Built in Test and Electronic Sensing

[Hydromechdude]

I have a general question regarding built in testing / failure sensing for electronic components. I realized I lacked this knowledge when reviewing an old failure report as I started wondering how failures are sensed. I'd like to do some research on this topic so that I understand the concepts and how these functions are performed.

The old failure report I came across contained the text below. I suppose the text below is more of general electrical question rather than controls. Would anyone be able to explain it in a watered down version?

"Our discrete outputs are supposed to provide a very low impedance to ground (ideally short) when on, and a very high impedance (ideally an open circuit)when off. In the On state, the low impedance prevents us from detecting the difference between normal current and no current, so we cannot tell if the
load is present or absent. We can detect the difference between normal and excessive current (overload, or load short). In the Off state, we cannot detect if the load is shorted to +28 V, since our high impedance limits the current flow."

Thanks,
HMD

 

[itsmoked]

In the On state, the low impedance prevents us from detecting the difference between normal current and no current, so we cannot tell if the load is present or absent.

This is not totally true. Since the 'low impedance' is NOT zero there is going to be a voltage present. That voltage will describe the current which - can be assessed.

We can detect the difference between normal and excessive current (overload, or load short).

??? Confusing. Contradictory.

In the Off state, we cannot detect if the load is shorted to +28 V, since our high impedance limits the current flow."

This is incorrect. The load is always going to be hooked to 28V with this type of output. The correct statement is "we cannot tell if the load is [blue]shorted to ground[/blue]".

That would be true. The only way to detect this would be to do "high-side current monitoring". There are lots of chips now for doing just this chore. That's a monitoring point between the 28V and the load. That way if any current is drawn from the source you can measure it. So, if the load were shorted to ground you would see the normal load current even though your output is NOT active.


Keith Cress
kcress -

http://www.flaminsystems.com

 

[Hydromechdude]

Therein lies my problem. I read it and it seemed reasonable to me, although I also tried to reason it to myself, as I thought it was a good statement.

So, any literature I can read up on to educate myself with this kind of stuff?

Thanks for the reply.

HMD