A R R O W   R E P E A T E R S
     The Smoke and Mirrors behind Repeaters

      W8KJ   Kevin J. Martin  146.97  ( 97 -vs- 96 / MARC 20Khz bandplan)
      N8BJX  Rick Green       224.38
      W8MNT  Blake Arnold     443.5
      WD8OBL Hugh Aller       439.25 ATV  (Line "A")
      W8MNT  Blake Arnold     902    ATV
      WD8DPA Mark Traveglini  Packet Digi on 220 MHz
        Dabbled in:
           6 meter
           10 meter (w/UHF crosslink)  --  Dan K2PUT putting back on the air
           Weather RADAR (beacon using Bell-208 modems)
	   ...much of which began on WB8UPM's (SK) tower

     BTW - Each had their own TRUSTEE
           Each had a seat on the Board (which outnumbered the elected BoD)
           Hence why we had a "Trustee Delegate"
	   Now we use a CLUB callsign & TD has been deprecated

     BTW - The "ARROW" used to stand for:
             "Amateur Radio Repeaters Of Washtenaw"
	     ...a spinoff from the HVARA Huron Valley Amateur Radio Assn.

Various Repeaters are either built out of the:
     "Junk Box"
     Are basically "Junk"
     Are "So-So" 
     Are "Rock Solid"

 - No shielding
 - No sub-band Pre-Selector
 - Does NOT do well around other repeaters atop same building
 - Interfered with AAPD (when local Osc was left running)

