So let's acquire the data and see the results. The high speed camera is synchronized exactly to the analog data therefore we can compare the switching times with analog voltage and current. The current is flowing through the fuse for approximately 20 msec before it really switches off 
and we can see nicely the when spark switches the fuse off. It is also very nice to see the position of the switch (on the left side of the fuse) as it goes to the off position 
.
 |
| Speed cam trigger |
The picture from the web cam is nice, but doesn't show this at all. Because the web cam is neither clocked nor synchronized with analog data, it is time stamped as the picture comes in to the computer. Thus, we can see the switch off with a delay of approximately 60 ms 
. In conclusion, web cams have two limitations: speed and time accuracy.
 |
| Web cam trigger |
However, there is a way to reduce the time inaccuracy by entering a camera delay. Usually this delay is quite constant and can be compensated. We can do this in the Hardware setup by entering this value in the Delay field in the table. We will still have a time jitter of each frame, which can be in the range of 30 ms or even more, when the system is at the limit its performance, but here we can't compensate it. If time accuracy is needed, we should look for clocked cameras like Dewecam or Basler.
 |
| Delay setup |
Now if we look at the repeated measurement, both cameras show approximately the same time of the event, however, the web cam shows only one frame when the fuse switched off while the Basler camera shows quite several frames to be able to also see the spark, the movement of the coil and the switch.
 |
| Speed cam with delay |
This solution, however, is limited to 250 frames per second. Some events require much faster cameras. These cameras store the picture in internal memory and after that to the video file, so we need to combine the picture and video in post processing.
