At what point do you consider design safety?

frank freedman

At what point would one consider an opto or magnetic isolator for driving an ULN2308 input? It is all on one piece of silicon after all and most stepper motor voltages will fry the controlling logic dending on the failure. Same for other higher power dedicated single chip drivers. This isn't something I recall seeing in prop or many other discussions. For example, power drive circuits where an FET punches through and takes out everything back to the opto isolators output including pcb traces in a particular medium frequency high voltage generator board.

I ask this tangent to the thread regarding stepper drive and yet another dead prop to see what others here think, do or perhaps, do not. And also circuit protection (and human) is a topic that does not seem to come up much on many forums such as this one. That was a lower power application, but some of the other projects in the forums can be at significantly higher power (AC line, high currents from battery banks etc.)

Heater.

A while back I was tinkering with a boost circuit, I wanted to see if I could get 200 volts from 12 volts using a 555 based circuit. The first breadboard attempt work surprisingly well. 150v lighting up a dummy load of neon lamp and resistor.

Until, that is, the neon lamp became disconnected. With the load gone the output voltage must have soared above the break down voltage of my MOSFET, that 200v found it's way out of the MOSFET gate and back down the circuit and took out the driver transistor and 555.

Annoying but quick and simple to fix.

So, sometimes taking safety precautions to prevent these kind of burn down is a time and expense saving idea. My next attempt had the voltage control feed back hooked up and tested before I started pushing for the higher voltages.

I imagine every case is different. For example if the circuit works and there is a low probability something bad happening and failure is cheap then why bother protecting it.

As for human safety. I'm not qualified to advise.

kwinn

Safety requirement depends on multiple factors such as intended use, voltage, current, location of equipment, etc..

Peter Jakacki

There is a difference between design safety and breadboard design "safety" as the latter is normally confined to the bench and most of the time we just want to test the functionality, not the safety unless it is a safety device. This is where I try not to over-engineer a design at this point so I can see what does and what doesn't work. However in the case of the ULN part taking out the Prop that was simply a matter of wrongly assuming that the clamp diode COM pin should be connected 3.3V and what happened after the motors were connected to 12V was entirely predictable.

I've never needed to "isolate" driver circuits for this reason, just careful layout and design is all that is required, but if you don't know what you're doing then you may tend to isolate and over-engineer and in the "just to be safe" process end up perplexed when something doesn't work.

Understanding circuit operation and datasheets along with electronic engineering in general is really a prerequisite for designing for safety which also depends upon how and where and by whom it is used etc.