Capacitive Coupling ???...

Status
Not open for further replies.
SG-1

SG-1

Senior Member
Location
Ware Shoals, South Carolina
I have a 120VAC control circuit using a Magnecraft W388ACPSRX-30 Time Delay Drop Out Timer.

The 120VAC signal wire is between 30 & 60 feet long. It is routed in wire bundles that are about 2 to 3 inches in diameter.

When the contact opens to de-energize the timer a pesky 10 volts AC remains across the coil. This is enough to keep it latched. :-?

I am measuring the 10 volts with a Fluke 27. If I insert my SV225 to convert the meter to a low impedance input the relay drops out when my probes touch the terminals of the coil.

I can also place a 25K ohm resistor across the coil to make it behave. This is about 10 times the resistance of the SV225.

Is capacitive coupling a possibility ? I cannot find mis-wiring.

The relay uses 2W max @ 120VAC
The Fluke measures the coil resistance at 10M ohms.

Thanks in advance for any help.
 
Well yes it can .

And also not just capacitive but could be a voltage induced in the bundle if ac is present in the wiring of that group .

If its a coil it can work both ac or dc try a diode & capacitor across it .

See if that makes it let go .

Its called a snubber circuit we used it on a generator once to eliminated false starting by inductive coupling thur wiring in a common raceway .

We always run controls away from normal ac wiring mostly today for that reason even low voltage and data controls you never know what it can pick up with mixed controls in one raceway or basket of wiring.
 
Last edited:
Is capacitive coupling a possibility?

Yes, it is possible... and quite likely the culprit.

There may be internal diode(s) on the input, and your DMM does not have enough voltage to exceed the forward bias voltage necessary to get an accurate reading of the coil's DC resistance... thus the 10M-ohm reading.
 
100314-1044 EST

The relay you specified is rated 10 MA @ 120 VAC input, or approximately 12000 ohms and may pull-in at about 60% of this. See http://www.magnecraft.com/library/archive/103_Section4.pdf page 4-17. I did not find the pull-in value so it is a guess.

If 60% is valid, then 6 MA would be sufficient for pull-in. Drop-out might be lower.

If other wires in the bundle are 120 V and we assume your 10 V is sufficient to hold the relay in, then about 110 V is across a series impedance to drive the 12000 ohm input. A resistance of 11*12000 = 132000 ohms from 120 V lines to this input wire would hold it in. Assume this was capacitive reactance, then this is equivalent to about 0.02 microfarads. It looks like about 0.8 MA is holding in the relay. Seems unlikely.

If I take a somewhat comparable basic relay at 24 VDC it drops out at about 5 MA. Divide by 6 an it is 0.8 MA. Maybe it is not too unlikely.

At 20 pfd per foot between two closely spaced wires and 60 ft we have about .0012 mfd per wire. So many wires might get us to 0.02 mfd, or higher voltages on some of these wires. You might also have resistive leakage current.

Put a 100 ohm resistor across your time delay relay input with no energization. Measure the voltage across this resistor to determine current. Next put a scope across the resistor and check its phase angle relative to AC supply voltage in the wire bundle, assuming it is all single phase. This may allow you to determine if it is capacitive or resistive current.

I seriously doubt that you have inductively coupled voltage of this magnitude unless there are very large currents in the bundle.

.
 
Gar,
That data sheet indicates it is a latched relay with an electronic timing circuit.
The 10 ma operating current may be for all the electronics ( pins A to B?) and may not be a good figure to use to calculate the actual input impedance of the pin 8 input used for an off delay function.

SG1,
Can you measure the input current to pin 8 under normal operation?
Am I understanding correctly that your control wires are from pins A to 8?
You are switching this input using a relay contact and not a solid state device is that correct?
 
100314-1351 EST

ELA:

I think with SG's residual supply voltage that this time delay relay never drops out of its on state. Does the electronics require something near the 10 MA specified?

My KUP11D15 24 VDC requires 30 MA to pull in. This is double the current spec for the Mangecraft 24 VDC unit. My guess is that most of the current spec is for the electronics, or maybe just an input load resistor, and that there may be a peak inrush current pulse to latch the relay. Once input power is removed there has to be enough stored energy to unlatch the relay. But this would also mean that the electronics for timing can not consume much power, and therefore why would so much steady state power be required during the period after initial turn on and until the trailing edge of the input power.

There need to be more experiments on the relay alone to learn what are its detailed characteristics.

.
 
Gar,
Is your KUP11D15 24 VDC a off delay relay with internal electronics?
If not then why compare to it?

The Magnecraft relay does not need to store energy. If you look at the wiring diagram it shows that input power remains intact in order for the off delay function to work.

You interrupt power to pin #8 only.