WHAT WE HAVE NOW: -------------------------- 146.96 -600 W8RP Bridgecom BCR-50 224.38 -1.6 (WAAM) W8PGW Bridgecom BCR-220 443.5 +5 W8RP Motorola ??? AREDN W8RP Ubiquiti/MikroTik SO-SO - P.A. burnout issues - No sub-band Pre-Selector + But can swap units if a TX or RX fails
IN THE OLDEN DAYS (1965-70): ----------------------------------- ALL repeaters originally output on 146.94 MHz (one per community) o Each had their own INPUT (146.34, .28, .37, etc.) o Hence being known as: 34/94, 36/94, 28/94, 37/94... o Later additional "Pairs" were coordinated (13/73, 28/88, 37/97) with distinct inputs 600 KHz away from the output Repeaters were *specifically* licensed - WR8ADU Repeaters MUST have a "Control Operator" at all times!!! (We had a Control Operator SKED, or the machine was turned OFF) Repeater had to LOG all transmissions (Cassette tapes by the hundreds!!!) Repeaters MUST have a "control phone line" $$$ or a Command Rcvr (LM567 Tone Decoder for DTMF from Radio Shack) NOTE: After the 20KHz Bandplan change, ALL 2-meter repeaters in Michigan have only TWO digits after the decimal point. And there are NO duplications across 145, 146, 147 MHz. 15 145.15 A2 16 147.16 Detroit 23 145.23 UofM 45 145.45 Chelsea 68 146.68 Livingston 76 146.76 Detroit 88 146.88 Jackson 92 146.92 Ypsi 96 146.96 A2 Monroe 146.13/73 moved to 12/72 "I-94" 146.94 was too close to Toledo 94, so moved to 92 MARC requires 150 mile separation for co-channel operation (Don't know if there was a Pig butchered for a BBQ celebration) Why do we have repeaters? What do they do for us? --------------------------------------------------------- **** The *WHOLE* idea is to have antennas HIGH IN THE AIR (period!) "ELEVATION IS EVERYTHING" -- THAT is how a repeater gives good coverage in the LOS bands -- But we are rarely LOS, so we depend upon 'reflections' -- MULTIPATH - Some relections arrive at the repeater 180 degrees out of phase because of different path lengths o Move your car 1/2 wavelength to go from noisy to solid xxx Show Multipath on whiteboard xxx The BASIC Repeater: -- In 1970's we BUILT repeaters from scratch (hence why we REQUIRED control phone lines) Nothing was store bought / Plug-N-Play / Did not built it with checkbook CONTROLLER was hard-wired LOGIC and/or RELAYS o Dana Whitlow K8YUM built 2m controller discrete IC's o CPU's didn't exist - affordably... xxx Show old 73 hardware xxx JUNK BOX
-- NO, we do NOT just hook a RX and a TX back to back -- First we need a COR "Carrier Operated Relay" (COS) o How do we know when a user is transmitting? - But let's start with how does SQUELCH work? TERMS: o White noise (atmospheric thermal noise) o Pink noise (cyclic clock) o "Full Quieting" on FM QUIETING and DEVIATION ---------------------- o Plus and Minus 5 KHz Deviation (Plus and Minus 2.5 KHz Deviation is NOT used in Amateur Radio) o Low audio is *generally* related to low Deviation o "Full Quieting" is a signal that FULLY overcomes the thermal noise and fully "quiets" the front-end of the Receiver - Hiss - Bacon Frying - Different from "Picket Fencing" (an effect of Multipath on mobiles) - "Clipping", when deviated beyond the bandpass of the Receivers I.F. o "Permakay" filters o IDC - Instantaneous Deviation Control (Motorola)
xxx Open SQ on HT's in classroom xxx Sample audio from DISCRIMINATOR Audio Hi-Pass filter - high freq component of white noise Charge a DC switch - set bias w/SQUELCH control xxx SHOW SQ-1000 Squelch/COR Controller xxx So now we can key up the TX ...but wait...there's more! We need a HANG TIMER (3 to 6 seconds) (smack TX up/down) We need a TOT (3 minutes typical) FAILSAFE SHUTDOWN 1) TOT 2) Control Phone Line -or- over air command 3) Internet to the "PDU" And we should have a COURTESY BEEP (1 to 2 seconds into HANG) o Why a Courtesy Beep??? Also indicates reset of the TOT BTW - The repeater transmitter MUST handle 100% duty cycle !!! ...but wait...there's more! We need to ID the transmitter - Every 10 minutes (Part 97) (doesn't have to be at beginning) - Usually done in CW, but can be done in Voice (digital "in mode") - No faster than 20wpm (but slower is nice for some) - But when? o Rules say WITHIN first 10 mins (220 rptr) o We prefer ID on first keyup o DMR / D-Star / Packet Digi done "In Mode" ...but wait...there's more! So how DOES the audio get from RX to TX? - 3db rolloff per octave called PRE-EMPHASUS and DE-EMPHASUS for better FM modulation (not needed nor desired in RPTR) - Too tinny and the "old timers" can't hear it - Too bassy and the the audio sounds too muffled especially in noisy environments such as in cars -- Hamtronics audio sucked! Tom Brey WB8COX designed Op-Amp audio shaping (FSA and COX each wired them up for 2m & 70cm) o Take audio from DISCRIMINATOR and pass to VARACTOR with some audio processing. Must also inject Courtesy Tone and ID into audio stream. -- Motorola