Repair, modification and alignment work

Work on OH9VHF and OH9SHF Beacons (August 2005):

Return to homepage

- Please Note: many photos on this page are exceptionally unpacked/resized to maintain good resolution for future reference !!!


- an overview photo of the beacons being run-in after overhaul with OH9SHF (ISEP rack) below and OH9VHF (2 high Euro rack) on top

- a diagram of the various interfaces and controls of the two beacons

- 14 years ago, when I built the OH9SHF 1296.945 MHz beacon, I put in a used fan than now subsequently ran it's bearings out

- so I put in a "new" used fan to see if we could achieve a similar 24/7/365 MTBF as before (this one has an integrated thermostat control)

- the output power had also dropped fatally, so a complete lineup was needed. The output power is now back up to the nominal 12 W

- we really could do with a 100 W class PA, so if you have one you can part with, please drop me a mail at oh2aue at sral dot fi

- I also now added a 56 degree Centigrade thermal cut-out switch to disable the PA if cooling is lost

- the beacon had an integrated circulator and 20 W termination built-in as a protective SWR measure, so it is safe (but bad practise) to run the beacon without an antenna

- realignment of the 432.315 MHz ovenized exciter was needed and also the hermetically sealed keying relay was sticky

- upgrading to a semiconductor keying switch was a mere fleeting thought as I had originally put the relay in a IC holder for easy maintenance

- so the relay was promptly replaced with a 14 year old spare part stocked specifically for this purpose :-)

- the 2716 (yes !) EPROM keyer is housed with the heavily modified DC0DA exciter board in a diecast aluminium box

- after alignment, the exciter output was back to the nominal output level of +26 dBm (50 ohm)

- also, the frequency was calibrated during this process and slightly tuned lower to take aging into account

- realignment of the BGY22 driven BAY96 varactor tripler took place too supplying the nominal 300 mW RF output on 23 cm - I just love these old reliable parts ;-)

- photo of the exciter and varactor tripler hooked up for final output power and spectrum checks

- reference service meter readings for nominal operation: SWR and output RF power


- OH9VHF was originally built some 15 years ago by OH2BYW (of the OH2TI group) and indended to actually become OH6VHF, replacing some really old hardware

- this beacon was upgraded from the older BLY90 45 W PA (down to 38 W after very careful initial alignment) to a more modern 80 W power amplifier

- this dual device amplifier can easily do 135 W, is specifed for 80 W continuous, but as we are amateur, we will settle for 100 W to the antenna cable...

- I decided to maintain the new amplifier (kindly arranged by OH3BK of the OH3TR clan) in original form to ease replacement in case of failure

- unfortunately the original mechanics did not fit the new 2 high Eurorack, so it had to come apart

- the keyer, oscillator + oven and the following buffer boards were removed and appropriate modifications made

- these three boards I then reintegrated into an old Mobira NMT900 prototype diecast enclosure (bottom view) that is very easily accessible but is still extremely RF tight

- here is a topside view of the module with the 2N6080 driver and 2N4427 predriver at the left, the ovenized LO in the center and the EPROM keyer to the right

- the 2N6080 has a 4 dB power attenuator at it's output serving two purposes: damping down the drive to the PA module and protecting the device from lack of loading

- looking at the bottom (front panel, without covers) edge, from left to right, you have:
- the power level control remained unimplemented (OM OH2BYW had originally designed a very stable 0, -10, -20 and -30 dB step concept into the beacon)

- also, the frequency was badly off after many years of unmanned 7/24/365 service, so the ovenized crystal was tempted back to nominal maintaining reliable starting

- power to the new beacon design is supplied via the 230 Vac available and is first switched down to -48 V (the Efore FSR6 was kindly donated by the OH2Z gang)

- the FSR6 is designed to operate as battery charger also - an option that may be added later

- the nominal voltage is now set for -48.0 V and the nominal current is 5.0 A for a solid 100 W RF output from the beacon

- this -48 V is then in turn switched to a highly protected 14.2 V regulated high current supply

- not everything goes this smoothly and checking the output spectrum (the original POCSAG BS has an integrated ~cubic meter diplexer) of the PA was a terrifying experience...

- the first three harmonics were at disastrous levels of -30 dBc, -40 dBc and -55 dBc respectively - somthing definitely needed to be done

- playing around with air dielectric coils and air dielectric trimmers showed unacceptable losses for this power class

- checking my stocks I decided to test some aluminium-backed Polyflon substrate and after some coarse calculations, came up with some approximate numbers

- taking out the Finnish equivalent of an Exacto Knife I came up with this

- then, using some hobby material that is actually steel wire, but really and truly silver plated (no idea of the thickness though) donated by OH3UW, turned the coils

- this low-pass filter has excellent characteristics and could cope with considerably more than 100 W with it's 0.15 dB insertion loss

- of course the rejection band attenuation is very high too, all the way up to a couple of GHz, but the darn thing just did not fit in the 21 mm left in the rack :-(

- after seriously comtemplating rectangular form (yes - truly !),  I decided to reorientate the coils, accepting a couple of tradeoffs - power handling and rejection

- the voltage handling is not so good of course and also the out or band rejection creaps up due to capasitive coupling over the coils, but nonetheless...

- ... the three first harmonics are now all at -75 dBc or more - plenty of headroom to keep any authority happy :-)

- mounting the low pass filter in a PCB material shielded box (without cover), it was ready for installation onto a 21 mm front panel donated to the project by OH2LH

- both beacons have now seen some 200 hours of operation on my bench and are ready for return to Pirttikoski, KP36OI

Spectrum Plots:

- please note, that the graticule plots and actual sweep plots do no overlay 100 % exactly !

- harmonic components of the OH9VHF beacon at nominal 100 W output level after the harmonics filter

- calibrated and measured levels of the OH9VHF output signal (100 W carrier output) harmonics with levels of approximately -75 dBc

- close-in spectrum of the OH9VHF carrier (spikes are from the two SMPS switch mode power supplies), levels approximately -75 dBc

- the OH9SHF beacon 432.315 MHz exciter output spectrum before buffering and varactor tripling (+26 dBm output @ 50 ohm), 200 MHz/div, 10 dB/div

- OH9SHF RF output after attenuator (1296.945 MHz, +41 dBm output @ 50 ohm), harmonic level better than - 70 dBc (other harmonics not meaasurable)

- OH9SHF close-in carrier purity with 5 MHz span (500 kHz/div) using a linearly regulated laboratory power supply (an SMPS is used in Pirttikoski)

Return to main page
Updated 05.09.2005