ISOM8610: ISOM8610 Isolation Performance

Alex Gold

Part Number: ISOM8610

Our design uses a competitor part which is tested, analysed and working but going obsolete. A potential replacement is ISOM8610. It is the only suitable replacement that appears available on the market due to various specification requirements. Functionally, the device behaviour is completely appropriate but the isolation may be a problem and the datasheet is unclear.

For our requirement, we must be able to test isolation withstand for 1400Vdc (or 1000Vrms) for 60s on each production item. This would be seen across the ISOM8610. I have read 8.2.4 in the datasheet and I need to be certain - is the graph implying that the device will withstand >7000Vrms when tested for 60s? And that the 500Vrms or 707Vpk is only if we need the voltage applied over a period of 20/30 years?

Thank you.

24 days ago

0 Eric Hackett 20 days ago

TI__Guru* 75972 points

Hey Alex,

Thanks for bringing these questions to E2E.To directly address your questions:

I have read 8.2.4 in the datasheet and I need to be certain - is the graph implying that the device will withstand >7000Vrms when tested for 60s?
- Not exactly, the graph in the datasheet is not a withstand voltage plot, and it's not meant to be interpreted s a specification for withstand voltage. It is a statistical projection based on lifetime testing, so it is a long-term continuous operating rating, not withstand test limit plot.
And that the 500Vrms or 707Vpk is only if we need the voltage applied over a period of 20/30 years?
- Yes, the 500Vrms is the maximum isolation working voltage, the continuous voltage the device's barrier is meant to sustain reliably over the product's lifetime.

For what you need, the table in the datasheet in section 5.6 is more useful for you. The UL1577 test and conditions sound similar to what you are trying to find, and this device is well over your limit of 1400Vdc (1000Vrms), rated at 3750Vrms withstand for 60 seconds.

I hope that all makes sense, and please let us know if you have any other questions.

Regards,

Eric Hackett

0 Alex Gold 20 days ago

Prodigy 10 points

Hi Eric,

Thank you for the reply.

This is sort of as expected; the component appears to be fine for use in this application. However, the issue this leaves me with is that in the Absolute Maximum Ratings table, the limits for t=60s for AC and DC respectively are 707Vrms and 1000Vdc respectively.

My customer and I need clarification as to what this means in light of the data in 5.6. In 5.6 it's 3750Vrms for t=60s, but 707Vrms in the Absolute Maximum Ratings. We need absolute clarity that the device will not be damaged by the 1400Vdc (1000Vrms) t=60s test that we need to do. As it stands we are forced to take the pessimistic case, which as written currently is that the device will be damaged by our 1400Vrms test. Is there any clarity to be had to resolve what looks like the contradiction between these two t=60s ratings?

Thank you for your consideration.

0 Eric Hackett 19 days ago in reply to Alex Gold

TI__Guru* 75972 points

Alex,

So, it's a bit confusing because this the result of a bit of faulty datasheet organization. But effectively, the absolute maximum rating seen in section 5.1 is correct, but the UL 1577 rating is also correct, and effectively 'absorbs' the rating seen in section 5.1. Essentially, both ratings are the same test, one is just the result of further certification testing. I know this is confusing, and I'm going to work with our internal team to update this datasheet.

So to answer your question, the device will survive the 1400Vdc (1000Vrrms) t=60s that you need to perform.

Regards,

Eric Hackett

0 Alex Gold 19 days ago in reply to Eric Hackett

Prodigy 10 points

Hi Eric,

Thank you again for the reply. I understand from what you've said that the part will survive the test and this is good for us. I do seek a little more clarity, important for us to be able to use the part in a design that is heavily scrutinised. We cannot, for example, test a few and demonstrate they are not damaged; we have to be able to show it beyond any doubt from the paperwork as well. Any issues like this will be assessed to take the worst case as the figure. I appreciate you have said the datasheet will be updated and I look forward to seeing the update.

In the meantime, to be 100% clear for our planning and to assure ours and our customer's compliance, will the new Absolute Maximum Rating move up to the 3750Vrms figure in 5.6 (or just beyond the 1000Vrms we need)? If you could assure us that is the case as a TI representative then I can take that in advance of the new datasheet and submit for an informal agreement to use by the other stakeholders.

Thank You Again,

Alex

0 Eric Hackett 18 days ago in reply to Alex Gold

TI__Guru* 75972 points

Alex,

Let me discuss this with our team further and I will get back to you on our plans to update the datasheet. What may be the case is that we remove the rows from section 5.1 entirely and just rely on the rating shown in 5.6, but again, I will discuss and get back to you.

Regards,

Eric Hackett