Contact Bounce

Maybe you could apply latch trigger unit?

If its a slow change then a capacitor across the contacts might suffice?

Otherwise you will need a small de-bounce circuit, a TTL device provides this function for up to 6 (I think) switches.

Its basically a comparator which when it first switches is disabled for a set time so no other signals can reset it... Works very well.

John.

Isn't it a Schmitt trigger? Latch trigger almost the same. You need only to organize specially its unlatching after contact bouncing finish.

NO.... a Schmidt trigger input just has a set hysteresis before switching...

Contact bounce can easily switch off and on several times in several milliseconds, so making the schmidt output switch as well.

Yes, you're right. I just picked up my old textbook. Hysteresis may be too short to cover average bouncing period. Though I used one being student long time ago for hercon relay signal handling.

Thanks John

What size of cap would you recommend for a 10mS pulse?

Alan

T=RC

Time (Seconds) = Resistance (Ohms) x Capactiance (Farads)

Hope this helps, you may want to tune these to taste, but it's hard to beat a good de bounce circuit. Thanks for helping me to remember the early days of basic electronics school!

Good, nice and simple logix

John.

Thanks, that means a lot . It's hard to give the "illusion" of being smart when you have to prove it.

First try this: Connect 100MFD 63V electrolytic capacitor parallel to relay. If still problem is not solved then you need a circuit. Take care to check polarity.

Do not try this idea Bolton, the surge current from a 100 uF capacitor will weld the contacts at worst and make the contact life appalling at best.

John

Other than the already good advice you have being giving, two methods i employ when interfacing PLC's and automotive devices is use of an opto coupler (it has a slow turn on/off time which may suffice) Toshiba TLP627 comes to mind.This also isolates the circuits. Secondly the use of a PIC micro with a programmable bounce and de-bounce time (firmware adjustable) this gives enormous scope and range.

Hope this helps

Is this a homemade PLC?, I didn't think so...

I can't believe that any serious PLC manufacturer will build a defective unit. bouncing will occur as a result from bad design (hard to believe,) or some sort of a problem at this specific PLC. I assume that there are several relays, Do they all bounce?, That would be a power problem. only one does the trick? that will point to a problem at this individual relay or it's circuit. If I were you I would:

1. Try to replace the PLC with an exact same unit. see if the problem persists.
2. Check voltage across the relay coil. It might be low for some reason. or pulsating. Easy to see with a scope.
3. Replace the relay.
4. Check the circuit that drives the relay coil.

I have seen some cases where the switched load, if large enough, will with it's in-rush current cause a line voltage drop large enough to cause relays to open for fraction of a second, and run havoc. If this will be the case stabilize the PLC feeding line.

And if all else fails, than choose from one of the many good advises given here...

Wangito.

Wangito, all switch and relay contacts bounce or chatter, its inherent in the design.

Switching a light bulb or whatever isn't a problem but if the contacts are used to switch a logic gate, for example, then the contact bounce is a problem.

Its not a fault of the PLC, relay or anything, its the mechanical design of the contacts.

Wangito,

If I was going to replace the PLC unit may as well by a transistor output module.

Alan

Motorola made a debounce IC years ago, but I don't think it's in production anymore, and I don't remember the number, either. I generally use a NAND, set-reset latch to debounce contacts. Of course, there's practical problems with the application of such a circuit, such as the output is high when active and open-circuited when not active, but these things can be overcome.

Let me know if you need assistance, and I'll draw a couple of diagrams for you.

I think that must be the chip I was thinking of Bill... I used it for a keypad input to a computer and it worked superbly...

Haven't been able to find the type number for it though... I did think it was a standard TTL part - must be wrong?

I'm pretty sure it was a CMOS part with a 4000 series number, but it's been well over twenty years since I used them, so I can't swear to it now. If it wasn't for CRS disease, Alzheimer's would drive me crazy!

Thanks for everyone's suggestions but this is exactly where we are at

I tried putting C1 in at various valueswhich just caused a permanent connection???

Don't really want to put another power supply in for cmos is there anything else?

Alan

I think the most important issue here is how often the 10mS pulse is arriving - so we can advise on the delay we can get away with...

Also is the CMOS counter a high impedance input,? as that would mean it might be easier to de-bounce a CR time constant at that point...

Why use two relays to drive a low impedance input? most relays need a fair number of milliamperes to make sure the contact plating conducts - if you are switching a milliamp or less contacts should be of the highest quality gold plated and even then I wouldn't recommend it...

John.

10mS pulse is as often as a unit is processed - between 60 and 90 seconds apart

The two relays in the diagram are 1 internal in the PLC and one internal to the tact time display(SSR) of which the data sheet can be found here.

One other point I have a feeling that the display is faulty because if for example the pulse is extended to 1S then the counter adds about 500 to the total (have one of our Japanese brethren getting circuit diagrams from the manufacturers to sort that out). What this means is the pulse must be short.

Alan

Ohhhhhhhhhhh Bolton that solid state relay is for ac or dc volts only above a certain value...

I doubt very much whether it can switch a practically zero current voltage...

Hold on I must check that again....

YES! as I thought the minimum on current is 20 mA and the off state leakage current is 1 mA

If you're switching a CMOS input you really need to have a resistor to supply enough current for the Solid state relay to switch... In fact it would be better to use a reed relay instead...

John.

There's usually V+ and Com outputs on most (if not all) PLC's I've ever used. What brand name and model is the PLC?

Allen Bradley SLC500 with 1746 OW16 output modules

Bolton,

To start with, I don't know of a module to plug in directly to an Allen Bradley SLC 500, but I have made special modules for a variety of PLC's. You say you have a 24 volt supply. I assume it's a DC supply. Sometimes it takes a little effort, but plug-in module bases are generally available for custom applications from the manufacturer. By the way, I did not find such module bases.

I did find, however, 1746-OB32 and 1746-OV32, source and sink (respectively) output modules which operate from 5 to 50 VDC, found at Allen Bradley and the installation datasheet (stating that they are transistor outputs).

There are ways to debounce circuits, and one of the sites showing such ways is found here (it easier for me to send you there instead of recreating similar drawings and pasting them in this post).

An NPN transistor with a pull-up resistor can be substituted for the amplifier in figure 2. The circuit in figure 3 is one of the NAND gate debounce circuits to which I referred in an earlier post.

Good luck!

If you are using an output relay from a plc, you will be able to programme a de-bounce. Assuming there is enough time between the pulses. The problem may be the meter is not quick enough for the pulse train?

No I won't! The contacts will still bounce when they close no matter what is in the program. If it was an external relay bouncing to an input then yes I could

Alan

Please check the 24v output is used twice in the PLC ladder diagram or PLC diagram may be faulty

And having done this and found it to be fine how might it have stopped/caused mechanical contacts bouncing?

The problem is nothing to do with the PLC or the program other than the fact that it is a relay output module and it uses 24vdc which means I can't use a cmos circuit

Alan

Cracked it folks.

Due to another issue which I hadn't mentioned used an Omron H5CX-L8SD timer in one shot mode for 5mS. Transistor output so no bounce

Data Sheet

Thank you all for your contributions

Alan