ITS4090

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
lock attach
Attachments are accessible only for community members.
GertH
Level 1
Level 1
First like given 5 replies posted 5 sign-ins

Hey!
Having problems with ITS4090.
When I use 12 VDC supply voltage there is no problem.
When the voltage reaches 24 VDC, the resistor R138 burns up. It does not happen on all circuit boards. Have tested 15 pcs before I finished the tests.
Q1 also burned up during a test.
What is it due to?
See the drawing below.

 

Thanks

/Gert

 

GertH_0-1662707051593.png

 

0 Likes
1 Solution
Anshika_G
Moderator
Moderator
Moderator
First question asked 50 likes received 250 replies posted

Hello @GertH ,

Sorry for the delay!
From the shared document, it is not clear what test sequence you are applying and what are the corresponding damages after each experiment (except for the resistor that is burned).
Also, I have some follow up questions for you :

  • What is the power rating of the used 27Ohm GND resistor?
  • What was the state of the resistor after being burned (I assume an electrical open)?
  • Were there any other damages observed next to the resistor (e.g. diode D31); or was in any of the experiments the ITS4090Q damaged itself? For the latter of course one would need to check this after the burned resistor has been replaced. If R138 is an “open circuit” it is clear that the ITS4090Q cannot work anymore. The loss of GND protection would be active in this case.
  • Is the sequence (how the 24V supply is plugged/un- and re-plugged) a normal operation condition of this application or is this only happening in these experiments?

 

Nevertheless, out of the described behavior one can draw some first conclusions:

The fault signature of the GND resistor being burned indicates a too high current flowing through the GND structure of the ITS4090Q. If the diode D31 is still o.k. it also indicates that the current flows from Vs (24V) to GND and there is no reverse polarity break through issue.

Where can this current come from?

Having an excessive current through the GND path from 24V to GND indicates the presence of overvoltage spikes. The assumption would fit to the fact that the fault is generated by de-plugging and re-plugging the 24V supply which could generate spikes. The ITS4090Q itself is protected against transient overvoltage spikes by the internal overvoltage protection which in essence a zener structure that clamps the voltage that the logic of the ITS4090Q sees to safe values. But during the time the overvoltage protection is active the current flow through the GND path needs to be  limited by an external resistor: this is the reason for having R138. The choice of the resistor needs to fit to the energy content of the transient spikes. Using 27Ohm is based on our experience a good value and should be suitable for most cases but also the power rating needs to fit to the energy content of the spikes.

How to verify whether this can be the root cause:

I think it makes sense to check whether spiking during hot-plugging could be the issue. This could be checked by applying externally a TVS diode between Vs and System GND and redo the experiment that causes the problem. The TVS diode used for this experiment should safely cut transient spikes below 60V (e.g. 39V TVS) as in this case the overvoltage protection will not become active.

If this experiment verifies the assumption in a next step one could check whether the problem can be avoided without the TVS diode by varying the power rating of R138 (or else even go to higher ohmic ones).  

Maybe as a last comment - In general - operating the ITS4090Q at Vs = 24V is no problem at all - the device is designed for this use case and besides quite robust.

Best Regards,
Anshika

View solution in original post

0 Likes
10 Replies
Anshika_G
Moderator
Moderator
Moderator
First question asked 50 likes received 250 replies posted

Hello,

I am analysing the schematic to find out the issue. Can you please share the voltage and current appearing across the resistor and transistor Q1 before burning up? 

By the first look, we can say the transistor Q1 and resistor are getting burnt due to excess current . But we need more data to analyze to find the exact cause.

Thanks and Regards,
Anshika

0 Likes
Hello,
I have done more testing.
Two cards worked but not the third.
Se testreport:
https://docs.google.com/document/d/1mYhNh0tjn8ImrbmYc5h0kS3N2FShrzk82Q_ZdDhoUoA/edit?usp=sharing

/Gert

Med vänlig hälsning / Best regards
Gert Hultqvist

+46 76 310 00 38
Office +4610 - 333 20 44


<>
Adductor AB - www.adductor.se
Production increase with the MUR system



<>


www.MUR-Box.com <>
0 Likes
Anshika_G
Moderator
Moderator
Moderator
First question asked 50 likes received 250 replies posted

Hello @GertH ,

Sorry for the delay!
From the shared document, it is not clear what test sequence you are applying and what are the corresponding damages after each experiment (except for the resistor that is burned).
Also, I have some follow up questions for you :

  • What is the power rating of the used 27Ohm GND resistor?
  • What was the state of the resistor after being burned (I assume an electrical open)?
  • Were there any other damages observed next to the resistor (e.g. diode D31); or was in any of the experiments the ITS4090Q damaged itself? For the latter of course one would need to check this after the burned resistor has been replaced. If R138 is an “open circuit” it is clear that the ITS4090Q cannot work anymore. The loss of GND protection would be active in this case.
  • Is the sequence (how the 24V supply is plugged/un- and re-plugged) a normal operation condition of this application or is this only happening in these experiments?

