Duplexer Overhaul for 7075
For more than 25 years, first analog then P25, the 147.075 MHz (+600 kHz split) repeater has covered the north of Wellington, New Zealand. The original machine was a home made crystal controlled 10W design by Peter ZL2ARW but the site now features a 100W capable Motorola Quantar.
With the increase in usage due to the growing interest in P25, the migration from the prototype P25 network based around an AstroTAC comparator in Hawaii, to the new system known as Quantar Site Connect linking Australia, Germany, and the USA to NZ, the shortcomings of the old Wacom WP 639 duplexer have become apparent. Poor receive sensitivity being the main complaint partly due to the temporary antenna in use and partly to increased duplexer insertion loss combined with lack of isolation.
Recently the old duplexer was pulled and a refurbishment session was held, attending were Peter ZL3TC, Ray ZL2RAY, Jeff ZL2JG, and John ZL4JY. At the same time another couple of duplexers appeared out of ZL4JY's secret stock for cleaning and evaluation, with the intention of choosing the best performing duplexer for 7075. Of the candidates the pale blue (WP639) and yellow (DB4060) duplexers were completely stripped and cleaned. The DB4060 was kindly donated by Mike ZL1BTB many years ago. The and low pass tuning rod assemblies were swapped so that the rods would be moved from their previous settings so that the finger stock would sit on different part of the inner rod. The high - low side swap fortunately gave slightly more exposed thread on one of the invar threaded rods that had been broken on the WP639 at some point in the past while a fractured UHF T piece was found and replaced.
The black duplexer is a WP629, designed for the 118 to 144 MHz band. In most respects it is identical the WP639 and the original datasheet is here:
Both use 125mm (5") cavities, 725mm (29") long for the low band unit and 625mm (25") long for the high band. The reject frequency adjustment stubs are the same length on the outside but the inner section is about 50mm (2") shorter on the high frequency unit. Lastly the RG-214 (Wacom 'modified' RG214) sections are 25mm (1") shorter for the higher band.
To tune the WP629 to 2m the RG214 sections were swapped for the shorter length.
After cleaning the WP629 and WP639 duplexers returned typically 1.5 to 1.8 dB pass insertion loss and typically 75 dB notches. The pale blue WP639 was slightly better in terms of insertion loss than the black WP629 (0.1 to 0.2 dB) and had silver plated tuning rods compared to plain copper on the WP629.
All tuning was done using an Agilent E5701A vector network analyser.
But the winner on the day after cleaning was the yellow DB4060 duplexer with the larger cavities, 6 dB deeper notches, and 0.2 dB less insertion loss due to the larger 200mm (8") cavities:
Ray starts work on the DB4060
Close up of the DB4060 tuning plunger
Further details on this device can be found here:
The article by N1BUG was very helpful, we removed all the old lubricant from the DB4060 and it was not replaced. It was easier to hacksaw the knobs from the threaded rod and then slot the rods for screwdriver tuning rather than replace the knobs.
Duplexer rack with new DB4060 in place
Next major work item at the site will be the rigging of the the new VHF and UHF antennas.
- John, ZL4JY