Bypassing an NPN Sensor...

Eric Nelson

Lifetime Supporting Member + Moderator
Join Date
Apr 2002
Location
Randolph, NJ
Posts
4,346
EDIT: Sorry, guys. My post seems to have no word wrap. Phil, any clue why???... :confused:

For some reason, I'm second guessing myself, so I'm hoping you guys will straighten me out. As usual, I'll make it a long, boring story... :D

I have a stand-alone label applicator that is currently using an NPN sensor (Banner PicoDot PD45VN6C50Q) to register the label to a mark on round tubes. Tube spins on a mandrel, and when it sees a mark, it spits out a label.

The mandrels are on a indexing turret, which presents the tubes to the labeler. Since I only want the labeler to look for a registration mark when the tube is in position, I control it through the main machine's PLC. Sensor goes to PLC input, PLC output goes to labeler's input. Output mirrors the input whenever I allow it, using a PERMIT bit. Basically:

| INPUT PERMIT OUTPUT
|----] [-------] [----------( )

Sometimes we don't need registration, so I also had a bit to override the input via the HMI:

| INPUT PERMIT OUTPUT
|----] [---+---] [----------( )
| |
| NO_REG |
|----] [---+

Of course, running the sensor through the PLC delays the response, but that has not been an issue until recently. We increased the rotation speed on the mandrel, and can no longer reliably detect the registration mark with the PLC. The input response time just isn't fast enough, plus scan time becomes an issue. The labeler has no problem reading the short duration of the signal, so my solution was to directly connect the registration sensor to the labeler, with a PLC controlled relay in series to enable/disable the sensor as needed.

NPN Sensor Labeler
+---------+
| DC+ |-------------------------> 24VDC
| |
| |
| N |
| P DC- |-------------------------> 0VDC
| N |
| |
| | PERMIT
| OUT |---------------] [-------> INPUT
+---------+
Yes, I could have moved the sensor to a high-speed input on the PLC, used interrupts, used an SSR on the output, etc., etc., but the bottom line is that I don't use the registration sensor for anything in the PLC, so there is no need to tie it to the PLC.

Okay, FINALLY to my question (assuming anyone is still reading)... ;)

This new setup works great, BUT there's no provision for running without registration. To run non-registered tubes, I have to bypass the sensor, allowing the PERMIT relay to give the registration signal. Can I simply install a selector switch across the sensor without damaging it? Like this:

NPN Sensor Labeler
+---------+
| DC+ |-------------------------> 24VDC
| |
| |
| N |
| P DC- |------------+------------> 0VDC
| N | BYPASS |
| | +--] [--+
| | | PERMIT
| OUT |----+----------] [-------> INPUT
+---------+

Or, do I need a second pole on the selector switch to also remove power to the sensor? Like this:

NPN Sensor Labeler
+---------+
| DC+ |-------------------------> 24VDC
| |
| |
| N | BYPASS
| P DC- |-------]/[--+------------> 0VDC
| N | |
| | +--] [--+
| | | PERMIT
| OUT |----+----------] [-------> INPUT
+---------+



Yes, it is a simple (and stupid) question, but it's bothering me... :(

🍻

-Eric

 
Last edited:
Wow, that really sucks. If my post wasn't already boring enough, you also have to scroll back and forth to read it!... :sick:

🍻

-Eric
 
I think you missed a *LADDER* code somewhere.

I don't think a 2nd pole is needed, afterall, the sensor shouldn't detect anything with no reg mark, right?
 
harryting said:
I don't think a 2nd pole is needed, afterall, the sensor shouldn't detect anything with no reg mark, right?
The sensor will stay connected, and it's output will likely turn on/off while bypassed. Actually, bypassing while the output is OFF is the part I'm concerned about. I honestly think I can safely override the output with power on, but the sensor is rather expensive, so I just wanna be sure... :oops:

🍻

-Eric
 
I am not sure I understand the issues, if you dont need the labeler why not just turn it off when not needed...ie a 2 position switch for registration/no-registration? OR if the plc controls the "process" and provides a "permit" for registration then you could use that permit with a relay to provide the power to the labeler and/or the sensor...should not hurt the sensor but if you want to make sure there is no false labeling I would "inhibit" or turn off the labeler too.
 
Maybe I'm missing something but - your 'permit' blocks the signal from going to the labeler - the NPN sensor can never pull low. But in both your PLC code and in the suggested 'bypass' you are essentially turning on the output all the time. I suppose the labeler does its thing on the LOW/ON -> OFF/HIGH transition of its input. Why not just run the bypass in exactly the same way as the permit?
 
As usual, I didn't explain well. Now I will attempt to further confuse you... :D

Non-registered does not mean no label, only that the label is not aligned to anything. Think of a toothpaste tube. When you get it, the end is crimped, so it has a 'front' and 'back' panel. Before it's filled and crimped, it's just a round tube with printing on it. Notice there's a mark at the crimped end (usually on the back). That mark is put there when the tube is printed.

The machine that does the crimping spins the tube until it see this regitration mark, then stops relative to this mark. Now when the tube is crimped, the printed text/graphic winds up in the proper 'front' and 'back' location.

With toothpaste, the text and graphics are usually printed at the same time as the registration mark. When WE get tubes, there is ONLY a registration mark printed. Our job is to install wraparound labels. We need to position our labels relative to that mark. When the tubes go for filling, they use the same registration mark to position the tube for crimping. In the end, the printing on the labels is properly positioned on the 'front' and 'back' panels.