 

Nevertheless, out of the described behavior one can draw some first conclusions:

The fault signature of the GND resistor being burned indicates a too high current flowing through the GND structure of the ITS4090Q. If the diode D31 is still o.k. it also indicates that the current flows from Vs (24V) to GND and there is no reverse polarity break through issue.

Where can this current come from?

Having an excessive current through the GND path from 24V to GND indicates the presence of overvoltage spikes. The assumption would fit to the fact that the fault is generated by de-plugging and re-plugging the 24V supply which could generate spikes. The ITS4090Q itself is protected against transient overvoltage spikes by the internal overvoltage protection which in essence a zener structure that clamps the voltage that the logic of the ITS4090Q sees to safe values. But during the time the overvoltage protection is active the current flow through the GND path needs to be  limited by an external resistor: this is the reason for having R138. The choice of the resistor needs to fit to the energy content of the transient spikes. Using 27Ohm is based on our experience a good value and should be suitable for most cases but also the power rating needs to fit to the energy content of the spikes.

How to verify whether this can be the root cause:

I think it makes sense to check whether spiking during hot-plugging could be the issue. This could be checked by applying externally a TVS diode between Vs and System GND and redo the experiment that causes the problem. The TVS diode used for this experiment should safely cut transient spikes below 60V (e.g. 39V TVS) as in this case the overvoltage protection will not become active.

If this experiment verifies the assumption in a next step one could check whether the problem can be avoided without the TVS diode by varying the power rating of R138 (or else even go to higher ohmic ones).  

Maybe as a last comment - In general - operating the ITS4090Q at Vs = 24V is no problem at all - the device is designed for this use case and besides quite robust.

Best Regards,
Anshika

0 Likes

Thanks for the answer.
I don't think it's due to a spike in voltage.
When I test I start at 12V and increase to 24VDC with a variable voltage supply. When I reach 24 VDC, the resistor R138 burns.
I tried testing in a number of different ways (nine boards failed :)) until I started shorting R138 and the diode. When I did that, everything worked, even when I connected 24 VDC via a switch.
I haven't had time to test if the burned cards work when I short-circuited R138 and the diode, i.e. jumper U31 pin 2 to GND.
I will do so shortly (I have been waiting for the answer I have now received).


Many thanks
/Gert

 

This is the R138 we use : IPC-7351\Chip-R0805

 

Can we use ITS4090 without resistor and the diode?

 

0 Likes
GertH
Level 1
Level 1
First like given 5 replies posted 5 sign-ins

Thanks for the answer.
I don't think it's due to a spike in voltage.
When I test I start at 12V and increase to 24VDC with a variable voltage supply. When I reach 24 VDC, the resistor R138 burns.
I tried testing in a number of different ways (nine boards failed :)) until I started shorting R138 and the diode. When I did that, everything worked, even when I connected 24 VDC via a switch.
I haven't had time to test if the burned cards work when I short-circuited R138 and the diode, i.e. jumper U31 pin 2 to GND.
I will do so shortly (I have been waiting for the answer I have now received).


Many thanks
/Gert

0 Likes
GertH
Level 1
Level 1
First like given 5 replies posted 5 sign-ins

This is the R138 we use : IPC-7351\Chip-R0805

 

0 Likes
GertH
Level 1
Level 1
First like given 5 replies posted 5 sign-ins

Can we use ITS4090 without resistor ande diode?

0 Likes
Anshika_G
Moderator
Moderator
Moderator
First question asked 50 likes received 250 replies posted

Hello,

It's good that the setup worked :).
But it is generally not advised to not use any resistor at all in GND path.  As it protects the device from fault events like loss of ground and reverse polarity protection. You can try connecting a higher ohmic resistor in place of diode in series with the resistor(27 Ohms) and check the results.

Best Regards,
Anshika

Thanks for the answers.
I will measure the current through the 27 Ohm resistor, which you previously advised.
What value do you recommend for the resistor to replace the diode?
/Gert

0 Likes
Anshika_G
Moderator
Moderator
Moderator
First question asked 50 likes received 250 replies posted

Hello,

You can choose 150 ohm resistor(also mentioned in the datasheet).

Best Regards,
Anshika.

0 Likes