SN6501 transformer driver vs build-in ISO1176T driver

Hi Henrik,

There is nothing schematically different between the SN6501 and the additional circuitry on the ISO1176T. With that being said the layout is more optimized in the SN6501 and therefore the on state resistances are slightly optimized.

The reason that we were recommend using the 1:2.2 turns ratio for the ISO1176T is so that it can be more generically used with any LDO and there is enough room for drop out. If you are using an external LDO and the drop is low enough for a 1:2.0 turns ratio, then there should not be any issues.  

Thanks,

John

Henrik,

since there is no current limiting neither  in the ISO1176T nor in the SN6501, I usually recommend a low-cost LDO with internal current limiting at the output. However, some of our customers use the ISO1176T without LDO and a 1: 2.0 transformer.

The unregulated output is simply buffered with a 1uF cap. While this voltage usually droops with increasing load current, the change in load current for isolated RS-485 applications does not change significantly because the common-mode is removed. Hence operations without LDO and the corresponding space and cost saving is possible.

Best regards, Thomas

Hi Thomas,

Thank you for you reply.

1) Our intention is to use a unified design for RS-485, CAN and Profibus isolated ports which can share most of the BOM. TI's parts seem an ideal candidate for such an approach, however the different transformer ratios recommended for SN6501 (1:2) and ISO1176T/ISO3036T (1:2.2) questioned that approach a bit. So I asked in this forum, hoping I can use the same 1:2 Midcom transformer for both SN6501 based interfaces and ISOxxxT based ones.

Since I wrote the original post, I did more research on the ratio calculations and came to the conclusion that TI's calculation for the SN6501/760390015 of 1:2 is very optimistic and probably even contradicting. 1:2 assumes a Schottky forward voltage of 0.2V @ 100mA while in your own SN6501 datasheet you specify this as 0.275V and most MBR0520L datasheets even specify rather 0.3V. Also in your engineering video "Engineer It - How to isolate a power supply for RS485 & CAN" Swaroop Vaidyanath assumes 10% accuracy for the 3.3V primary supply but then uses 2% in the calculation to achieve the 1:2. 

Following your published formula in the SN6501 datasheet and using the proper max values from the datasheets the required ratio looks very different with 1:2.18 and make the Midcom 760390015 not an appropriate choice any more. However you promote the SN6105 and the Midcom 760390015 as matching parts and ideal partners. 

All very confusing and giving not much confidence in the TI recommended solution.

2) Regards your recommendation to use an LDO, yes our intention is to use an LDO as we have to guarantee a high level of interoperability with different equipment. So standard compliance is a requirement and this includes correct voltage levels for RS-485 or CAN. I also would be worried to use a 1:2 transformer without LDO as the voltage would definitely be outside the 5.5V maximum operating voltage of the ISO1176T. Not sure how others deal with this, but we tend to design within the datasheet's operating conditions.

Kind Regards

Henrik