MSR2000 audio is conditioned for best performance for First Responders (noise, sirens, fire engine pumps, etc.) -- We want "crisp audio" -- Fully deviated +-5KHz without clipping xxx SHOW ASSORTED CONTROLLERS xxx o We saw home-brew... (used on old 73) o ID-O-Matic -- configured using serial port and a text-based menu o Pacific Research RI-1 (used on old 96) -- configured using over-the-air DTMF o Pacific Research RI-300 (used on 74) -- configured using Windows98 and S-cmd macros o MCC RC-1000 (used on old 92) o CAT-1000 (used on 145.15 and 146.72) o Arcom RC-210 (used on 92 -and- future 74 and 96/220) -- configured with DTMF or "RCP" Windows software (also used on Chelsea 45, Manchester 98) Arcom RC-210 controls THREE repeaters at once !!! o CW and Voice ID's o Canned messages/announcements at set dates/times o Over the air Programming (fewer trips to the site) o Can link any of the ports (2-meter with 220 during Net) o Status reports: 11 (full status -hourly) & 14 (short status) o Time (55) o DTMF pad test (85#####) o Voice quality playback (15) o Temperature/Voltage/state-status (31, 32, 33, 50, 51) o Control external relays, etc. +++ I use the state status (Door, Grid) to change the Courtesy Tone
VOTERS: o Monolithic (one Receiver - one Transmitter) o Multiple Receiver sites (UHF links) o A "Voter" samples which receiver has the lowset white noise ============================================================ What about the RF? What about Interference between TX and RX? !!!!!!!!!!!!!! -- Front end saturation o Worse w/today's WIDEBAND hear-all receivers Commercial RPTRS have Tuned Front End before 1st Mixer Helical Coil front end (1st RF not swamped with other RF energy) o No wiggle room to detect the desired signal o This is called "desense" because we have DESENSITIZED the receivers ability to discern a weak signal in the presence of a strong signal o Want PRE-SELECTORS in Receiver (Motorola)
BASICALLY... Cavities to notch out TX energy - "NEAR FIELD" energy swamps front end of Receiver - NOTCH out 146.96 from reaching the 36 receiver - PASS only 146.36 to the Receiver *** It's a whole presentation unto itself *** - >80 db ISOLATION is desired (82 db for a 25 Watt TX) - Usually about 1.5 to 3 db INSERTION LOSS on the RX cans
We CAN share both Transmit and Receive on ONE antenna !!! - PASS the TX freq on the TX string of cans - NOTCH the TX freq on the RX string of cans - Jumpers between cans are 1/8 or 1/4 of the *opposite* freq What about *other* transmitters on the site getting into the RX? Rooftops and Towers are covered with TRANSMITTERS Paging, 154.25 fire dispatch, WTWR 98.3, WQKL 107.1 Use "notch" CAVITIES (a.k.a. "cans") to filter out undesired signals - Cavities to "pass" only the receive frequency (all other signals are outside the "passband") - $$ EXPENSIVE $$ high-Q filter -- $300 146.36 pass filter - $$ CHEAP $$ 1/4-wave shorted stubb -- $3 solution RX----------T--------DUP | | 1/4 wave at Broadcast freq (consider Velocity Factor 0.62) | |--- shorting Push-Pin | - PASS the TX freq on the TX string of cans - NOTCH the TX freq on the RX string of cans - Jumpers between cans are 1/8 or 1/4 of the *opposite* freq PROBLEMS: o De-sense / Front End overload o Mixing products / Every diode/rust junction is a MIXER o IMD (RF mixing products interfering with input) o Site noise / Motors - Lights (CFL at Chelsea) o Line loss o Insertion loss o LAND LORDS!!! Tower Plaza $400/month per repeater ($1200/mo) WTWR gave access in exchange for property maint. WVMO gave access in exchange for Weather RTTY !!! Don't abuse the relationship !!! COORDINATION BODIES: ============================ M.A.R.C. -- Michigan Area Repeater Council o Coordinates for lowest interference o Coord 6-meters up through 1.2 GHz o Band plan for LINKS and Packet o Keeps record of UHF links between repeaters o 2-meter 150-mile separation o Work with neighboring State coordination bodies o QUADRANTS around the State o PL coordinated by Quadrants o We submit a TDS Report every 2 years to maintain our coord o 2M and 70cm are FULL!!! 220 *may* have openings + Uncoordinated like 5.15 get *NO* protections + FCC backs the Repeater Councils above uncoord. REPEATER PERFORMANCE COMPARISSON ==================================== Rptr Power Dup Loss Coax Loss Ant Site ASL Ant AAT ERP @ ASL ---- ----- -------- ------------------ ---- -------- ------- ------------ 96 25 ~10 W 100' LDF4 -1.2db 6db 866' 150' 63 W @ 1016' 74 30 -6 W 200' LDF4 -2.4db 6db 1030' 180' 72 W @ 1210' 92 ~20 -10 W 60' LDF4 -0.6db 6db 850' 170' ~40 W @ 1020' 15 100W -zero- 400' 1-5/8" -0.8db 8db 935' 350' ~600 W @ 1285' +++ RF Transmit "FOOTPRINT" and Receive "FOOTPRINT" TEST STANDARD: A mobile with a 3db gain (5/8 wave) mag-mount at 25 Watts o EQUAL TX and RX Footprint o *should* hear the Repeater the *same* as it is being heard by the Repeater --- end ---