This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

SN6501: Dual isolated 12V and 5V output - 5V or 12V input ?

Part Number: SN6501
Other Parts Discussed in Thread: ISOW7821, SN6505B, , SN6505A, ISO7721

Hello,

Hope someone could help with this topic and some questions below.

In our application, we have Vin = 11-16 V, a linear regulator which makes the 5V, and a ISOW7821 which makes the isolated 5V.

Since now we have to add an isolated 12V, we would like also to replace the ISOW7821 (because too expensive). I would use a SN650x because of its circuit simplicity and the small size of the package. The configuration I was thinking is the one like this, with Vin = 5V and two transformer in parallel driven together, to make a 5V with a reinforced isolation through the Transformer 750313734 and the isolated 12V (11 it’s also okay) through a 760390012 Transformer with a bridge rectifier configuration.

  1. Can the system work correctly with current consumption of max 20 mA @12Vout and max 75mA @5Vout ?
  2. If yes, how the SN650x can prevent saturation of both transformers?
  3. Can be used SN6501 or it’s better to use SN6505A or SN6505B?
  4. Are there alternative / better /cheaper transformers?
  5. Would you prefer 12Vout with 4 diodes voltage doubling configuration and a transformer ratio of 1.1:1 or 1.3:1 or a standard 2 diodes config and a transformer ratio of around 2.5:1 ?
  6. Could you suggest a different but still smart/small solution with a Vin = 11-16 V ? maybe also replacing the linear 5V regulator (5V non-isolated max current around 150mA) ?

Thank you in advance for what you can answer

  • Hi Alberto,

    Welcome to TI E2E forum and thank you for your interest in TI isolated products.

    Since you are isolating both 5V and 12V, it does make sense to replace ISOW7821 with ISO7721 and use SN650x based solution to generate the isolated 5V for ISO7721. Please see my inputs below to all the questions you have listed out,

    1. Yes, since SN650x is an open-loop power supply solution, we do expect it to work fine when multiple transformers are being driven by just one SN650x device. We have a few other customers also do the same.
    2. The device doesn't have any dedicated feature to prevent saturation and hence, this needs to be taken care by choosing the right transformer for given requirements.
    3. SN6501 can be used to achieve this but using SN6505B gives you more margin for both input and output currents. Although SN6501 is available in 5-pin SOT-23 package and SN6505B in 6-pin SOT-23, you could always use SN6501 in SN6505B footprint and still have everything work just fine. Hence, it maybe worth going with SN6505B to start with.
    4. SN650x based solution is one of the simplest and cheapest external transformer based power supply solutions, I don't have anything else to suggest from our product group.
    5. Both the configurations should be fine, although you might find transformers easily for the former than the latter.
    6. We currently do not have a device to support 11-16V direct input hence, the solutions discussed above are the best that I can think of.

    Please do also note that since SN650x based solution is an open-loop solution, you might need an LDO at the output if you need better output voltage regulation. Let me know if you have any other questions, thanks.

    Regards,
    Koteshwar Rao

  • Hi Koteshwar,

    Thank you for your fast feedback, please be so kind to evaluate my further questions regarding point 2) and core magnetisation.

    Consider I use SN6505B which has V-t_min=7,6 us and then I chose the transformers above mentioned which have V-t = 11 us. Basing on datasheet requirement selection they are the right transformers to work with SN6505B, correct me if I am wrong. Should I consider other parameters?

    However, the core magnetisation during each switching cycle depends on Ton-Q1 and Ton-Q2 inside the SN6505x and I cannot control them. Also, Ton-Q1 and Ton-Q2 are ideally the same, but in reality they cannot be exactly equal. Hence, an offset from an imbalance in flux density swing can be generated. What does guarantee that this offset is restored and does not increase in time until the saturation of the core?

    If then I chose to replace SN6505B with SN6501 which has V-t_min = 9.1 us, is it not too close to the V-t transformers value of 11 us ?

    Thank you in advance,

    Alberto Carlotti

  • Hi Alberto,

    I am Manuel, a colleague of Koteshwar's.

    The transformers mentioned in the original post are compatible with both SN6501 and SN6505B, even in voltage-doubling configurations. The V-t product calculation example and equation in the datasheet includes 10% margin for power supply tolerances, which is sufficient for most cases. Both of these transformers are also listed in the Recommended Isolation Transformers... table of the SN650x datasheets.

    Although D1/D2 switching times are matched as closely as possible by design, magnetic flux imbalances due to imperfections like differing trace impedances are addressed in the Core Magnetization section of the SN6501 datasheet. When there is a small flux imbalance in the push-pull transformer, the resistance of the output switch sinking higher power will increase (due to heat from the current rise), reducing voltage proportionally at the primary windings of the transformer and keep voltage across both windings even. In conditions where loads are extremely unbalanced, transformers usually saturate and either recover with time once balance is restored or require a power cycle.

    We recommend using SN6501 with transformers that have V-t product >9.1Vus for 5V input operation, including the parts mentioned above (750313734 and 760390012). This V-t product margin is enough for operation under recommended conditions.


    Thank you,
    Manuel Chavez

  • Hi Manuel, 

    thank you very much to for the clear explanation, this clarified my doubts. 

    Thanks again also to Koteshwar.

    Kind regards,

    Alberto Carlotti