tattle tale

There is never an exact same time... that is impossible

That said how fast it this thing? I have no idea... how fast is electrcity? again I have no idea (even though I am a genius)

I am going to meet with Ron in the near future so I may have to lend it to him for some bench testing

for those who still doubt ...

gosh ... I thought we were through with this ...



again, I apologize for not making myself more clear ...



moving forward ... I still say that my original “all go off at the same time” test is perfectly valid to prove the concept - but for simplicity’s sake, let’s nail this thing down with an apples-to-apples experiment ... I just used some of my student wiring lab equipment and hooked up EIGHT Normally Closed pushbuttons in series with each other ... I used these in a common holding circuit to “seal in” a real world contactor ... here’s the schematic ...








I had a MicroLogix1000 handy with TEN inputs ... I connected input 9 to monitor the AC Line ... I connected inputs 8 through 1 to monitor the “stop buttons” ... I connected the last input (0) to monitor the contactor’s auxiliary contacts ...



and here’s a couple of snapshots of the setup ...









when I energized the contactor, ALL of the Micro’s inputs lit up ...



and here’s the quick “down and dirty” program which I used to record the status of ALL of the input bits as soon as ANY bit went OFF ... (note: if anyone has another program they’d like to try, just post it) ...








with everything ready to roll, I pushed “stop button C” and naturally the contactor dropped out ... I recorded the data ... I reset the contactor ... I reset the trap bit ...



I pushed “stop button C” and the contactor dropped out ... I recorded the data ... I reset the contactor ... I reset the trap bit ...



I pushed “stop button C” and the contactor dropped out ... I recorded the data ... I reset the contactor ... I reset the trap bit ...



I did this same test ONE HUNDRED times in a row ... here’s the data that was recorded ...







now even though I pushed the SAME button (C) EVERY SINGLE time, the data that was recorded was NOT always the same ...



I’ve highlighted the actual bit (6) that caused EACH and EVERY drop out condition ... a “good” test would have given a decimal value of 896 ... and most of the time (81%), that’s what was recorded ... BUT ... (and here’s my point) the data was NOT ALWAYS correct ... specifically, 19% of the time a RANDOM bit pattern resulted ...



for a specific example, take a look at the data in test #44 ... that one is especially troublesome ... if we had never covered the issues that I’ve brought up in this thread, what conclusion would most reasonable technicians come up with while considering the data from that particular test? ... I submit that the most logical conclusion would be that the contactor’s auxiliary contacts are causing the contactor to drop out ... but that perfectly logical conclusion is WRONG - because the ONLY button that I ever used to drop out the circuit was button C ...



now an electrically knowledgeable technician might look at all of this ERRONEOUS test data and say: “Wait a minute. If a midstream bit goes OFF, then why don’t all of the bits downstream of that bit go OFF also?” ... bingo! - that’s a very good question ... now how are we going to explain the various “random bit patterns” in those erroneous tests? ... I submit that you can NOT adequately explain those data readings WITHOUT understanding the concepts that I keep harping on ...



regardless of how you choose to slice it, my point has been made ... again ...



personally, I think that I said it best back in post #36 ...

in a nutshell: PLCs are fast - but they are NOT instantaneous ... and (KEY POINT) some of their inputs may react quicker than others ... many (most?) people don't know that ...

Click to expand...

in the spirit of the forum, I’m just trying to make more people aware of a POTENTIAL issue that many (most?) technicians never even consider ... anyone who doesn’t believe what I’m saying, is perfectly welcome to set up the test and run the experiments for themselves ... I assure you that I am NOT making this up ...



my question is this: how on earth would you program the PLC to interpret this RANDOM data and come up with an accurate assessment of which was the “first out” bit? ... (and before anyone suggests incorporating a “filter” or “settling” or “debounce” arrangement, please consider how that might adversely affect the ability to spot a defective contact which is only very briefly and intermittently open) ...



now as for the “ready made” Tattle Tale that Geniusintraining kindly showed us in post #44 ... I really do NOT know how that’s going to work ... maybe it will work PERFECTLY ... I really hope that it does ... but as GIT said:

I am going to meet with Ron in the near future so I may have to lend it to him for some bench testing

Click to expand...

good to hear that you’re finally coming down this way, Mark ... by all means bring that thing along with you and we’ll test it out while you’re here ... I’ll post the results on the forum as soon as the test is done ... personally I wouldn’t bet more than pocket change on the outcome one way or the other ... but on the PLC Tattle Tale being discussed in this thread - well, that’s a different situation ... I still maintain that a “First Out” tester based on a PLC is not going to be 100% reliable ... it might work well enough to be useful - but if more than one input can electrically change state simultaneously, then personally I wouldn’t bet the rent on its accuracy ...