We don't always register to a printed mark. Usually, we register the label to the flip cap on the tube. We position the label relative to the 'thumb notch' (the places where you stick your thumb to open the flip top). Yes, it DOES matter to the product designers... :rolleyes:

Once in a while, we don't need to register to anything. Just wrap the label around the tube, period. This is what I mean by 'non-registered'. I still ONLY want the labeler to spit out a label when there is a tube present, which is why I have the PERMIT signal. I simply want to ignore the registration sensor when running in 'non-registered' mode. To do this, I need the registration sensor ON all the time, hence my BYPASS switch.

🍻

-Eric
 
I think I understand now

I think I get it now, when in registration mode the sensor operates ONLY when it detects a registration mark then the label is applied. The sensor has to be ON though for "ANY" label to be applied.

Gee Eric you are probably better at this then I will ever be but if the sensor is not going to be "used" per se I would bypass it completely...ie take power away from it and send a signal direct from a switch, relay or the plc to enable the labeler.

I reckon the simplest method would be a 2 position switch with dual contacts, one contact feeds power to the sensor for registration and other would bypass sensor (remove power to it) and send signal direct to labeler.
 
rsdoran said:
I think I get it now, when in registration mode the sensor operates ONLY when it detects a registration mark then the label is applied. The sensor has to be ON though for "ANY" label to be applied.
Exactly!... (y)

Bottom line, I guess my question is:

Can I safely connect an NPN sensor's output in parallel with a selector switch?

For some reason, I haven't completely convinced myself that the answer is yes... o_O

🍻

-Eric
 
I would not but can not offer a definitive reason why. My suggestion was use the selector switch with the feed power to the sensor on one contact...ie when in registration mode the sensor is powered...when switched to other contact, power to sensor is OFF and 2nd contact sends signal to device.

If the sensor is not being used I think its better not to power it, just having it powered lowers its life cycles whether it actually cycles or not....IMO.
 
Thanks, Ron. Your last post gave me another idea which I'm content with (finally). Similar to my original 2-pole idea, but this time I'm isolating the sensor's output.

NPN Sensor Labeler
+---------+
| DC+ |-------------------------> 24VDC
| |
| |
| N |
| P DC- |----+--------------------> 0VDC
| N | | BYPASS
| | +--] [--+
| | | PERMIT
| OUT |-------]/[--+---] [------> INPUT
+---------+

Now, why didn't I think of this in the first place?... :oops:

🍻

P.S. I'd still like to hear any thoughts from others on my original idea about why it is / is not okay to do that... :nodi:
 
This is getting aggrevating...the post keeps messing with me. I was thinking more on these lines.

NPN Sensor Labeler

+---------+
| DC+ |---------------------------> 24VDC
| |
| |
| N | reg
| P DC- |-------] [--+--------------> 0VDC
| N | |
| | +--] [--+
| | | No reg PERMIT
| OUT |----+------------] [-------> INPUT
+---------+


Reg is a dual contact 2 position switch (or 2 pole relay) activates the sensor when in reg position 1 and turns off/bypasses the sensor when in position 2. The graphics could be more descriptive but you get what I mean.
 
Last edited:
Your drawing is what I tried to show in the last drawing of my first post. I should have grouped the contacts to show that they were both on the same switch. This might clarify:

NPN Sensor Labeler
+---------+
| DC+ |----------------------------> 24VDC
| |
| | BYPASS
| N | +-----+
| P DC- |------|-]/[-|-+-------------> 0VDC
| N | | | |
| | +-|-] [-|-+
| | | +-----+ PERMIT
| OUT |----+-------------] [-------> INPUT
+---------+

Note: Contacts are shown with the switch in the 'registered' (un-bypassed) mode.

My concern with this circuit is that, although the DC- of the sensor is not connected, I am connecting the output of the sensor to 0VDC.

When in bypass (non-registration), the sensor is essentially connected to the power supply like this:
NPN Sensor
+---------+
| DC+ |--------------> 24VDC
| |
| |
| N |
| P DC- |--- No connection
| N |
| |
| |
| OUT |---------------> 0VDC
+---------+

I am concerned with current flow from OUT to DC+ damaging the sensor... :oops:

🍻

-Eric
 
Last edited:
My original thought was break the 24vdc but I dont technically see a problem with there being "0vdc" on the output....techically the sensor has no power so can not bias therefore it becomes a plain old blocking diode.
 
I agree... there is nothing wrong with that little trick. I've done it myself and it has been on the machine for over 5 years now with now problems to the sensor or otherwise.

BTW... I asked the same questions when I did it.
 

Similar Topics

Hello everyone, I am not new to this forum but this is my first post. At work we are working with a simatic safety PLC. Altough this has great...
Replies
11
Views
6,277
Hi, I have a series of subsystems as parts of one machine that runs product. Transfer table (TT) moves product from one subsystem to another. TT...
Replies
2
Views
1,547
I was recently asked a question that I wasnt sure was the proper way to answer it. I was asked if there was a input sensor malfunctioning and I...
Replies
7
Views
3,770
I need to fire up some old DOS-based Westinghouse PLC software from circa '92 or '93. It uses a parallel port Rainbow Sentinel dongle, won't go...
Replies
1
Views
4,521
I have a new line reactor where we have let the smoke out of one winding. Its a new reactor and the phases are balanced. It appears to be a...
Replies
7
Views
3,458
Back
Top Bottom