The potential difference between primary and secondary wiring of an isolating transformer

Lets say we have an isolation transformer powering our circuit. What voltage is a multimeter going to show when we put one probe into mains ground, and the other one at the output of the isolated (secondary) transformer wiring.

• I would say: nothing useful. – Marko Buršič Aug 20 '16 at 11:51

What voltage is a multimeter going to show when we put one probe into mains ground, and the other one at the output of the isolated (secondary) transformer wiring.

Theoretically 50% of your mains AC voltage, providing the multimeter has infinite input impedance and the secondary is disconnected from any load.

This value also assumes that there is no grounded shield between windings and that the net electric field is evenly distributed in one layer of windings.

All manner of other subtle issues can raise or lower this figure because it is based on inter-winding capacitance. For instance, because it is likely that the primary and secondary are wound using multiple layers, the main electric field imposing on the secondary will tend to be more towards ground potential or more towards live potential. This is dependent completely on which of the mains wires connects to the layer closest to the secondary. Swapping live and neutral on the primary could easily give different results.

Should this measured voltage be a worry? No because it is sourced through the inter-winding leakage capacitance and this will be in the order of 100 pF to 1 nF (hand waving alert!).

Assuming a real world multimeter (~1MOhm), and a professional grade isolation transformer you will measure zero between either output wire and ground if the output is not bonded to earth/neutral. If it's a typical low cost isolation transformer then there is at minimum an electrostatic shield between input and output windings so the output winding has a capacitance to earth/frame. If it's a high quality isolation transformer the output winding will have top/bottom and side shields (called triple shielded).

The capacitance to frame/ground/earth in a professional grade isolation transformer is likely in the 10-20pf range from output winding to ground and perhaps 1pf or less to the input winding. Here's a diagram of a double shielded transformer, the two shields capacitance is in series: transformer diagram http://electrical-engineering-portal.com/wp-content/uploads/construction-of-shielded-two-winding-transformer.gif

Here's a commercial grade transformer as an example of the input to output isolation that can be achieved: http://www.consulneowatt.com/products/transformers/ultra-isolation

• That example transformer has a "Shield against strong lightening" - wow, it even resists anti-gravity! – Andrew Morton Nov 14 '16 at 19:10

Isolation depends on the ratio of coupling capacitive impedance from primary to secondary and , mutual impedance to ground such as line filters and the quality of the shield to reduce the coupling capacitance inside.

If you were to analyze the transfer impedance, you would find it quite high relative to the load but leakage current may be as much as your load's common mode filter to ground (0.5mA) if that is connected.

If you are looking for RF isolation, then a CM choke of suitable size may be added.

One test you can do is put a large plastic cap across the meter and measure the voltage and if zero, then remove and measure AC current in mA scale to compute the effective impedance coupling of V/I for the frequency being used.

I agree with Andy aka, and also agree with the "nothing useful" comment. BUT, there could be static charges that would add or subtract from the capacitance divider. I sometimes have seen an intentional resistance of several megohms across an isolation barrier, to limit the voltage due to these extraneous charges. The input resistance of the multimeter would accomplish the same thing.