to see Photos of
the Repeater Equipment
in London.
Amplifier
Final Amplifier
---
---
Hamtronics LPA 2-25R (25 Watts)
VoComm VVC100-25RF Continuous-Duty (100 Watts)
Options: DVR Voice ID and Message Recorder
In August 2005, the Hamtronics receivers and transmitter were modified to flatten the frequency response, remove the annoying start-up delay and repair the erratic and unpredictable voter performance. The link receivers and the voter card, which were mounted in a single chassis, were repackaged into separate Vector EFP modules and mounted in a Vector 19" rack assembly. Special attention was given to make the audio as high quality as possible. To have transparent voting performance, it is imperative that all of the receivers have matched levels, frequency response, and linearity. Here is a graph showing the response of all three of our receivers. The local 2 meter receiver, which is not linked, has a similar response. The responses are very closely matched, and any differences are not detectable by ear. It is not noticeable when the voter switches receivers, which may happen several times per second on a fluttering signal.
Also added was a CW identifier to indicate which receivers are voted and how many are "hit". This operation is explained in the FAQ.
to see Photos of
the Repeater Equipment
in Mt Sterling.
The controller (schematic here) actually uses two microcomputer chips - one controls the operation of the system (DTMF tone commands, time-out timers, etc.) and the other one is dedicated to detecting and decoding the CTCSS (PL) sub audible tone. This is my first attempt at writing DSP (Digital Signal Processing) code on a microprocessor. The PIC12F675 microcomputer that I used (it cost $1.33) is not considered a DSP processor, but I am familiar with PIC programming, so I gave it a try. There was enough processing power available to sample the incoming audio 2640 times/second, convert the analog sample to a 10 bit digital word, perform a modified FFT (Fast Fourier Transform) using the Goertzel Algorithm, and detect proper tone presence. It works as good as any commercially available CTCSS decoder that costs 50 times as much!
This controller was the prototype and testbed for development of the firmware that was eventually used at the other two remote receive sites in West Jefferson and Plain City. A printed circuit board was developed for those sites.
Future plans include combining the functions of both microcomputers into just one. This is just a personal goal to see if I can.
The GE Mastr-II transmitter (circa 1972) components were removed from a trunk mounted mobile radio. The final amplifier module and heatsink assembly was fastened to a 19" rack panel plate, and the exciter board was mounted in a Bud Box and also fastened to the panel plate. See photos here. The audio and phase modulator circuits are not used. The crystal oscillator has been modified for direct FM modulation. This provides for a very linear and wide frequency response that is required for proper voting. (Accurate voting depends on the presence of the high frequency white noise to be compared to the other receivers in the system.) A small interface board to key the transmitter by the controller was also installed in the Bud Box. Here is the schematic.
The transmitter easily puts out 20 Watts, but is set to about 10 Watts to be sure it stays cool. I installed two cooling fans on the heatsink assembly, but found that they are unnecessary so they are not used.
The Motorola Micor receiver is considered by many experts as one of the best receivers ever made for repeater duty. The receiver used in our system has been modified for use in the two meter ham band. The helical resonators (5 of them) were rebuilt to be resonant at our operating frequency. (The radio was originally designed for 150.8 to 162 MHz.) The receiver board has just the RF, IF and detector sections only, and is mounted into the 19" rack-mounted Motorla Spectra-TAC chassis for convenience. A modified Motorola Audio-Squelch board is used just for the carrier squelch detection and is fed to the controller card. Un-deemphasized audio directly from the buffered discriminator is also fed to the controller card.
The Power Supply is from a Micor high-power base station, rated at 35 Amps, and is more than adequate for this application. Total load is less than 4 Amps so it never even gets warm. It is a switching type power supply, so efficiency is very high.
to see Photos of
the Repeater Equipment
in Plain City and
West Jefferson.
Hustler G6-144B (W Jefferson) (Spec Sheet - PDF Format)
PC-Tel BMOY-4405 (W Jefferson) (Spec Sheet - PDF Format)
Here is a graph showing the before (Hamtronics Method) and after frequency response. Before the modifications, the frequency response varied more than 15 dB from 100 Hz to 3 KHz and sounded very "tinny". After the changes, the frequency response varies less than 3 dB over an even greater frequency range, and repeated audio sounded just like the input. Similar modifications were performed on the Hamtronics equipment at the transmitter site in London later in 2005.
The UHF link transmitters are only 2 Watts, but that is plenty of power to link back to the London transmitter site over 10 miles away. Link signal quality is very good.