Ron Beaufort said:

... I still maintain that a “First Out” tester based on a PLC is not going to be 100% reliable ... it might work well enough to be useful - but if more than one input can electrically change state simultaneously, then personally I wouldn’t bet the rent on its accuracy ...

Click to expand...

Ron, thanks for taking the time to test this. I had been thinking of trying to find time at work to try it out. My ML1000 test unit performed just about like yours, although your method of testing was much more thorough. I only used mine a handful of times on a variety of real machine problems as opposed to your 100 test runs.

As a tester, it was much better than trial and error, and the best thing we had to use at our disposal.

Pointing out the exact reasons for its drawbacks is a great exercise in understanding how a typical PLC operates and I am sure many will benefit from your experiments, and especially your detailed write-up.

Thanks again,
Paul

Eureka!

does anyone else have those experiences where you go to bed and your head keeps working all night even while you’re asleep? ... my head does that all the time ... my wife can’t understand why sometimes I wake up more tired than when I went to sleep ... anyway, at 4:00 o’clock this morning I woke up with the following schematic perfectly drawn out in my head ...

suppose that we use ISOLATED input bits for the PLC “First Out” tester ... that way we can wire them up like this ...





and now at the instant that our “problem” switch (C) opens up, the ONLY bit that will have a voltage (a difference of potential) across it will be bit 6 ... that should make it a lot easier to capture the changed status of that single ON bit and accurately record it as the “First Out” culprit which dropped out the relay ...

disclaimer: I haven't had time to actually test this approach out, but I can't see why it wouldn't work ... I'll try to test it tomorrow and post any significant results ...

of course the MicroLogix1000 we’ve been talking about isn’t going to give us the isolated input wiring that we need – but as long as we can come up with the correct hardware, then I’ve got a very strong hunch that we could get this thing to work reliably even with a PLC ... as I kept saying, it’s not the PLC program that’s been causing the problem, it’s the unpredictable response of the PLC inputs when two or more of them electrically change state simultaneously ... using isolated inputs should eliminate the problem – since only ONE input will electrically change state at a time ...

now then ... I looked (briefly) on the North American Manufacturing website for the ready made Tattle Tale that Geniusintraining brought up – but I couldn’t find it listed ... but even without seeing how the thing is designed, I’d bet significantly more than pocket change that it uses ISOLATED inputs similar to what I’ve drawn above ... and in that case, I’ll bet that it’s going to work just fine when we get around to testing it ...

and to OkiePC ...

Pointing out the exact reasons for its drawbacks is a great exercise in understanding how a typical PLC operates ...

Click to expand...

I’m glad that you see it that way, Paul ... that was my main reason for bringing up all of this voodoo in the first place ... I’ve had to help debug one or two systems over the years where these concepts were an issue ...

once in awhile you’ll hear an old-time hardwire electrician talk about a control problem caused by a “relay race” or a “contact timing” issue ... I guess this is the solid-state PLC equivalent of that type of situation ... the random and intermittent nature of the beast makes it pretty hard for most people to track down a problem caused by this effect ... as with most things, once you’re aware that something like this exists, it gets a lot easier to troubleshoot ...

Ron Beaufort said:

now then ... I looked (briefly) on the North American Manufacturing website for the ready made Tattle Tale that Geniusintraining brought up – but I couldn’t find it listed ... but even without seeing how the thing is designed, I’d bet significantly more than pocket change that it uses ISOLATED inputs similar to what I’ve drawn above ... and in that case, I’ll bet that it’s going to work just fine when we get around to testing it ...

Click to expand...

First off thanks to all for providing a very informative thread. I was unable to find the item at North American MFG but did find the one referred to in post 20.
http://www.1212designs.com/FOI_Compact/default.htm
and as you stated it does use Isolated Inputs.

I think I will order one and see how it works.

thanks again.
bob

Plan B

First off, thanks Ron very informative. I have seen that problem before and never actually knew why it was happening. Since I started as an electrican I went back to what I knew and used a meter. A fluke 85 series meter to be exact.

