In the works - a 222 medium power ampifier using a Russian GI-7B tube

                    - a GS35B amp, maybe for 6 meters, possibly HF

                    - a weak signal source for 432, 1296, 2304, 3456 and 5760

                    - a 1 watt 2304 amplifier with 26 dB of gain

                    - low noise preamps for 902, 2304 and 3456

                    - a 2 watt 3456 amplifier

                    - a 1 watt 5760 amplifier

                    - a legal limt + differential T tuner for HF (started)

                    - stripline VHF/UHF directional coupler (started)

                    - a 10 MHz reference system locked to GPS

                    - a thermal watt meter (built but needs testing)

                     - a 10 GHz transverter

Pretty ambitious, huh?

  

2304 transverter kit

This kit is a W1GHZ design, uses a 720 MHz LO board and the transverter board uses copper pipe cap filters, 2 on the LO chain, one common one for TX and RX and another one dedicated to the TX. LO drive is +7dBm, 144 MHz IF drive is 0 dBm and TX output is approximately +7 dBm. I used the simple RF detector board to tune the filters.

The 3456 transverter looks identical, with the exception of a couple of components and tuning. These kits are a breeze to asssemble, but leave a bit to desire in the perfomance department as they are, a good low noise preamp and external TX amplifier will be needed to do any QSO's much past line of sight.

  

RF microve detector

I aquired a set of low power transverters for 2.3, 3.4 and 5.7 GHz that still needed some work and wanted a simple way to check relative output. Wally, W9OBG, sent me a couple of home brew detectors that would not measure some of the lower levels I was wanting to look at, so I do some checking and purchased a few Agilent HSMS 2852, a zero bias detector diode. The package is a dual device SOT-23 pacakge, pretty small to work with. I created a slimple PC board that gave me room for the SMA input, a 50 ohm load, the SOT-23 diode package and a cap and resisitor. The larger board is one using simple hot carrier diodes. With 1 milliwatt at 50 MHz into the smaller detector using the zero bias diode, I get 220 mV DC out. The one using the larger hot carrier diodes only has 61 mV DC out. At levels much below 1 milliwatt, the hot carrier diodes do not produce any voltage, the 2852 produces a detectable voltage to well below -10 dBm. The circuit is very simple and could be built on a connector. The center pin of the connector has the 50 ohm load resistor to ground, one diode goes to ground and one in series with the output, a small cap, I used a 220 pF to filter the ripple and a small resistor to ground, i think I used a 68 K ohm. 

  

  

2 meter amp conversion project

A shot of my old 2 meter amplifier (W0KT called it a telephone booth due to the cabinet it was in), converted from a Quintron QT-8505 PA. 650+ watts out with about 5 watts of drive. I normally use a 4 dB pad before the driver board. I can probably remove the driver board and use a full 25 watts of drive and still get full power out. The original design used a preamp/IPA board and the exciter only provided 200 milliwatts of power. Power supply requirements are 28 VDC at 70 amps and sits below the PA chassis. The fans draw air in the front of the power supply and blows up the fins on the back of the PA chassis. The heat sink is so massive I never bother to run the fans (too noisy as well). A very simple conversion and with the paging market all but gone, it is possible to find used transmitters for a reasonable price.

The RF devices are all MRF174's. Note the simple 3-way spitter and the complicated 3- way combiner! Later versions of htis PA used a splitter that was much like the combiner, it offers better isolation between PA board inouts. The 200 watt PA boards are 2 MRF174's in a push/pull confuguration. These devices are a FET and act a little like a tube. Bias them to about 250 mA of current per device, somewhere around 3.5 VDC. The bias switches negative during standby. The lower section is the low pass filter and the directional coupler. The amp is fully metered and the cast chassis is an immense heatsink! I'll add a photo of the control and meter compartment later.

  

© 2017 NG9R

Bio
Projects
Created with the QTH.com SiteBuilder.