power, R927 is TX Limit.

Get a good pair of power cables, start turning the power up until you get to about 2W above rated. Then turn down the operating level. On mine, Limit is the blue pot, Operating Level is the red one.

I shouldn't have done it, but I cranked them both wide open. I got 58.3W out of a 40W rated radio. I didn't leave it that way -- I set it to 15W, then noted that I'm losing about a watt and some change in a RG-174 jumper. The service monitor only reported 13.3W.

The power supply was providing 13.73VDC, and because I was doing a power calibration, I hooked the radio up with a #2 welding cable and some half-inch ground braid I'd recovered and had handy. Losses inside the radio made the voltage at the back of the interface connector at 13.55VDC at 15W, and 13.45VDC at 45W.

[And then I found out the RG-174 jumper was bad and replaced it. Read more below...]

Those pictures are from this excellent write up:

Here's another:

Connector pinouts:

How I mocked up a control head:

Again, I suggest a service manual, but if you don't have one and you've seen one in the past, you might be able to wing it. 73 and good luck...

TinyTrak3 Manuals:

TX Delay Settings:

My actual info on TXD comes from listening to the audio unsquelched and squelched while transmitting packets. The Mitrek is up and on the air in about 60ms. Since it's a crystal-based radio, it's on-channel as well -- no warm up needed. So a TXD of 8 or 80ms is long enough for the radio to send about 20 ms of preambling.

At 9600 BPS, these radios can be backed off to TXD 6 or 7 and used at speed.

The venerable Alinco DR-1200 needed a TXD of 33, mostly due to the broad and sweeping behavior of the VCO. That thing transmits Chirp FM spread-spectrum before it gets on frequency! It's at power before the VCO stabilizes. I'm surprised the FCC let them get away with it.

Most radios should be able to do a TXD of 33. If you listen and notice that the first part of the signal sounds like two alternating notes -- it is. That's the preamble. In practice, you want that to be as short as possible, while still decoding packets correctly. Much of that depends on the other radio returning from transmit to receive quickly. This is why the DIGIs now transmit immediately after hearing a packet. Because if you sent it, you're still waiting to swap back from transmit to receive, and the digi is taking advantage of the pause to blurt the message out.

I set the TinyTrak to a TXD of 18, as shown in the link above. The issue, again, is not the Mitrek, it's every other radio out there. Even with the reed relays, the Mitrek is a fast piece of 1970s technology.

SSID settings:

In my case, I went with KE4AHR-12, ("one-way trackers"). Since it vehicle-borne, it will change to -9 sooner or later, and my bicycle will be -12 or some other SSID.

TinyTrak generic:

03/28/12 Edit:

Per N8DEU, I backed the deviation off to 3.5KHz. I found the RG-174 jumper in the case was bad, and replacing it with another pre-made one caused the power out to jump from 13W to 18W with no change in current. After retuning the Mitrek to 25W (out of the box) the current draw was about 6.25A at 13.6V. The cigarette lighter is fused at 10A, so this gives me a few amps for operating another radio and charging a few mobile devices.

As deviation goes down, power efficiency goes up, however gains are on the order of 3dB, if that. You need at least 6dB to notice the difference.

I use a 1/4-wave antenna for my tracker. Its relatively small for the band used, and gives good performance out to certain angles. My perspective at this time is more coverage-testing of the existing nodes, and the dipole allows me to communicate using knife-edge diffraction. Now if only we'd implemented the system using 9600 BPS instead of 1200 BPS....

On the receive side, I mocked up a control head using a 10K ohm resistor between "Detected Audio" and "Squelch Wiper". A 3.3K ohm resistor connects "Squelch wiper" to audio ground. This gives a fixed squelch response. Then I connected the + side of the upper audio capacitor on the interconnect board to the backside of JU-3C, which connects to L18 -- the third large pin in the Mitrek control cable interface connector. From there, I routed the signal into the Signal Present line on the TinyTrak and set the voltage at about 3V. Reprogram the TinyTrak to invert DCD, and then set the bias so that it operates in a reasonable fashion. Net result is that the reciever works correctly when there is noise at -124dBm, or when there is a signal at -119dBm. The receiver actually starts hearing around -130dBm.

High Power RF Amps

This is a great PDF:

Local copy:

And a good companion to this article:


Thoughts on biasing MOSFETs:

June 5, 2012 edit:

Ludicrous power!


About this Archive

This page is an archive of entries from March 2012 listed from newest to oldest.

February 2012 is the previous archive.

April 2012 is the next archive.

Find recent content on the main index or look in the archives to find all content.