Beginners question about trafo VA rating
- Thread starter jimk04
- Start date
- Status
No, you must parallel the secondaries to use the full 300VA, assuming you only want the voltage given by each secondary.
Be careful with matching the phase - use an AC voltmeter with the secondaries in series to determine this. First check each secondary and verify the expected voltage. Then series them. If you measure zero volts ac across the two free ends, then you can join those free ends together for use in parallel. If you measure 2 x the voltage of a single secondary, swap the leads on one secondary and check again for zero volts AC before you parallel them.
Edit: you also get "all the VA" if you use the secondaries separately or in series.
Be careful with matching the phase - use an AC voltmeter with the secondaries in series to determine this. First check each secondary and verify the expected voltage. Then series them. If you measure zero volts ac across the two free ends, then you can join those free ends together for use in parallel. If you measure 2 x the voltage of a single secondary, swap the leads on one secondary and check again for zero volts AC before you parallel them.
Edit: you also get "all the VA" if you use the secondaries separately or in series.
Last edited:
To ubergeeknz's good reply I will add that connecting in (simple) parallel, two even similar windings may still leave cross-currents between the two windings. If the secondary windings are wound bifilarly (the two wires are put on the core/coil-former at the same time and next to one another) simple parallel connection is OK. As you most likely do not know how the transformer is produced, I will suggest you to put individual rectifiers for each winding and join them after the rectification. Then, one winding cannot send energy into the other winding.
If there is sufficient difference between the two secondary windings, though they are supposed to give the same voltage, the difference will cause cross-currents that may work as a partly short-circuit of the transformer. Then you may have the problems you describe. Rectification will always cause a voltage drop but when you use separate rectifiers for each winding, the voltage drop will be a little less. Most important is that the separate rectifiers prevent cross-currents and even slightly different windings will share the load current quite well due to the resistive voltage drop.
Yes, the rectifier outputs are joined before the main filter caps.
Yes, the rectifier outputs are joined before the main filter caps.
With modern, industrially produced transformers, paralleling is a non-issue: the number of turns is exact, especially for low-turns, low-voltage windings and even if the windings see different fluxes due to slight geometric asymmetries, it is of no consequence: the difference means that a significant leakage inductance exists between the windings, and this inductance acts as a ballast to equalize currents.
To summarize: if the windings are bifilar, paralleling is OK because a different number of turns is impossible, and if they are wound separately (and have the same number of turns), paralleling is OK too because they will naturally balance.
To be sure, just measure the voltage between the terminals you are about to connect after you have joined the first pair: if you read <100mV for a low voltage winding, it is OK.
If you still have doubts, make the final connection through an AC ammeter: if the current is smaller than 5% of the rated secondary current, then everything is fine.
The current balance under load will be much better, and will be perfectly sufficient.
All of the above applies for a single transformer with two windings: paralleling two separate transformers requires more attention (but remains possible)
To summarize: if the windings are bifilar, paralleling is OK because a different number of turns is impossible, and if they are wound separately (and have the same number of turns), paralleling is OK too because they will naturally balance.
To be sure, just measure the voltage between the terminals you are about to connect after you have joined the first pair: if you read <100mV for a low voltage winding, it is OK.
If you still have doubts, make the final connection through an AC ammeter: if the current is smaller than 5% of the rated secondary current, then everything is fine.
The current balance under load will be much better, and will be perfectly sufficient.
All of the above applies for a single transformer with two windings: paralleling two separate transformers requires more attention (but remains possible)
If you have 100VA transformer with 2 identical secondaries you can pull 50VA from each, but if you only use one you can theoretically pull about 70VA since then the heating of the windings from I-squared-R losses is the same. However this assumes concentrating heat in one secondary is no worse than spreading it over two secondaries, which is an optimistic and flawed assumption.
However the primary heating is less at 70VA than at 100VA, which helps.
Say the two secondaries are 25V 2A, with 1 ohm resistance, the I-squared-R losses in each at 50W are 2x2x1 = 4W per winding. With one running at 2.8A its 2.8x2.8x1 = 8W in one.
However if you are not sure about how well the windings can lose their heat you shouldn't overdrive a single winding this much - at the end of the day the maximum temperature inside each winding is what counts, and this can only be checked by measuring or detailed thermal modelling. For instance on a toroidal each secondary might be a single layer winding at/near the surface and thermally they are well coupled to each other and ambient.
On an EI core the inner secondary might be the limiting factor thermally, and you might not be able to figure out which is innermost.
So ultimately follow the datasheet, or be prepared to experiment. Its fairly easy to measure a winding's average temperature by measuring the resistance change after reaching thermal equilibrium, but its a time-consuming process compared with buying a bigger transformer!
However the primary heating is less at 70VA than at 100VA, which helps.
Say the two secondaries are 25V 2A, with 1 ohm resistance, the I-squared-R losses in each at 50W are 2x2x1 = 4W per winding. With one running at 2.8A its 2.8x2.8x1 = 8W in one.
However if you are not sure about how well the windings can lose their heat you shouldn't overdrive a single winding this much - at the end of the day the maximum temperature inside each winding is what counts, and this can only be checked by measuring or detailed thermal modelling. For instance on a toroidal each secondary might be a single layer winding at/near the surface and thermally they are well coupled to each other and ambient.
On an EI core the inner secondary might be the limiting factor thermally, and you might not be able to figure out which is innermost.
So ultimately follow the datasheet, or be prepared to experiment. Its fairly easy to measure a winding's average temperature by measuring the resistance change after reaching thermal equilibrium, but its a time-consuming process compared with buying a bigger transformer!
- Status