(From Wireless World June 1940, page 310) UHF Instruments
TWO new instruments for use in the ultra high frequency field have been introduced by Marconi-Ekco Instruments, Ltd. One is the TF517 signal generator, which is a development of the Model TF390 but with a frequency range from 130 to 300 Mc/s. The output is 0.1 volt max. and the calibration accuracy ± 1 per cent. An attenuator with a range of 100 db is accurate to ± 2 db or 2 uV. Internal modulation is supplied at 400 c/s, 30 per cent.
The other instrument is a diode indicating wavemeter, Type TFG43, with a range of 20 to 300 Mc/s in four bands, with an accuracy of ± 1 per cent.
WWII production or perhaps a little before. Output is from a push pull pair of KT61 driven by L63 phase splitter, so not dissimilar to high quality amplifiers of the 1930s and 40s. Range 0 to 20kc/s in one range. Rectifier (missing) is MU12/14 and oscillators are VR65 (military version of SP61).
I plan to restore this for use both as a signal generator and test amplifier. Though the original KT61s both work, they are both quite tired and not well matched. I'll replace these with something cheaper and more readily available.
To the best of my knowledge BFOs are no longer used as signal generators. As can be seen from this example it's a big and costly device with twice the number of components needed to do much the same by more ordinary means. Most likely the original objectives were stability and accuracy. The BFO achieves accuracy through its unusual ability to go right down to 0Hz and stability through the use of two identical oscillators. The BFO approach did have other important applications in demodulating single side-band and CW (morse) signals and in mine detectors. Mine detectors developed into metal detectors. Another application of the BFO was in an early electronic music instrument - the Theremin.
Update - 26th July 2009
The pair of KT61 have been replaced by home made adaptors (bases from a couple of dead 6V6G) and new 6CH6 pentodes - as these are near enough electrically equivalent.
This is a much later (early 1960s) instrument than the two above. Of course it has an earlier ancestor in the wartime TF 144G.
Once again I found this on Ebay, and again it wasn't too expensive. The big problem with these Marconi instruments is that being so well built, they're extremely heavy - this one weighs 60lbs (27kg). Fortunately it was local, having spent the last part of its working life helping to service radios for Dartmoor Rescue.
Powering it up soon revealed a nasty fault. I expect it was working fine last time it was used but moving it in the boot of my car down our bumpy lane probably did it no favours. Here's what happened - it powered up fine, the signal looked good on the 'scope too. Then when I changed bands everything was fine until I hit band 'E' and no signal, no meter reading, then a little smoke. I quickly switched off and opened up to inspect. Nothing looked bad so I switched on again and saw smoke from R233 soldered to the carrier interrupt switch. I threw the switch and the voltage regulator tube lit up.
I slept on it and in the morning did some tests. It didn't take long to establish that all bands but E were fine but switching to E very nearly completely shorted out the HT supply to the RF oscillator; and I didn't get any meter reading. The lack of a meter reading almost certainly meant the thermocouple heater had burned out.
The fault was traced to a short circuit on the coil turret. The insulation on the wires on the E coil assembly had hardened, cracked and then allowed a short. It was necessary to remove this section of the turret and replace the insulation. Once done it still didn't work! Now the fault was an open circuit! This was traced to the main winding, it had burnt out right at the beginning of the first turn. Fortunately I was able to get about 1mm of bare wire exposed and managed to solder a new lead on, then put some glue on the joint. Aluminium foil was used to protect the windings from the heat. This worked, leaving only the problem of the failed thermocouple heater.
This cell was made in 1956 by Tinsley & Co. According to their web site they were the first to commercially manufacture Weston standard cells in England - back in 1905! Those must have been to an earlier pattern as mine seems to be based on a design patented in 1930 by L J Seitz, Jr.