7.1.11 Turns Ratio
The AT3600 offer two basic alternative ways to confirm that the transformer has been assembled properly, with the appropriate number of primary and secondary turns.
Turns ratio is the preferred test for signal, pulse and switched mode power transformers, where the normal operating conditions require only small excursions of the B-H curve, never extending beyond the linear regions.
(For line frequency transformers, designed to operate over the full extent of the B-H curve, including the non-linear regions, the preferred method is to use an open - circuit voltage test to check for the correct numbers of turns on each winding.)
Clearly a turns ratio test cannot tell you the actual number of turns on a winding, only the ratio between one winding and the next. You should therefore have at least one inductance test in your program to give confidence that the absolute number of turns is correct as well as the ratio.
To measure turns ratio, a test source voltage is applied to one winding, the energised winding, and the voltages across two other windings (one of which may also be the energised winding) are measured using harmonic analysis. The turns ratio is measured by dividing one measured voltage by the other, and making a compensation for the effects of winding resistance.
It is recommended that you choose the winding with the highest number of turns as the one to be energised. A possible exception to this rule is when you wish to measure the ratio between two windings which should be accurately matched at 1:1. In this case it may be better to energise a third winding with a lower number of turns, to ensure that any measurement errors apply equally to the two windings under test.
You can specify the signal to be applied to the energised winding to have a frequency in the range 20Hz to 1MHz, and an amplitude from 1mV to 5V.
The recommended test conditions depend on the inductance of the energised winding; they are given in the table below assuming that the energised winding is the one with the highest number of turns:
|Inductance of the
|Preferred test signal
|100nH → 1uH
1uH → 10uH
10uH → 100uH
100uH → 1mH
1mH → 10mH
10mH → 100mH
100mH → 1H
1H → 10H
10H → 100H
100H → 1KH
1KH → 10KH
The Test Conditions for Turns Ratio Measurement
Notes: The signal is applied to the primary winding, or the winding which has the largest number of turns.
However, if by doing this, the expected voltage on the winding with the smallest number of turns falls below 1mV, then the test voltage should be increased.
This may also require an increase in the test frequency so that the current taken by the driven winding does not become too large, but in general this frequency increase should be kept as small as possible to avoid problems caused by stray capacitance at high frequencies.
WHERE MATCHING IN GROUPS IS IMPORTANT:
In some transformer designs, the turns ratio between a primary winding and a secondary winding is not as important as the ratio between different primaries or different secondaries.
To make the most accurate measurements in such cases apply the test signal to the primary winding and measure the turns-ratio from primary to one of the secondaries.
Then, leaving the primary energised as above, measure the turns ratio between the secondaries.
Next, energise a secondary winding (possibly at a different voltage and / or frequency depending on its inductance) and measure the turns ratio between the various primaries.
In this way windings which should be matched are treated equally during the test.
SPECIFYING THE TEST LIMITS
When specifying turns ratio tests, it is preferable to avoid limits which are unnecessarily tight, and which may therefore lead to measurement difficulties.
For example, if two equal secondary windings should have 10 turns each, the ratio should be 1:1. One turn in error would produce a ratio error of 10% or -10% (i.e. 11:10 or 10:11), and therefore limits of +5% and -5% would be suitable to detect the error.