Self Resonant Frequency continued
PI Spock,
Thank you for the information. I should elaborate. We have made a dozen or so groups of transformers to Power Integration specifications, and have tested them. For Inductance (for DI-143, ours measured 644 microhenries), leakage inductance - 14.5 microhenries(which is a confirmation of the proper insulation between windings), the expected DCR resistance we meet EVERY spec. However, when we measure the Self Resonant frequency (700KHZ), it is too low. This has been true for four of the designs compared to the PI specification. Mind you, we are not a garage type operation, we are in the business of making transformers. Is it possible that the calculated Self Resonant frequency is missing another factor that explains why our built transformers would not be meeting the calculated specification? Using the data provided, and your calculation, what should the Self Resonant frequency of our sample be?
Our hope is to clarify this issue, as we intend to offer numerous transformers for Power Integrations chips. If there is something that we are missing in construction, it is imperative we discover it.
Sincerely,
Tj Klotz
Comments
Hello PI-Spock,
On August 10th, 2009 we received the final transformer specifications for DI-143, which does not include a layer of tape between the first and second layer of each Primary winding. Furthermore, this transformer specification has half of the total Primary wound, then the bias winding, followed by the Secondary winding, a shield, and completed with the second half of the Primary winding.
If we had the latitude to deviate from the Power Integrations specs, there may be a few options to increase the Self Resonant Frequency.
Our concern is more for the 1MHZ minimum resonant frequency specification using the construction instructions provided by PI and still being a valid design for the DI-143 application. Is the Self Resonant Frequency a critical parameter for a valid transformer design? (all other specs are met) Any idea what the sample built by PI for this application MEASURED for Self Resonant Frequency?
Thank you,
Tj Klotz
The leakage inductance is within spec. What is more important now is the no-load power consumption of the supply. If this is within your limits then the resonant frequency does not matter so much. 700 kHz is sufficiently high for the transformer.
BTW there is a way to specify tape between primary layers through the software. On the primary and bias dialog there is a check-box (below the bias section) which allows you to add a layer of tape between 2 primary layers.
Do you have a layer of tape between the 2 primary layers? If the resonant frequency is on the lower side and the leakage is also low, this points to high intra winiding capacitance. This can be lowered by adding a layter of tape between 2 primary layers.