[MrFixr55]

That sounds strange and almost physically impossible.

Ohm between the primary and that secondary, and check voltage between one leg of the primary (preferably neutral and each leg of the secondary in question. Resistance between primary and secondary should be infinite. Voltage between the primary and this or any secondary should be zero (a slight resistance, such as 10 - 100K Ohm may have to be placed in parallel with the meter leads if a DVM is being used. The 5V and 6.3V windings should both have very low resistance as they are short windings with rather thick wire, making it physically impossible for that winding to put out 107V.

Is this the original transformer?

If you find a short between the primary and secondary of this transformer, check the wiring. You may have accidentally put the primary and secondary in series, in a "buck" mode.

Several companies, such as Arvin and RCA were not above using a transformer with a high voltage secondary to power a series filament string/ on some models. However, this is rare.

When Philco (likely the first as they were a major manufacturer of car radios) and others started using 6Volt tubes for AC powered radios (6V tubes were originally designed for car radios an the filaments were rated for 6V even though they served as heaters for a cylindrical cathode), they followed the practice of having a grounded center tap on the filament (heater) secondary winding of the transformer. Either this was found to be unnecessary at the time or improvements in tube design rendered it unnecessary to have a grounded center tap. Therefore the center tap was eliminated and one leg of the 6V winding was grounded.

Several transformers of this vintage including from the 80 and 81, as well as from the 37-84 should work, as well as others of the vintage. The 42 (6F6 draws more filament current than the 41 (6Kg but this is a 4 tube radio. A transformer form a 5 tube radio should fit.

Interestingly, both the 84 and 37-84 list the same part number for the transformer (32-7180). However, while the schematic for the 84 shows the center tap for the 6V winding to be connected to the center tap of the B Supply winding (and technically not at chassis ground potential, the 37-84 (same set, only uses Octal tubes) does not show a center tap for this secondary but has one leg at chassis ground.

This radio is an interesting set. It is a direct descendent of the 80 Junior, a radio that caused almost as much of a stir as the Philco 20 (which started the "Cathedral" radio craze). The 80 was designed in the throes of the great Depression to be cheap but work. It was actually a superheterodyne with a36 tetrode "autodyne" convertor (1st detector) but no IF stage, a 36 grid leak detector with adjustable regenerative feedback, as a second detector, no first AF amp, and a 42 pentode output stage. The closest RCA competitor, the R28 (and others) Radiolette was a superior constructed Superhet, using a 58 tuned RF amp, a 2A7 (just released) Pentagrid Convertor, no IF Amp, a 58 biased "power" detector and 2A5 output stage. the Philco 80's purpose was as a "loss leader" to get people in the store where the sales rep would attempt to "upsell" the Customer to a more expensive Philco radio. However they sold thousands of these, starting at $19.95, and eventually dropping the price to $14.95. This caused RCA to sell its R28 "Radiolette" almost at a loss, and they still could not compete with Philco. Between these radios, the 20, the 70, etc., Philco became the #1 selling brand. They updated the 80 with the 81 (Pentode 1st and 2nd detectors and "Police Band", then your 84 (same as the 81 but no police band), then the 37-84 (same as 84 but with Octal tubes).

Having serviced an 80 and owning an RCA R28, I was very impressed with the 80's performance and tone.