This driver is used (2 per module) in a fairly new product we have been shipping 10k modules in past 6 months. Until recently, we have had no driver failures. Now we have two returns from a customer with a driver failure on each.
On one unit, one channel of a driver turns on when the input goes high but does not turn OFF when the input is removed.
On the second unit, one channel of a driver does not come on when the input is high. Additionally, the inputs are typically driven high by applying 12V through a 10K ohm resistor. The inputs usually clamp around 6V. On this channel, the output goes all the way up to 12V - there is no clamping at all.
By the way,
1. The output is OFF initially. It comes ON when the input is asserted and will remain ON until power is removed.
2. VDRV is nominally 12V.
Have you ever seen this type of failure and if so, what was the cause?
NOTE: This question and answer originated from a customer support case which was handled by us. We regularly review support cases and add any helpful ones here for all to benefit from.
1 ACCEPTED SOLUTION
Accepted Solutions
The substrate of the chip is the supply. My concern is that any positive transients on the supply pin VCC (VDRV), as long as the input is held low, may have an impact. In fact, positive transients on the supply can find their way to the I/O via parasitic components in the IC. That is why we recommend to add RProt resistors at the pins. Here is the modified schematic with my suggestions.
Described change will probably help to avoid damage of the input drive circuitry by transients on the supply. I say probably because it mentioned transients would have to have very fast rise time. But there are other possible root causes to consider: it could also be an ESD event.
Electrostatic discharge (ESD) can damage an input during handling (assembly of your board) and not demonstrate the effects of the damage until the device has had some time in the field. We call this "walking wounded". Reconfiguring the input drive circuit will not fix this issue. In fact, you will have to make sure that strict ESD procedures are in place. This subject is something you will want to investigate in-house...
But while on the subject of ESD I noticed that your schematic did not show any capacitors on the outputs at the connector pins. To increase the module ESD capability a simple 100V, 22nF cap can be placed at all of the pins at the connector. This requires a small ground plane around the connecter where the caps are connected to each of the pins. This suppresses ESD at the connector and increases the robustness of the module both for ESD and EMC. While you are making the mods to your input control circuits you might want to add these small caps as well.
The substrate of the chip is the supply. My concern is that any positive transients on the supply pin VCC (VDRV), as long as the input is held low, may have an impact. In fact, positive transients on the supply can find their way to the I/O via parasitic components in the IC. That is why we recommend to add RProt resistors at the pins. Here is the modified schematic with my suggestions.
Described change will probably help to avoid damage of the input drive circuitry by transients on the supply. I say probably because it mentioned transients would have to have very fast rise time. But there are other possible root causes to consider: it could also be an ESD event.
Electrostatic discharge (ESD) can damage an input during handling (assembly of your board) and not demonstrate the effects of the damage until the device has had some time in the field. We call this "walking wounded". Reconfiguring the input drive circuit will not fix this issue. In fact, you will have to make sure that strict ESD procedures are in place. This subject is something you will want to investigate in-house...
But while on the subject of ESD I noticed that your schematic did not show any capacitors on the outputs at the connector pins. To increase the module ESD capability a simple 100V, 22nF cap can be placed at all of the pins at the connector. This requires a small ground plane around the connecter where the caps are connected to each of the pins. This suppresses ESD at the connector and increases the robustness of the module both for ESD and EMC. While you are making the mods to your input control circuits you might want to add these small caps as well.
