TIOS101: TIOS101-3 failed short circuit test

Hi KPK,

I repeated this same test on an EVM here and was unable to reproduce the failure. I only had one unit available, so I have ordered a few more to support additional tests.

In general the device should be protected from these kinds of overcurrent conditions via two mechanisms: the output current monitoring (and automatic shut-off) as well as thermal shutdown. These should help reduce the likelihood of this kind of failure.

Can you please explain a little more how the short circuit is applied? If the short occurs through a cable, then there is a chance that voltage spikes could occur due to the way the inductance of the cabling responds to a fast change in current (which would occur, for example, when the TIOS101 driver shuts off). If you had oscilloscope probes capable of supporting higher voltages, you could check for these kind of transients by monitoring the OUT pin and 24-V supply while the short is applied.

I'm also curious about the power supply used. This condition results in a large and relatively quick change in load current. Some power supplies may respond to load transients with overshoots or ringing on their regulated output. This could again be checked via an oscilloscope.

Regards,
Max

hi Max,

Below is the information: 

I_LIM_ADJ pin is shorted to Ground.  This means Rset is zero

The short was done using a simple cable. No load on 'out' pin during the short.

To give you some more insight of the testing, below is the oscilloscope measurement during the short.

DO_NFAULT = NFAULT signal from IC (pin 2)
DO_Valid_LO = it is a signal that monitors the output voltage (this is our internal requirement). The monitoring is using simple transistor based solution. Goes low if the output is HIGH.
DO_OUT = output (pin 8) of the IC,

I agree that there was overshoot and ringing observed during the measurements. But this was observed when FAULT signal went low.

Kindly check and let me know if you can draw some conclusion out of it.

Best Rgds,

kpk

KPK,

I should get the new units tomorrow, so I can check a few more parts then.

In the meantime, I'd be interested in understanding the overshoots a little better. The ringing on DOUT shown in your screenshot corresponds to when the output driver shuts off (as signaled via a falling edge on NFAULT), so it likely due to current rapidly changing across a cable inductance. There may be a similar phenomenon as well when the current switches back on (~15 ms later). Do you think you could check both of these conditions and zoom in a bit more on the oscilloscope's horizontal scale so that we can see the full amplitude of the ringing?

My concern is that the overshoots are damaging the part, either directly due to over-voltage stress or by triggering the internal transient protection clamping circuits (which would then try to sink current from your DC rail and become damaged). If this were the case, then a solution as simple as some additional capacitance on the OUT pin may help.

Max

hi Max,

My PCB are limited sets and I cant afford to damage anymore of them. Hence I will not be doing any more short circuit test until I have IC samples which already requested. I should be to test it next week once i get the sample to replace.

at the mean time, can you run just one test for me?

1. Keep the output in tristate and short to zero. (EN pin is low or open)
2. Same as 1 but short to 24V

Best regards,
kpk

KPK,

I tried this test and did not notice any unexpected behavior. With the EN pin tied low, the output of the device acted as a high impedance whether 0 V or 24 V was applied. Neither condition seemed to damage the device.

I also repeated the previous testing of shorting OUT to VCC (set to 24 V) with RSET = 0 Ohms, EN high, and TXD high (so that OUT is pulled low) across three different EVMs. I saw the correct behavior from all of them. The driver recognized an overcurrent condition and shut off automatically, and then retried on a periodic basis. When the driver was enabled it would pull the OUT line down by 5-10 V (not all the way to 0 V since my power supply is stronger than the TIOS driver). Upon automatic disable of the driver, I did not observe any overshoots as the output voltage increased back to 24 V. (My shorting wire was relatively short - about 0.5 m).

Regards,
Max

KPK,

I was just curious if you had any further progress on resolving this. Have you received the replacement samples and conducted the testing you had planned?

Thanks,
Max

hi Max

Thx for follow up:

Yes. just got the sample IC to test on the board. Looks fine for us. It indeed looks like the damaged IC was one isolated incident and hence can be ignored at present.

Can I check are there any inherent differences between high and low side drivers? In other words, if I set maximum output current to 200mA then what value should I expect for high side driver output and low side driver output? Maximum variations in % or absolute value are ok. I am sure TI should have the data (either graphical or tabular format) over the entire range
I would like to know the differences because this could be a safety critical issue from our design point of view. We design based on IEC 61508 standard. We need this information for our design parameters.

Best regards,

KPK