Last year Meg and I returned to the UK for Christmas, bringing along 47GHz and 76GHz transverters to try and work a UK station who had previously built transverters for these bands but had noone to work.  A QSO was had with ease across the garage on 47088 MHz but no signals were identified either way on 75976MHz, so an opportunity for a UK first was lost. After hearing that the 76GHz UK first had been completed, I turned my attention to the next band up, 142GHz.

 Whilst roaming the flea market at Dayton I happened to stumble upon a a surplus Hughes 47449H-1002 G band harmonic mixer "condition unknown" with chipped paint for a modest price making it worth a gamble. This would expand my spectrum analysis capabilities to cover 145GHz, so my enthusiasm to construct for the band exploded. Realising we were due go on vacation in Crete in mid June 2000, stopping off in England on the way, I realised it would be an ideal opportunity to try for a "first". As insurance I would need two transverters. The problem was I only had 4 weeks to build them! The good news was that on testing the mixer seemed to work fine, despite it's tatty appearance.

The only published equipment design available was the DB6NT one in DUBUS, which required drive at 11GHz then had multipliers using four FETS to generate 35GHz which then drove the shm diode mixer with a 144MHz IF.  At Microwave update 99 I had obtained some surplus multipliers which took an sma input of 7dBm at 12GHz and produced +17dBm output in 36GHz in wr28 and ran off +8V. This module could be used to drive the mixer directly saving me space and complexity so  I would only have to mill a small box to contain the mixer PCB. The incomplete drawings in DUBUS were taken and modified to accomodate the changes. As I wanted to use an unmodified IC202 to drive the transverter the DB6NT IF pcb could not be used, so I employed a 1.5"x1.5"x3.5" diecast box with the modified circuit built dead bug style on the bottom of it, on a piece of double sided unetched PCB. The diodes were mounted on the board with fine silver loaded epoxy, applied with the point of a needle whilst viewing under a microscope. I found it useful to practice trying to mount a small piece of wire on a scrap pcb to perfect the technique before finally trying on the diode. After mounting the boards were baked at 150F for 30 minutes to cure the adhesive. The picture shows the DB6NT transverter in it's 8.5"x4.5"x3" diecast box. The G4DDK004 is at the top with the WDG009 multiplier at the bottom. The mixer assembly is the aluminium box centre right, sitting on top of the IF Interface

The next decision was what frequency on 142GHz? A search of my stocks of 50 degree C specification crystals produced a pair of 99MHz crystals originally bought for a 24192/432MHz transverter which could be used in the DDK004/WDG009 oscillator/multiplier I intended to use. The DDK004 produces 2376MHz at 7dBm which drives the retuned WDG009 multiplier to +14dBm at 11880MHz. The commercial multiplier operating below its design frequency, produces 35640MHz at only +14dBm but this was considered enough to drive the mixer diodes  When mixed with 144MHz an output of 142704GHz is produced. The assembly was hooked up to the analyser and after about 30 mins of problem resolution, centred around power supplies it produced around 20uW of RF, assuming I am achieving the conversion loss quoted by hughes for the harmonic mixer (40dB) was correct.

 I was starting to run short of time, and would not be able to build a second DB6NT transverter so the possibilities of using the Hughes mixer as the second transverter was considered. The DDK004/WDG009 chain from my 76GHz transverter was recrystalled and retuned to produce +18dBm at 11879.5MHz and fed into the mixer via an isolator for stability. Another Identical 144.5MHz IF interface was built in a diecast box, the whole unit was mounted in a recycled 6"x10"x4" diecast box (it used to contain the G4BPO/P 23cm mhp 15 years ago!) The difference between the two IF frequencies was deliberately chosen to stop the IF radios working each other.

Initial tests were done with 30dB horns using WR7 from Eisch electronics. The transverters were first set up at opposite ends of the Texas garden. SSB Signals were good and after letting the LOs warm up for 30 minutes the frequency stability was excellent. After this initial success one of the transverters was put into beacon mode placed on the shack window aimed at a local line of sight landmark about 500m away. I then drove to the site positioning the car for maximum signal and good signals were finally obtained with enough signal in hand to suggest 1km was feasable. I also took this opportunity to swap the receive horn for my Procomm 25cm dish; signals improved by around 15dB, which is around what was expected.

On the eve of departure the 2 transverters were packed in bubble wrap along with a pair of IC202, a Garmin GPS unit and a box of cables and put in the suitcase. This equipment, even without the Procomm dish, totally eliminated taking any scuba equipment to Crete! Luckily DFW security did not find the equipment  too suspicious when the x-rayed it. Leaving Dallas in a cloud burst, which delayed departure for 1 hour, I was pleased to arrive in hot and sunny England, it looked ideal 142GHz weather! The lower power Hughes transverter was set up on the window ledge of the G7FRE shack pointing down the garden where the DB6NT transverter was set up, powered from the mobile. Signals were just as loud as over the same distance in the USA,  showing the rigs had not suffered in their travels.

I then drove off to a previously identified overpass on the A2 to the west of G7FRE. Talkback was maintained on 433.55FM using a pair of VX1R. Signals were considerable weaker at the new G4FRE/P location and peaked up with the horn slightly off boresight but at 1812UTC 52 given 53 received reports were obtained. From GPS the distance was 1.29km, both locations were within JO01BX. There was probably enough signal in hand to extend the distance, but by this time the effects of jet lag were showing and a hasty retreat ensued. The rigs subsequently returned to the USA with me on June 19, it still worked on arrival!

Since my return another oscillator/multiplier chain has been built to replace the I stole from my 76GHz Transverter. If space permits I hope to bring both 76GHz xverters to the UK for Christmas 2000; having completed a QSO with them over tens of km in the ARRL January 2000 contest in flat Texas the UK possibilities are intriguing.