TSC888 failure modes

[Aung Ko Ko Thet]

Hi All,

Is there any one who has good experience in using high side current sensors in bi directional DC Motor Controls?

We are using TSC888 high side current sensors on the 12V DC Supply lines to 7 DC Motors, 4 bidirectional and 3 unidirectional. We use L293 ICs as motor drives.

There has been sensor failures within two weeks of running the system continuously. (Approximately 80 hours). The sensors always fail in such a way that the output shows 10 times the actual value.

I would like to know possible root causes for this from any one who has experience in using TSC888 in the motor controllers.

Thanks in advance
Aung

 

[John S]

Hi All,

Is there any one who has good experience in using high side current sensors in bi directional DC Motor Controls?

We are using TSC888 high side current sensors on the 12V DC Supply lines to 7 DC Motors, 4 bidirectional and 3 unidirectional. We use L293 ICs as motor drives.

There has been sensor failures within two weeks of running the system continuously. (Approximately 80 hours). The sensors always fail in such a way that the output shows 10 times the actual value.

I would like to know possible root causes for this from any one who has experience in using TSC888 in the motor controllers.

Thanks in advance
Aung

Probably due to over-voltage. We had a similar failure using a TS1100.

John

 

[Jamie]

Hi All,

Is there any one who has good experience in using high side current sensors in bi directional DC Motor Controls?

We are using TSC888 high side current sensors on the 12V DC Supply lines to 7 DC Motors, 4 bidirectional and 3 unidirectional. We use L293 ICs as motor drives.

There has been sensor failures within two weeks of running the system continuously. (Approximately 80 hours). The sensors always fail in such a way that the output shows 10 times the actual value.

I would like to know possible root causes for this from any one who has experience in using TSC888 in the motor controllers.

Thanks in advance
Aung

We out actually seeing some circuitry, it would be hard to say but I
am leaning towards underrated parts or, EMF noise that results in HV is
getting back to the sensor circuit via the motors.

But my guess would be that a shunt resistor is opening,

Jamie

 

[John S]

We out actually seeing some circuitry, it would be hard to say but I
am leaning towards underrated parts or, EMF noise that results in HV is
getting back to the sensor circuit via the motors.

But my guess would be that a shunt resistor is opening,

Jamie

Our shunt resistor value is .004 ohms. It is less than a zero ohm
substitution part in most cases. The Kelvin connections are now as good
as they can get on this second board layout and there have been no failures.

Note that the common-mode specifications are very important here.

Are there any spikes on the line that will exceed the common-mode specs
of the chip? My bet is a huge YES!

So, I don't think you have used these parts nor have you studied the
specifications in detail.

 

[Aung Ko Ko Thet]

Hi John,

Thank you so much for your kind interest and comments on this matter.

We have found two things:

1. We have carried out validation measurements to make sure that nothing would exceed the absolute maximum ratings. Only one problem, there is a negative going ring 20 us perioud (4V pk-pk) for 60 us when the system is power cycled. It is due to the low quality 12V Switch Mode Power Supply. The rating of the IC says the common mode minimum is -0.3V. The negative going peak of -2V is already exceeding. So we have a preliminary conclusion that this could be one of the root causes of IC failure.

2. The ESD spec: of the IC is quite low, HBM 1kV and MM 100V. So it is moresensitive to ESD than other ICs, as we the system is open and the PCB is exposed to human interactions for this moment in time, it has a high ESD risk.

Please let me know your opinion also

Regards
Aung

 

[Aung Ko Ko Thet]

Hi Jamie,

Thanks for your kind comment.
Actually, there has been no failure in the shunt resistor. After replacing the IC, the current measurement systems works fine. It is basically due to the IC which somehow malfunctioned.

Regards
Aung

 

[Jamie]

Hi Jamie,

Thanks for your kind comment.
Actually, there has been no failure in the shunt resistor. After replacing the IC, the current measurement systems works fine. It is basically due to the IC which somehow malfunctioned.

Regards
Aung

You need to do a couple of things.

1.
You need to provide a clamp diode on the supply line to grab
any below ground problems. This is a common issue with flaky supplies
and people connecting things backwards. A high current schottkey should
work for that. One of my cases, I had to use an active clamp to suppress
high currents due to ground faults from other equipment.

2.
The current sensor you are using has a diff voltage of 2 volts and if
for some reason you get a high current pulse, which does happen for
brief moments, you can exceed this. In cases like this, it is more than
likely the front end does have diodes internally however, I didn't see
if the unit actually has sinking resistors inside? So maybe some extract
protection is in order, like R's of your own connecting the inputs to
the shunts? Or, two diodes over the shunt to prevent the raise of
differential voltage.

3.
You may want a big fat suppresser on your DC buss. The drive IC's you
are using more than likely have diodes in the bridge to capture the
regen energy and the motor could get into a free spin condition there by
generating more energy back to the supply. When people are involved it
is easy to get a motor load to be free spun out of control! (load swing)

So going back to (1. issue), a high wattage zener of 12 volts is in
order or active clamp (dynamic Brake R circuit), like I use..


Jamie