Wurth SMPS Transformer
Switch Mode Power Supply Transformers (SMPS)
Worked Example of Suitable Tests
Overview of SMPS Transformers
The emergence of Switch Mode Power Supplies, and the drive for smaller, cheaper power conversion has meant that the transformers that lie at the heart of any SMPS are under increasing demand to be cheap and quick to produce and test whilst still remaining safe and reliable.
The power semiconductors used to switch the supply into the transformer, and the IC’s to control the switching frequency have both dropped in price and increased in reliability and performance over the past 20 years. This still leaves the transformer to provide two basic, and historically conflicting, functions.
First, as the isolation barrier from the supply to the user, it must be proven to isolate the supply under large potential differences.
Secondly, it must also have close coupling of the windings (i.e. low leakage inductance) to keep losses to a minimum and hence keep efficiency up.
Wurth Electronics manufactures a variety of good examples of SMPS transformers, for a range of different types of SMPS.
Here we will look at the 750811290, a transformer designed for flyback SMPS configurations.
SMPS Suggested testing
SMPS - AT Editor Schematic
The part is represented using the AT Editor schematic to the left.
The centre-tapped primary is automatically recognised and drawn when a winding is added to the schematic with a pin number that was already defined (in this case pin 3)
As we wish to core performance under 3.15 Amps DC bias, the fixture also connects a DC1000 to pins 4 and 2. This means that any HI POT tests we deploy will have to use pins 2,3,4 as the LO terminals (see HPAC test later)
SMPS - AT Fixturing
As the transformer has standard PCB mounting pins, it is suited for fixturing using Kelvin Pins. These grip horizontally on each pin, so a clamp is not needed to keep the transformer in place.
Kelvin pins allow for quick fitment of the UUT, as well as giving us optimum accuracy needed for the measurements, as the effect of pin contact resistance and any fixture wiring can be automatically compensated out of any results.
The picture shows the part fitted to a custom 12 pin kevin fixture.
The Fixture also has 2 x 4mm sockets for connecting the Voltech DC1000 DC bias source. These are wired in the fixture to pins 2 and 4 of the socket, and are then used in the program for test 10 to measure LSBX.
SMPS - AT Fixturing, with additional 4mm sockets for DC100 connection for the DC bias LSBX test at step 10
SMPS - AT Test Program
The program first checks the individual coil resistance of each winding to check they are below a specified maximum.
This is followed by four turns ratio checks to confirm the correct number of turns, phasing and general transformer operation at 10 Khz
The inductance of the primary is then checked at 10Kz, followed by the same test but with 3.15A applied (using a Voltech DC1000) to confirm the core does not saturate.
The correct core gapping and winding placement is confirmed using a Leakage Inductance test.
Finally the isolation is confirmed using a HPAC test at 4.5kV AC, 50Hz for 1 second.
Pins and Conditions
|1||R||DC resistance||Pin 2-4, test for 600 mOhms +/- 10%||To check the winding resistance is below a maximum. Also acts as a check of correct wire gauge and good termination.|
|2||R||DC resistance||Pin 6-5, test for 110 mOhms +/- 10%||To check the winding resistance is below a maximum. Also acts as a check of correct wire gauge and good termination.|
|3||R||DC resistance||Pin 8-10, test for 570 mOhms +/- 10%||To check the winding resistance is below a maximum. Also acts as a check of correct wire gauge and good termination.|
|4||R||DC resistance||Pin 9-11, test for 460 mOhms +/- 10%|
|5||TR||Turns Ratio||Energise pins 4-3,0.1 V 10 kHz. Check turns ratio 4-3:3-2 to be 1:1 -/+ 6%||To check correct ratio of windings from each side of the primary centre tap|
|6||TR||Turns Ratio||Energise pins 4-2,0.1 V 10 kHz. Check turns ratio 4-2:9-11 to be 1:1 -/+ 2%||To check correct ratio of windings from all of primary to one of the secondaries|
|7||TR||Turns Ratio||Energise pins 4-2,0.1 V 10 kHz. Check turns ratio 4-2:8-10 to be 1:1 -/+ 2%||To check correct ratio of windings from all of primary to the other secondary|
|8||TR||Turns Ratio||Energise pins 4-2,0.1 V 10 kHz. Check turns ratio 4-2:6-5 to be 6:1 -/+ 2%||To check correct ratio of windings from all of primary to feedback winding|
|9||LS||Inductance||Energise Pins 4-2, 0.1 V, 10 kHz, measure inductance to be 461 uH +/- 10%||Check core material and assembly accuracy|
|10||LSBX||Inductance with DC Bias||Energise Pins 4-2, 0.1 V, 10 kHz, apply 3.15 A DC. Check inductance to be >368 uH||Check core does not saturate under DC. Hence allows you to prove on every part that L drop under Bias is not greater than published 20%|
|11||LL||Leakage Inductance||Energise pins 4-2, 0.1 V, 10 kHz. Check leakage inductance to all other coils is less than 12 uH||Checks coupling of coils is good to minimise leakage|
|12||HPAC||AC Hi-Pot||4.5 kV AC, 1 second, Pins 8,9,10,11 Hi, Pins 2,3,4,5,6 Lo. Check current <5 mA||To check isolation as per datasheet. Note that the primary is kept LO as this has the DC1000 attached. See DC1000 user manual for best practice with HI POT using a DC1000 simultaneously.|
|AT5600 Run time 4.01 sec|
|(AT3600 Run time 8.51 sec)|
LSBX test Inductance under DC BIAS.
In this example the datasheet specifies the need to check inductance under bias as it is a flyback transformer. Other types such as push-pull or forward converters would not have a test need for LSBX so the program would be simpler.
The factors governing L response under DC current are the number of turns, the core material and the core air-gap chosen. As these factors are already checked by the LS and TR tests, some customers may chose to only check LSBX at the design stage (see our DC1000 page on design testing using a DC1000 with any LCR meter), or occasionally sample testing using the AT5600 Audit Test function.(see our Audit Test page).
However, some users may wish to retain the LSBX test on 100% of parts, due to component use (e.g. Military/ Medical)
AT Test Results for SMPS