Some great advice from somebody else - not me

you can buy a real module to convert from NPN/PNP and back. This is cleaner than the transistor/resistor hack I've done in the past. -
http://www.woertz-usa.com/pnp-npn%20data%20sheet.pdf

Worked once for a manuacturer who told me he'd rather use them Sinking devices for his less experienced technician didn't get has much imbeded smoke released with them.

And it was thrue.

Make a short and the Input resistor will keep the smoke insideyour devices.

Personally I try never to use them.

A, the great PNP versus NPN (a.k.a sinking vs. sourcing) debate...

I, for once, am sticking with NPN inputs. For once, we are using a lot of touch probe type of signals, so grounded negative is mandated, not optional.

A short in input circuit may start a machine? Well, if it is safe to start (E-Stops are reset, guards are closed etc.), it's a nuisance, not a tragedy. If it can start for any reason (a shorted input wire is just the same as an idiot pushing "Start" button) while it is not safe to start - wel, it is just a poorly designed machine and switching over to PNP sensors ain't gonna make it any safer.

For any other sensor - just check them for break and make. A shorted input will cause an alarm. Old technique, works fine.

If I had to start from absolute scratch, I might've considered PNP. But there are so many machines already in the field... and the sensor makers somehow don't seem to bother themselves with some kind of clear and visible markings to distinguish a sourcing prox from a sinking one... Can't even imagine what kind of mess for the maintenance guys would be if some equipment was one way and some - another.

A short in input circuit may start a machine?

Click to expand...

Oh dear!!!! Operator starts generator (power has not failed and set is in manual), short says mains circuit breaker is open, manually close generator circuit breaker because mains breaker says it is open, generator is out of synch, generator leans over very hard and dances across the floor!!!!

NPN inputs, no thank you!!!!!

Great article by Jim Rowell. I wish all designers could read 33 and 35.

Grounding DC Power Supplies?

When we talk of grounding power supplies, are we talking about chasis ground from the power supply or the DC-(COM) to ground? I don't refer DC-(COM) as ground because the point of reference to the power supply is not gound. Would that be correct? And if DC-(COM) were connected to ground, aren't you creating a antenna for noise, which will cause power supply issues?

I think Jim Rowell really is saying that the power supply commons should be grounded.

He obviously doesn't work with sensitive analog sensors.
Analog sensors need to be powered by their own isolated power supply. A high impedance resistor can be used to keep the floating power supply from floating too far. Analog circuitry uses differential signals so the absolute voltage is not as important as the relative voltages. HOWEVER, the voltages must still be within the range of the input device which is why a high resistance resister between the isolated power supply command and ground is needed.

As Jim Rowell pointed out, you must make sure that neither the isolated 24 nor common are ever grounded. I will add, don't put DC devices on the isolated analog supply and don't daisy chain commons. Device commons should be home runs to the power supply common. Think about it.





My pet peeve is the use of AC photo cells. My 'integrator' gets wound up a little more when people ask me why we don't have AC inputs. My 'integrator' winds to the limit when people connect AC inputs to the DC inputs and let the smoke out.

BobB said:

short says mains circuit breaker is open

Click to expand...

... while in reality it is closed - or am I missing something? So we have either the mains CB being monitored only for open position (and not monitored for closed position) or "both signals on" situation not causing an alarm. Poor design, IMHO, especially for such potentially hazardous type of system.

BTW, consider a short on PNP input system, blowing the fuse or sagging the voltage on a whole group of inputs. I am not sure that would be inherently any safer.

... while in reality it is closed - or am I missing something? So we have either the mains CB being monitored only for open position (and not monitored for closed position) or "both signals on" situation not causing an alarm. Poor design, IMHO, especially for such potentially hazardous type of system.

Click to expand...

Having control of the complete design would be very nice. However, in reality, some bright spark electrical consulatant has put out a pile of specs for the job. One of them comes to us and the others go to other people, including main switch board builders (bus bar stuff). Invariably someone tries to save a cent or two and we finish up having to work with what we are given.

Complain bitterly, point out the safety issues - design issues - faults in design. No one wants to listen as it will cost money. Keep all the paperwork, e-mails etc etc to cover ones own a**e and our insurance company who would try to get out of paying anyway.

Not ideal but a fact of life. I love jobs where I can have some input into the overall design and build in all the safeties possible. Does not happen often I can assure you.

Safer to use positive inputs and forget negative inputs all together.