| ??? 03/31/05 03:49 Modified: 03/31/05 03:59 Read: times |
#90723 - No, not fundamentally, but... Responding to: ???'s previous message |
Sasha said:
What makes optoisolation too dangerous? Optoisolation isn't that dangerous that it should never be used for sensoring mains, of course. But that what you will actually find on your PCB finally, often is! Why? Not all optoisolators are VDE classified and if so, then not all packages are VDE classified. But even if so, the layout with solder pads can finally violate VDE laws again. So, what looks proper in datasheet can be a disaster on your PCB... More, DIL package style optocouplers do not look it to carry dangerous potentials on its pins. Instead of a transformer. This is a plain psychological issue. More, you cannot simply touch the pins of a transformer, not so simply as you can touch the pins of an optocoupler... VDE certification does only focuse on the isolation characteristics of the naked optocoupler. But with MID400 you have additionally to connect a resistor in series to mains. And if you apply this MID400 with 230V mains you have to add about 47kOhm. Not mentioning that this results in a heat dissipation of 1.1W, this resistor has to withstand 6000V and more! Do you really think that this 47kOhm will resist 6000V and more, even if this is only for a short period? Again, this resistor will also be in danger to be directly touched. All these issues make me think, that a transformer is safer than an standard opto, even if datasheet about naked opto tells it's safe. By the way, the transformer I talk about is only 22mm times 24mm in area. This will not relevantly waste much more space than the opto plus resistor. As advantage Geert will have an interface which is well weaponed against even the worst bursts and surges. Much safer than a standard opto. Another issue often totally overlooked is ESD. As the capacitance between input and output of a good opto is in the sub picofarad range any ESD event hitting the barrier will fully float it with extremely dangerous voltages. I have seen many optos in the past, which were killed just by ESD events. Why am I so sure, that ESD events were responsible for damage? Just simply, because these damages occured in winter months, where we have very low humidity from time to time. Best conditions for developping lots of electrostatical charge being eager to destroy helpless optos... Kai |