If you have one of these handy little yellow boxes in your tool box and it has fresh battries (this can be very important). You can check series circuits like this. Basically what you want to do is put your meter in parallel to the suspect switch. Switch meter to whatever voltage the system is using. You will notice a zero voltage reading on your meter. As long as the switch stays closed that reading will stay zero. If the switch opens up then you will get a voltage reading.
No nobody in their right mind is going to stand there and watch the meter. That is where a function on the mter come in. It is the Min Max function.

Remember when we hooked up the voltage was zero because the switch was closed. Well that means when we turn on our min max function the highest voltage is going to be zero (now for those that see some minor mili volt signal that is ok also, we are looking for a bigger change so we are going to call the noise reading zero). Ok now we wait for the event. Event happens, what does our meter read?

Well if Max reads zero then you are on the wrong switch, move to the next one. If it reads the voltage on your circuit or something close to it then you have found your grimlin.

While this does not check all the switches at the same time, It does not lie. I think it could work in conjunction to your relay checker and confirm your findings.

Note: To hamper an interlock (series circuit) you do not always have to have a complete break in the circuit. Depends on the minum voltage required for your coil or input.


Note 2: You can use the divide an conquer with the method described above to narrow down your search. Basically just pick a section of switches and go across those. If you get a zero reading after the event you can elimate those or if you get a voltage reading then the bad componet has to be one of them.

Fyi

You are correct American Manufacture does not sell this anymore... not sure why (did not work very well??)

any way here are a few pics of the panel on the inside.. they list 3 different ways of wiring up their device, I hope the pictures are ok (from my phone)

..and here’s the quick “down and dirty” program which I used to record the status of ALL of the input bits as soon as ANY bit went OFF ...

Click to expand...

Ron, I would like to see what happens if you use the method that have used before. I use a different condition as the trigger. Instead of looking for the First changed input (a losing proposition as you have proven several times now), look at the state of the contactor. Use only the state of "Input 0" in your program, the one from the contactor seal-in. When that one drops out, it means the contactor has had time to physically drop out, which means that several PLC scans have passed. That allows enough time for the actual open switch to be identified, as the inputs should settle out by then. Record the other input bit states only AFTER Input 0 goes off. Then compare each bit to see which were still closed. The first non-closed bit is the first place to check for a problem.

Greetings Lancie1 ...

I'd love to try out your idea ... is there any way that you can help me out by posting the exact code that you want me to run? ... just write it for a MicroLogix1000 with 10 input points and post it ... you can follow the wiring that I drew out in post #47 ... I'll keep the existing wiring hooked up until I hear from you ... or at least for several more days ... I've got to straighten up the lab sooner or later after my last on-site training gig ...

got to go ... the Red Cross called up and they need a dose of AB+ ...

Ron,

I will try to write a Mircologix program for your test setup, but make no guarantees. I am having some health problems right now and have trouble concentrating for more than a few minutes at a time. It seems my thyroid is working overtime (a so-called "hot node"), which tends to make me nervous, irritable, and with the energy of a wet dishrag.

take care of yourself ... we can work on this later ... my back is to the wall with year end business matters - and stuff like that ...

hope you're feeling better soon ...

Ron,

I managed to concentrate long enough to modify an existimg program. I tried to match your wiring. Please test it. Perhaps someone might have a better method.

Correction for Burner First Out Reset

I see now that my brain wasn't thinking as clearly as I thought...How can one know at the time?

I tried to put in a reset for the first out, but unfortunately it will reset the N7:0 word every time CR energizes. The original program has several functions that reset the First Out, and anyone who uses this will have to make sure s/he has the proper input on Rung 004 to reset it.

I'm new here but as it looks like Gary narrowed it down for you already. In my pre-PLC days we used the relay solution either having the relay open or close for the tattle tale.

With PLC's I now consider it good practice to have the downstream conductor of any chain, usually the E-Stop chain go to an input with all of these inputs to the same word. Then mousetrap which opened first in logic. Current design typically uses a second contact for annuciation but that provides no path to troubleshoot when the chain is faulting do to a failing device.

On a related item is an intermittant short to ground that blows the primary fuse. If it does not violate safety put a suitably sized incandesent lamp in place of the blowing fuse and when the lamp gets brighter via the short go search for the curcuit thats dragging it down.

RSS

I hate to dredge up this old thread - but I just got a call from a new member who doesn't have RSLogix500 and can't open the file I attached to post #25 ... so I'm attaching it as a PDF below - and also including a copy in PLC-5 format ...

Craig, if this isn't exactly what you need, let me know and I'll try again ... I enjoyed talking with you ...