The AM Forum

We all know that the ARC-5 (Command Set TX) especially the BC457-459's were used as VFO's for all kinds of AM, CW and even SSB rigs. I have been stareing at the schematic since I was Johnny Novice in the 1970's.

I have redrawn the schematic of the oscillator portion for your edification. It is nothing more than a standard grounded plate Hartley oscillator using a triode tube as you can see. Something is wierd here though. The engineers did something strange with the filament.

Questions:

Why did they tie the filament to the cathode?

Why did they lift the fillament off ground?

Why did they not do this with chokes?

rather than the elaborate tertiary winding on the main VFO tank and why didn't this mess up the stability?


Hint: Think big Airplane at 20000 feet in heavy flak.


Any bright ideas guys? And no, I do not think I know why but I have some ideas.

Mike WU2D
 

I think it has something to do with raising the filament to the same rf potential as the cathode, so that the cathode-filament capacitance doesn't affect the oscillator.  It may also reduce FM squeal(not hum) from the filaments (I think they ran off 400~ a.c.).

I once thought the filament current would heat up the coil and make the oscillator drift, but when I rewired the oscillator to feed the filaments directly, I recall it didn't work as well, but it was so long ago I don't remember what was wrong.  I rewired it back to original.  I used an 80m version of the ARC-5 xmtr as a vfo.  Even cut the chassis down to include just the osc section and not the final.  It was reasonably stable but not perfect by any means.

                         

Mike I don't know about the B17 or B27 but the 28 volt bus on airplanes that I have played with are very dirty with all kinds of high frequency switching noise.
115 VAC 400hz bus is also very dirty. I would think chokes on the heater inputs would isolate common mode noise from being coupled from heater to cathode C modulating the VFO. This floating system may have made the oscillator design easier????? BTW chassis ground is also pretty dirty. At 20000 feet the skin of the aircraft is just a big return lead. 
BTW modern aircraft made of composite has soem real issues with EMI and of course lightning.

I seem to recall owning a bunch of these in my Novice days... nothing I recall says that these were particularly stable or good VFOs at all! They were otoh, cheap, plentiful and cheap and plentiful.

             _-_-WBear2GCR

The military airplanes of WWII had 28 VDC electrical systems.  The 110/400hz systems were a somewhat later development (late 50's IIRC).

Most early 50's airborne radios still had DC dynamotors (the ARR41, ART13, etc).  I have a manual for a R-45A which is a similar-appearing but completely different radio from the R-45 - has miniature toobs - which lists a 110VAC/400 Hz power supply.  This radio is most likely mid-50's vintage.
 

28 volt bus could easily have 2 to 3 volts of high frequency noise riding on it plus 600 volt spikes every time a relay opens.
dyno brush noise must have added to the fun.

Yup, there was most likely plenty of electrical noise on a B-17 with 4 generators and 4 make-break regulators and a ton 'o' sparkplugs a-firin.

John,
I'm pretty sure the P&W Wasp had shielded spark plug leads. (early stealth)

Ok we have a few theorys. I quizzed several hams and engineers at work too. Here is the rundown so far.

1. The 24 VDC may be quite dirty in the aircraft. Forcing the filament and the cathode to be together could reduce hum or noise modulation. 
 
2. The intense vibration in the aircraft could cause the filament to vibrate in respect to the cathode; causing FMing.
 
3. The Filament could actually short to the cathode - this connection would allow the circuit to possibly continue operating.
 
4. Could running DC through the tank possibly stabilize the tank at low temperatures?
 
5. The circuit may have not had enough gain at some part of the band or bands and this connection may have improved electrical performance somehow.

I want to take some baseline data.
I have two unmodified command transmitters, one is a BC-696 3-4 Mhz and the other is a T-19/ARC5 3-4 MHz Navy. I will do a warm up test on each of them. Keying them every 5 minutes over two hours to take a frequency counter reading. This should give us an idea of how stable they are in ambient conditions.

Mike WU2D

Hey Mike,
Get yourself a copy of MIL STD 461E and look at CS114 and CS115 tests to see how stuff is tested for this kind of threat.
There are also some ground injection tests out there but I have not run them often may also be a CS test or part of RTCA DO 160E  You may also have to go back to 461A. I seem to remember 1 volt of RF on the ground (calibrated at 50 ohms) for 461A

It is not often that I get an answer from the horses mouth but it happened - I got an E-Mail from Steve Williams, who has reproduced the entire radio suite of a B-29 in stock condition. aafradio.org

He quoted this right out of a design manual for the radio.

 Section IV, Theory of Operation, Paragraph 4-42,
Master Oscillator Circuit.

"T53B is a twin winding on the master oscillator coil,
wound from ground to the cathode tap in order to
maintain the heater at the same r-f potential as the
cathode, so that variations in cathode-to-heater
capacitance within the tube will not affect the
frequency of oscillation."


There she is - Mike WU2D

P.S. I'm still going to test the two radios and will report the results.

ALL military stuff made in those days and for many years after had completely shielded ignition systems. If it had a gasoline engine, it had fully shielded ignition, no ifs, ands or buts. Including coaxial spark plug wires and plugs. I have an engine in my collection from an aircraft APU and even the magneto is sealed in a metal can.
                                 The Slab Bacon                                                                                       

I have taken VFO frequency stability data on 5 command set transmitter units (actually 4 but one was recapped and got a new tube and was re-run).

The testing was over a 2 hour period. One unit was 100% stock with banging relays. Two were the Handbook conversion which keys only the finals (osc on the whole time) and one was with the osc on only during the tx with the amplifier keyed. All transmitters were on 75 Meters except for unit # 4 which was a 20 Meter Conversion which doubles from a 40M tank.

All were set to 15 Watts Output Power.

After a couple minutes of heater warm up, the initial frequency was recorded, then data were taken every couple of minutes over the first hour and less frequently the second hour. 15, 30,  60, 90 and 120 minute data is shown. The drift between test points is called intervals. Interval 0 for instance, is the initial 15 minute warm up drift in Hz.

Power supplies were unregulated but the heaters were DC (12 or 24 VDC depending on the unit).

Natural convection in ambient 27C with all covers intact - no fans.

The final test was a key down old buzzard test which had a 6 minute duration. This was done after the 2 Hour warm up.

Here is the data.

Mike WU2D

Old EMI tests had you set up a 28 volt relay as a buzzer and wrap the leads of the relay around the harness of the UUT. The buzzing contacts would generate 600 volt spikes on the relay leads. When this was wrapped around a test harness it would induce about 150 volt spikes on the harness wires. This is a very effective test to see how noise effects performance. This test was later replaced by the more controlled CS115. This would be interesting to see how the oscillator handled crud.