Cornell Dubilier Electronics "Powercon" 12A4 Vibrator Inverter

This inverter was obtained very inexpensively from ebay. Condition was excellent, and like most of the U.S. made inverters I've bought, looked like it had not had much use.
My impressions are that a lot of people bought inverters on impulse for "when we go camping" but never actually did so. One may wonder why I collect U.S. made inverters when their 110-120V output would seem inappropriate for Australia. There's several reasons; 1)I have a small quantity of U.S. made appliances, 2)Vibrator power supplies are something that have interested me for a long time, and, 3)Many modern items fitted with a switchmode power supply are designed for 100-240V operation, so can be used on a 115V inverter such as this.
Vintage inverters are far more common in the U.S than in Australia, so they come up on ebay frequently and usually attract few bids. If vibrator power supplies are of interest, they are something worth collecting.
Input is 12V DC, and output is 115V 60c/s at 340mA. Having output current specified instead of power is unusual. It is equivalent to 39W. Possibly, the designers were thinking of loads with low power factor as a reason for expressing output ratings this way.

As can be seen, this is a very compact unit. The space inside is well used.

Cornell Dubilier made vibrators themselves, and needless to say, that's what's used in the inverter. There are a couple of interesting aspects to the design, however.
One is that the vibrator is wired in. It is supplied with flexible leads which are connected into circuit by means of wire nuts. It would therefore appear that Cornell Dubilier thought vibrator reliability was good. Not having to use a socket does make for more compact construction, and eliminates socket troubles. The output socket location does limit the kind of plugs that can be inserted; e.g. no round shaped plugs, plugpack transformers, etc. It also does not accept polarised or three pin plugs. However, it must be realised that most low power appliances of the era did have a flat non polarised plug.

The vibrator is a wire connected unit.

I did not open the vibrator as it's of the crimped can type. It did not need contact cleaning; operating perfectly straight away. The second interesting aspect of design is the circuit:

Normally, the transformer centre tap is fed from the active battery terminal and the vibrator reed is earthed. However, in this inverter it has been reversed. Electrically, it's the same of course. However, vibrators where the can is connected to the reed are obviously unsuitable. Possibly, earthing of the transformer primary acts as an electrostatic shield and this was the reasoning for doing it this way.
The circuit is conventional, with the usual switch, fuse, and input filter condenser. 10A for the fuse is rather high given the output rating of 39W. A 5A fuse would be more appropriate.
There is an unused tap on the transformer secondary; an orange wire taped up. Why a lower voltage might be required from it is not clear. As is usual, the input is non polarised, but one side of the supply is connected to the chassis so this needs to be taken into account when used in a vehicle or with earthed equipment.
The buffer capacitance consists of a 3uF oil filled paper type, as well as two series connected .47uF's. All up, about 3.2uF. The .47uF's are for output RFI filtering. Both output socket pins are balanced above earth, which lessens the noise output. Sensibly, the vibrator is a series drive type (CDE 6567)
The output waveform is typical.

Loaded with a 240V 40W light bulb, this is the output waveform. Peak voltage is 162, but because of the "dead time", the rms is about 115V.

Performance of this inverter is good, and very noise free. No repairs were required. I left the paper condensers in situ as the can filled oil type is unlikely to leak enough to cause problems, and as the two .47uF's are in series, they are not being exposed to particularly high voltage. I use this inverter to run this Gilfillan 68F off my home solar supply.