5
\$\begingroup\$

This datasheet seems to promote specific TVS diodes as good for absorbing an automotive load-dump. This graph seems to support the idea:

enter image description here

However, these other graphs seem to contradict any load-dump absorption capabilities:

enter image description here

Extending the line on the third graph to 200 ms (half way point to 400 ms of load dump event, worst case) shows some ~1.6 kW capability.

For the SM8S24A, worst case peak power will be 38.9 * (79 - 38.9) / 0.5 = ~3.1 kW - which is almost double what the part can withstand even at room temperature.

I have noticed this other oddity:

Maximum Peak Pulse Power Dissipation (10/1000 μs) P PK 6600 W

Maximum Peak Pulse Power Dissipation (10/10000 μs) P PK 5200 W

In the 3rd graph it looks like they are using a 10 μs rise time and then just stretching td up to 100 ms? ISO talks about 10 ms rise time, not 10 μs. Can someone shed light on this? Is the part actually up to the task or is this some marketing trickery and nothing else?

Are Vishay and Bourns related companies? Here is their (Vishay) 6.6 kW diode, pretty much the same thing:

enter image description here

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Vishay and Bourns are separate, but many components like these are equivalent and multiple-sourced. See also MOVs, there's not really anything you can improve upon, they all use the same materials, there's just small variations in ratings between product lines and manufacturers. TVS of course are tighter spec'd so you have less variation between equivalent products. To get different curves, you'd have to have material differences, like thicker leads/tabs, or heatsinking (just barely relevant in the 100s ms regime), etc. \$\endgroup\$
    – Tim Williams
    Oct 7, 2022 at 5:45

2 Answers 2

Reset to default
3
\$\begingroup\$

Is the part actually up to the task or is this some marketing trickery and nothing else?

I think I see what they have done. You need to look at the small print in ISO 16750-2 on page 12: -

enter image description here

The key word in the lower highlighted sentence is "or". In other words they are assuming the engine alternator is revved up to full-speed producing a peak of 101 volts when the load is shed.

Under these circumstances, you choose \$R_i\$ to be 4 Ω.

The SM8S24A device's terminal voltage could be as high as 38.9 volts (as you understood): -

enter image description here

But, the current (101 volts limited by 4 Ω) is only 15.525 amps hence, the power dissipated by the device will be no more than 604 watts.

I now don't see any inconsistency with the extension of your 3rd graph either: -

enter image description here

Even at 400 ms, the maximum power that can be handled is still above 1 kW.

\$\endgroup\$
5
  • \$\begingroup\$ Maybe, maybe not. First graph in my question - does it not suggest they went even beyond worst case in the ISO (which, as you pointed out, is 79V with 0.5 ohm) ? \$\endgroup\$
    – kellogs
    Oct 6, 2022 at 18:17
  • \$\begingroup\$ @kellogs maybe but, it boils down to them either mis-selling the part or maybe they just haven't put work into understanding the ISO. Or maybe, when the supply is 79 volts, the duration of the pulse is only 40 ms (as per the table I embedded in my answer). And, the true half value duration of the pulse is closer to 20 ms when applied to the 2nd graph in your question. \$\endgroup\$
    – Andy aka
    Oct 6, 2022 at 19:13
  • \$\begingroup\$ @relayman357 I just don't know what I was thinking LOL (fixed) \$\endgroup\$
    – Andy aka
    Oct 6, 2022 at 19:46
  • \$\begingroup\$ @kellogs are we done with this Q and A now? \$\endgroup\$
    – Andy aka
    Jan 20 at 16:03
  • 1
    \$\begingroup\$ Yes, since a good while ago. \$\endgroup\$
    – kellogs
    Jan 25 at 15:10
1
\$\begingroup\$

The td of ISO 16750-2 is not the same as the td mentioned in the device you are referring to.

  • td of the ISO is from 10 % rise to 10 % fall of load dump.
  • td (or better said tp for pulse duration) is duration from 50 % rise to 50 % fall of exponential pulse

So the energy in the both case is really different.

I am not sure that the load is exponential by the way.

For the 3rd graph, it is a standard figure in TVS: the peak pulse power derating versus tp for exponential waveform but it is not necessary linked to ISO. This curve helps to calculate the power capability of the TVS for any surge duration for exponential waveform. Then you have some rules to convert tp from exponential to rectangular or sawtooth or whatever. You can get more details in this AN316

I hope it clarifies a bit.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.