Just for discussion, in older breaker type systems both windings in the coil formed a stepup transformer. One side of both windings was connected together internally and brought to the negative pole. Negative and positive was written on the cap of the coil to aid with installation.
When the charge in a coil collapses it creates a large back-EMF, which creates a high voltage pulse on the grounded side of the high voltage secondary winding. This is not harmful to a points breaker system, because you have two strong metal contacts to ditch the high voltage spike through.
In an electronic system the 'low' side of the coil is switched through a type of transistor, which can be damaged if exposed to such high voltages (although some have internal protection), so some electronic ignition systems don't use a 'commoned coil' approach, but instead both the primary (low tension) and secondary (HT) windings are left isolated and do not connect together anywhere. The switching transistor in these circuits doesn't need to ditch the HT back-EMF pulse, only the pulse generated by the collapsing field in the primary (easily shunted through a diode)
In this latter case, both coils are floating, just as in a conventional isolated transformer. The primary is switched on and off, and the collapsing field induces a much higher voltage in the isolated secondary.
I think this type of circuit suits a limited number of spark arrangements, because the diagram I saw shows that on the 3VD, one spark plug is connected to one side of the secondary, and the other secondary connection goes to the other spark plug, meaning that the secondary circuit consists of a secondary winding, each end connected to a single spark plug cap, and the circuit completes through ground at the spark plugs. This is different to a points based coil system.
The quick way to check is to measure the resistance of the secondary winding at the coil. If the coil has commoned windings there would be no resistance between the secondary terminals (assuming a good coil), so measuring between both spark plug lead outputs at the coil should read close to zero Ohms. But if the coil is the type which uses windings which are not commoned together, a resistance check will reveal the resistance of the high turn secondary coil, which will probably be a few Ohms.
A confirmation check is to measure the resistance between the high voltage output terminals and each low voltage primary winding input terminals, in a fully isolated (non polarised) system there will be no continuity between either input terminal and either output terminal.
Where polarity might matter in the use case of a fully isolated coil,but possibly not enough for the manufacturers to bother to put terminal polarities on the coil's case like they always used to, is that it has been found that with a hot sparkplug a higher voltage is achieved if the spark polarity is negative at the hot centre of the plug relative to the grounded outer case. (some claims state upto a 15% higher voltage). The statements I have read about it were along the lines that it made little difference really, but maybe at higher revs there seemed to be more power.
I cleared out some 3VD parts recently and wish I had kept a coil for testing now !
EDIT, this is a common circuit for some twin cylinder ignition circuits (this from a 2CV...pah !) showing what is sometimes called a 'double ended coil' , with the exception of the points system, the 3VD wiring diagram I have seen shows the coil used in an identical fashion to how it is used here.......which then leads me to question the truth about a negative pulse being better, because in these systems one plug is always positive and the other is always negative, and in that case the polarity of the primary low voltage side is irrelevant https://en.wikipedia...llumage_2cv.svg
Edited by fixitsan, 08 November 2020 - 11:28 am.
Sorted and on the Road Mick 
