relay wont close consistently with open collector input

[jmacion]

Hello,

We have a device we are interfacing with that has an open collector input which are labeled as "pin2" and "INTLK" shown below. We are trying to activate a relay coil when the open collector pulls the coil down to 0V when its active. The problem we are having is that on about 1 out of 25 times we activate it, the RLY5 coil get stuck in the NO position and we can get it to return to NC even when powering down the whole system and then back on. Its confusing as we would think reset to the NC position? We know its not fused as periodically it resets itself and starts working again for another 25 or so runs.

We have measured the impedance of the open collector by putting a 2kohm resistor in series with it. We measured 8VDC over the 2k resistor which means the open collector is around 4kohm (has 16VDC drop over it). We thus assume that we cannot use a relay with a 24VDC coil (must activate at 19VDC) since we have a 24VDC rail and the drop over the coil will not be enough. We did replace the 24VDC relay with a 12VDC relay coil shown below and are having the same problem. What is very odd is that the original configuration was a DIN rail 24V coil Relay and this works 100% of the time, the part number is shown below.

Parts:
G5V1DC12: coil: 12.5mA/12VDC, R= 12/12.5=960ohm, must operate at 9.6V
G5V1DC24: coil: 6.3mA/24VDC, R= 24/6.3=3.8kohm, must operate at 19.2V
G5V1DC5 - we didnt try this one but shouodnt have to go any lower?

This is the DIN rail relay. I dont understand why it always works (for 6+ months).
Phoenix 2903370, 9mA/24VDC coil, R= 24V/9mA=2.66kohm. Minimum switch=5VDC

We have been building this circuit over and over and it still having intermittent behavior. We have built At least 4 PCBs with different parts changed out on each with no luck. There is no way the parts are damaged due to soldering as we have parts even on the PCB. We need help.

I am ready to break down and start crying on this one and return my college degree. Any help is appreciated.
Do we need a common emitter configuration to drive a relay (this cant be done as we dont control the source)?

thank you
Jim and Reshma

 

[BrianE22]

Are PIN2 and Interlock actually getting to zero volts?

 

[jmacion]

yes, they are getting pretty close to 0, around 0.2VDC

 

[LionelHutz]

So, the relay turns on but won't turn back off even with all power removed from the circuit? That points to a mechanical relay problem, not an electrical problem.

 

[jmacion]

it Lionel - its done it consistently in about 10 different relays on a total of 4 PCBs.

I think the coil not getting enough voltage to move to its primary position and then cant move back to the NC position.

 

[LionelHutz]

It's one or the other. Does the relay switch and close 10 to 5/6 when the coil is powered or does it stay with 1 to 5/6 closed?

If 10 to 5/6 closes and then won't open again even without power it doesn't sound like an electrical problem to me.

 

[3DDave]

I recall hearing of a similar problem once - turned out the relay was magnetizing a steel plate next to it and the field in the plate kept the relay from returning. The ability to retain position was partly dependent on the amount of vibration in the machine.

 

[jmacion]

It stays in the NO configuration without power which is surprising as NC is the default which is expected with power off.

We use this relays many PCBs designs for decades without problems. This is a new design that interfaces with an open collector which is new.


Its something about hte open collector causing this problem, its not powering fully and this is causing the end result. I dont want to focus to much on the end result as I think the problem is how its powered electrically.

I think we may need a FET but I rememvber having problems with the FET. Tere has to be a way to do OC on relay coil.

 

[jmacion]

Dave - Interesting problem. WE dont have any magnetics nearby. This is very simple.



I think the problem is how to use the Open collector to power the relay.

 

[3DDave]

Well an electrical problem that is intermittent and not component related is certainly not a common one. You didn't mention if it was the same open collector driver and multiple boards or every board has an integral open collector driver on it.

 

[3DDave]

By the way - if it is a problem with the open collector you need to tell us what that is.

You also haven't put any effort into measuring the current during operation; telling specifications doesn't help without that information.

 

[BrianE22]

The NC contact isn't welding, is it?

 

[jmacion]

The common collector inputs are part of a COTS system. We have 5 of them and they all do the same thing.

I think the issue that the PCB mount relays just do not work with a common collector configuration as the "turn on" voltage is just too high while the DIN rail relay one turns on a very low voltages. Even with a 5V PCB mount relay, it will still not have voltage to turn on. There is something about it not turning on all the way where it gets stuck in NO position.

I think we need a PCB mount SSR or something else or ortherwise will have to use the DIN rail mount parts to interface with the common collector and then bring the outputs into the PCB.

 

[LionelHutz]

the open collector is around 4kohm (has 16VDC drop over it)

they are getting pretty close to 0, around 0.2VDC

What voltage is on the relay coil when the relay coil is powered by the open collector?

It makes no sense for the relay to stick on without power "due to the open collector". The Phoenix relay is 6A rated and the Omron relay is 1A rated. Are you welding the contacts?

Why don't you just switch to a relay that matches the Phoenix one?

 

[jmacion]

Hi Lionel
the phoenix is a DIN rail mount and the ones we are having problems with are on the PCB. Thery are much smaller.

I think we build another PCB that does not have the problematic PCB mount relays though and then use the DIN rail PCB as an interface.

we are sure at this point it has to do with under powering the relay. It is even making a funny sound when we try to turn it on.

I think we are going to try a SSR or antoher technology that is lower current burden on the open collector.

 

[LionelHutz]

I can't see the relay that's plugged into the DIN socket being much bigger.

 

[hacksaw]

Have you checked the time variations of the voltages with a scope?