Japanese I\O

[James Mcquade]

all of my designs are based on sinking inputs and sourcing outputs.

for sinking inputs, power is applied to the sensor, the sensor then feeds the plc input which sinks the signal to the co common terminal. less power required with these types of inputs.

for outputs, the plc output module supplies power to the device and also requires less power.

I have never seen a Japanese design using sourcing inputs and sinking outputs, doesn't mean they don't. I just haven't seen one.

regards,
james

 

[Geospark]

Once again, the original statement, which I agree with, is that unexpected machine operation could result. This has NOTHING to do with SIL.

Original statement...

From what I have seen most Japanese equipment uses sinking I\O. This has never made sense to me, even dangerous. Even with a isolated supply if just one input becomes grounded then the supply is no longer isolated and the input would turn on correct ? Can someone explain why they do this ?

Steve,

I agree with your assessments at post 12 & 15, but the original statement was made at post 1, not at post 7. The OP stated that they believe it to be possibly dangerous to use sinking I/O. The contributors, as far as I can see, have been basing their replies on this statement.

There a fail safe for digital input/output/analog SIL-1/2/3/4/5.

I am curious to know though, what is Safety Integrity Level 5 (SIL5)?
There is only SIL 1-4 as far as I'm aware.

But can anyone spot a contradiction in that opening statement?

George

 

[osmanmom]

Until 4 only

 

[osmanmom]

I bring the SIL because I not interested to continues the discussions
The posted highlight the issue.
Somebody give a statement and please give the prove.
Please descript why use SINKING input or output are dangerous.
While using SOURCING less dangerous or safe.
I glad to know

 

[wdd0422]

Isn't unexpected machine movement dangerous ? Anyway, the explanation that it uses less current makes sense to me. I am more comfortable with PNP inputs and sourcing outputs.

 

[JerryH]

I'm at my second Japanese company now. Both companies purchased all of their equipment from Japan and all is Positive ground.

For a Positive grounded system you should use Sourcing Input Cards and Sinking Output Cards.

For a Negative grounded system you should use Sinking Input Cards and Sourcing Output Cards.

This is done for safety reasons to prevent any unexpected inputs from energizing if an input wired is shorted to ground and prevents any output device from energizing if an output wire is shorted to ground.

I suppose if your DC supply is floating you could go either way.

 

[Steve Etter]

I stand corrected and apologize. I misspoke by referencing the original post when, indeed, I was refering to Post #7. My bad.

wdd0422 - No, unexpected movement is not necessarily dangerous. That's is what the references to properly designed systems and SIL come in. If the machine has the proper safeties in place, machine movement - either intended or inadvertent - will not be dangerous.

osmanmon - For this example, I refer to +V and -V as my source and common, respectively.

An NPN input circuit is referenced to +V and the controller sees a "true" when the switched -V leg completes the circuit and "false" when it does not. Because it is prefered by many (most?) designers for all commons to be bonded to ground, if a short occurs between the common wire and ground on an NPN input circuit, the controller will see a "true" condition. This false "true" is much more likely to be seen on an NPN input than it is on a similarly designed machine with PNP inputs where a shorted input wire would have to make contact with +V.

Does that make sense?

Edit - I see JerryH said much the same while I was typing

 

[BobB]

Confusion reigns (as usual)!
I hate the descriptions sinking and sourcing - they are used in different ways by different manufacturers. What one calls sourcing another calls sinking. One of my pet hate descriptions.
While NPN transistors (a much better description) generally switches with lower power I was told years ago that NPN transistors in the early days were much cheaper than PNP transistors - that was when I was doing an electronics course in the 1960s - and that was why there were used.
I do understand that if you switch a negative to an input rather than a positive and the 24VDC is grounded an unsafe condition can occur IF the negative of the 24VDC power supply is grounded. Easy fix, I NEVER ground the negative of the DC power supply. For a start if you use a switch mode power supply the internals present an isolated supply - grounding the negative removes the isolation and creates the possibility of a dangerous situation.
That being said I also normally use positive switching inputs but negative switching inputs do not bother me either due to the isolated power supply and do not really care if the outputs are NPN or PNP transistors. They only go to relays or lights with my designs anyway.
Never had a problem with this at all and never had a dangerous situation arise.
Same thing with 24VAC for switching contactors, double wound, isolation type transformers once again with the 0 volt output not grounded. Never had a problem with this either.
Added advantage I see from the non grounding is far less possibility of noise from a dirty main supply into the control circuit.
A lot of my jobs are generator control systems running on 24VDC battery supplies and chargers. I do ground the battery negatives which run the general control circuit but then go through a 24VDC/24VDC switch mode power supply with the output left to float. Added advantage here is that if the battery voltage drops to 19VDC the switch mode kicks it up to 24VDC and keeps the PLC system going - extends the time the system can operate on a loss of mains power.
Hope this makes sense.

 

[shane.horan]

So you have NPN inputs at 24V, sinking to 0V to give a logical true. If your ground is linked to the 0V of the power supply, then an inadvertent ground connection to your input terminal gives a logical true. Which would be bad.

So if you have NPN inputs, use a floating 24V supply to prevent the inputs sinking if they get grounded. Lovely.

My question is, if the input terminal gets a good solid ground to your machine's steel frame, won't you get a transient current until the potential difference between the power supply and the frame is equalized? Milliseconds maybe, but enough for the PLC to see a blip? I'm genuinely curious. I've used both PNP and NPN inputs but always with a floating 24V. I've never had a false positive but is it possible?

FWIW, I believe NPN transistors are still faster than PNP (for technical semiconductor reasons), which is why they're common in high-speed signals like motion control interfaces and encoders.