Bove And Beyond

Pete Putman KT2B 3335 Fieidstone Drive Doyiestown PA 18901

In Praise of 13cm

This month's column goes above, beyond and way out with a Hook at a nifty transverter kit for t he 13 centimeter (2304 MHz) microwave band. It1 s the LMW Electronics 2304TRV1K, manufac* tured in England and imported for the USA by Down East Microwave of Troyr Maine.

LMW manufactures a variety of UHF products, including assembled transverter units for 903. 1296 and 2304 MHzr with power outputs in the 2-6 watt range. The options are endless, including outboard preamplifiers and internal sequencing boards. I elected io build up one of the kits (A) to save a few dollars and (B) to t ry my hand at some microwave construction,

Transverter Schematics

The heart of the LMW transverter ts the Universal Local Oscillator board, or ULO {see Figure D.This «s a stable LO for microwave use that uses relatively few parts, is easy to fire up, and delivers plenty of output.

Both the 1296 and 2304 transverters use the ULO. For 1296 operation, the crystal frequency is 96 MHz, which multiplies 12 times to 1152 MHz f 1152 + 144 = 1296). For 2304 opera-lion. a 90,667 MHz crystal multiplies 12 times to give 1088 MHz. This is doubled to give 2176 MHz (2176 + 144 = 2304) on the transmit and RX mixer boards .

VHF and UHF Operation

Figure 2 is the schematic of the RX mixer Tl, a 2N918 is used as the oscillator powered from an 8-volt regulator. T2 is a 2N5179 which works as a buffer/doubler stage and drives T3. a BFR91 tripler, Another BFR91 works as a doubler and the output is fed to yet another BFR91 Class-A amplifier. The outputs are derived from a 3 section filter at two points, providing both low and high level LO injection,

I bought the ULO, transmrt mixer, receive mixer, IF amplifier and a chassis from Bill Olson of Down East during the PackRats Hamarama in October 1987 Each individual board came in a ZipLoc bag with instructions a schematic and several bags of parts.

Parts only Partly There

The first thing I did was to check the parts against the parts list and found a number of items missing from each board. For example all five tow-level amplifier stages and a mixing diode were missing from the TX mixer. Some capacitors and a mixing diode were missing from the RX mixer as well. On the other hand the ULO kit contained two 2N5179 transistors instead of one which turned out to be a happy accident, as the 2N918 oscillator transistor simply would not work.

Bill said there were problems with kit packing in England and he began packing the kits on this side of the Atlantic. The missing parts showed up promptly from Maine via First Cfass mail, and I set about to work

Bfr91 Manufacturers

Photo A Top view of the ULO board. Note the crystal osctilator to the far left. TX/RX outputs fie on the right.


This took little time to assemble. It shouldn't take more than about 2 hours to wire one and briefly test it. As mentioned earfter not detecting output from the 2N918, I replaced it with Ihe spare 2N5179. and the output shot right up! All stages are easily tunable with a diode probe and VTVM, One construction note: The imerstage coupling capacitors after Q3 are extremely smali and fragile chip capacitors, which cannot withstand much heat. The instructions suggested heating and [inning the PC board pad. allowing solder to flux over the ends to make a secure connection.

The pictorial diagram shows the approximate position of the trimmer capacitors for fun output. Setting them as prescribed, with just a bit of tweaking, yielded just that. I connected each output to my Boonton 92 RF Millivoltmeter for the final alignment^not necessary but very helpful The manual claims 40 mW at point T and 10 mW at point "Xl\ which seems like quite a bit of LO injection at first glance!

There is, however, a good rea-

Local Oscillator Schematic

Figure 1 The Universal Local Oscillator schematic—the heart of the LMW transverter m

t ovTTvts

Figure 1 The Universal Local Oscillator schematic—the heart of the LMW transverter son for it. Both the TX and RX mixer boards double the LO frequency before injection. In the case of the RX mixer, that is accomplished by a pair of Hewlett-Packard HP2817 hot-carrier di-odes-certainly not a scheme with lots of gain, but effective. The TX mixer board also uses two HP2835 diodes as a ring mixer, with an NEC 85637 transistor as the doubler.

i measured nearly 50 mW from the "Y" port and 12 mW from the HX" port, so the manufacturer's specifications were up to snuff The output was also checked with a frequency counter and trimmed to 1080.00 MHz after sitting for 30 minutes. The display still sat on 1080.00 MHz 30 minutes later— very stable indeed-

RX Mixer, IF Amp

Photo B shows both these boards in the completed transverter case. LMW describes the RX board as a "Low Noise Pre-amp. Fitter and Mixer" stage. The design is very simple (Figure 3). it uses just four active components a Hewlett-Packard HXTR3645 bipolar transistor (Ti), an NEC 85637 bipolar device (T2), and Ihe two HP2817 diodes mentioned earlier.

T1 functions as a low-noise (2 dB NF) amplifier stage dnving T2. which also has about a 2 dB noise figure The 2300 MHz signal is then mixed in the jnterdignai filter with the doubled LO signal to provide output at 144 MHz. and not much output at that. Total conversion gain is only on the order of 8 dB with a 4-dB system noise figure, which is the reason for the IF amplifier stage.

This is nothing more than a 2-meter low-noise preamplifier using a BF98t MOSFET to devel op nearly 20 dB gain with about a

1 5 dB noise figure The total sys

1296 Mhz Diode Mixer

Photo B, 2320 PRE RX mixer (left) and 144P IF amplifier (far right) boards in the transverter case. The diode mixer/doubter is to the immediate nght of the RX mixer.

tem conversion gain then works out !o about 28 dB with a 5-6 dB noise figure I found this level tolerable for 2304 weak-signa! work.

Both boards go together in a snap! The bias networks and DC connections are soldered below the RX mixer board, strung in between the vanous feedthrough capacitors, (One note about these feedthroughs: the pfe-dritled holes were too small and had to be enlarged with a hand drill. Be careful to not damage the etched inductors and tuned lines.)

The two transistors should be installed last, since they are extremely smalt and fragile, t tack-soldered one lead on each device to hold (hem into position on the board then gently flux-soldered over the leads to minimize heat problems.

The II1 amplifier board assembly is conventional, but one must take care to protect the gates against static electricity when in-sraliing the BF981 Unplug the soldering iron and make sure the board is grounded, Assembly lime for both boards is two hours The IF amplifier tunes up easily when using a signal source at 144 MHz and a receiver, since maximum gain is close to best noise figure.

The RX mixer board requires a signal source at 2304, which could be the 16th i harmonic of 144 MHz. Dave Mascaro WA3JUF of the PackRats was nice enough to tune my RX board up with a microwave oscillator and spectrum analyzer probe, after which he determined the noise figure to be 4-5 dB. and the conversion gam io be 8 dB. Tnese are close to the specs Incidentally LMW does make an outboard GaAsFET preamplifier using an NEC 72084 device for 2304 MHz. I strongly recommend purchasing it, as it yields about 10 dB gain with about a 1.5-2-dB NF, and makes a big difference >n received signal strength

The IF amplifier input may need retuning once it's connected to the output of the RX mixer board. The ULG input to the RX mixer will almost ceriamly need repeaking. When \ initially tested the RX mixer on the air, I1 couldn t hear a peep from two nearby stations which shoulo have been S-9 plus. A quick check with the Boonton 92 and counter showed insufficient injection of the 1088 MHz signal lo the HP2817 diodes. Repealing VC9-VC11 on the ULO board brought it up another 10 dB which made all the difference as

+1 0

Post a comment