So my point is that you need to measure the input current to pin8 in order to estimate its input impedance. This is the impedance of a input trigger circuit (not to be compared to a relay coil drop out voltage).
 
100314-1510 EST

ELA:

The KUP is a P&B standard DPDT 10 A relay. It would represent the spring forces for the 10 A contacts, and approximate energy to close the magnetic circuit.

My mistake I did not look at the off delay diagram. The control input could be a much higher input impedance. Thus, much less capacitance than 0.02 mfd might keep it on, and as you pointed out if the source to pin 8 is a solid-state relay its leakage current might keep the relay on.

The solution to correct the problem is probably a 10 K 5 W wirewound resistor from terminal 8 to B. 5W instead of 2 W because it reduces the surface temperature. An IRC PW-5 is an inexpensive axial lead device, 3/8 x 3/8 x 7/8.

.
 
ELA said:
...
So my point is that you need to measure the input current to pin8 in order to estimate its input impedance. This is the impedance of a input trigger circuit (not to be compared to a relay coil drop out voltage).
Click to expand...
If the drop out contact at the far end of the control wiring has a normally opposite mate (i.e. double-throw NO/C/NC), or an equivalent, perhaps the answer is to just sink the wire to ground (or whatever "B" terminal voltage is) upon the trigger event.
 
ELA said:
SG1,
Can you measure the input current to pin 8 under normal operation?
Am I understanding correctly that your control wires are from pins A to 8?
You are switching this input using a relay contact and not a solid state device is that correct?
Click to expand...

The input switching is being done with MOCs ( mechanically operated contact ) from a MV circuit breaker cell.

All the wiring through those harnesses are for control. They are inside a swirchgear cabinet. No 12470 applied to the bus yet.

I have put the relay on the table for an experiment as suggected.

If I use the circuit breaker to turn on the 120VAC then the relay acts normal.
If I slowly turn down the voltage it drops out at about 50VAC. Pick up was about 60 VAC.

If I suddenly turned the voltage down. It would stay latched right down to zero. It takes a very specific rate of decay, but I found it after a few tries. It is very repeatable. I am using a 3 volt bell to detect the change in contacts.
 
Smart $ said:
If the drop out contact at the far end of the control wiring has a normally opposite mate (i.e. double-throw NO/C/NC), or an equivalent, perhaps the answer is to just sink the wire to ground (or whatever "B" terminal voltage is) upon the trigger event.
Click to expand...

All the MOC contacts are simple "a" or "b" type.

I am using a 25K ohm resistor now across the coil A & B to allow me to continue with the rest of the circuit.
 
Where afore-time I was thinking capacitive coupling I now clearly see the rate of decay across the coil can stop the device from dropping out. Thumping on it with my finger did not help any either. I had to reapply the 120VAC and then de-energize it.

New question: Can the wire length provide that same rate of decay ? Looks like it can, as of now.
 
ohmhead said:
Well yes it can .

And also not just capacitive but could be a voltage induced in the bundle if ac is present in the wiring of that group .

If its a coil it can work both ac or dc try a diode & capacitor across it .

See if that makes it let go .

Its called a snubber circuit we used it on a generator once to eliminated false starting by inductive coupling thur wiring in a common raceway .

We always run controls away from normal ac wiring mostly today for that reason even low voltage and data controls you never know what it can pick up with mixed controls in one raceway or basket of wiring.
Click to expand...


This timer is a little more complicated than a simple coil. This device used to be an agastadt, no problem.

It is listed for 120VAC. The breakers have a 125VDC control circuit. Then there is the 120VAC control circuit. Both schemes have many contacts in the other scheme. Both schemes operate simultaneously. Kind of like solving simultaneous equations in school only a lot louder.
 
SG-1 said:
Where afore-time I was thinking capacitive coupling I now clearly see the rate of decay across the coil can stop the device from dropping out. Thumping on it with my finger did not help any either. I had to reapply the 120VAC and then de-energize it.

New question: Can the wire length provide that same rate of decay ? Looks like it can, as of now.
Click to expand...

I asked if you are using pins A to 8 as your control input in a previous post?
Can you please confirm this?
That is required for the off delay to function properly.
 
ELA said:
I asked if you are using pins A to 8 as your control input in a previous post?
Can you please confirm this?
That is required for the off delay to function properly.
Click to expand...

Thanks for your persistance.

Sorry missed that.

No we are using A & B. to simply de-energize the device. I will turn my attention to that immediately.
 
SG-1 said:
All the MOC contacts are simple "a" or "b" type.

I am using a 25K ohm resistor now across the coil A & B to allow me to continue with the rest of the circuit.
Click to expand...
Form A contacts are NO.
Form B contacts are NC.

You can effect a SPDT by using one of each and making one terminal from each common. Typical such arrangement would be, but not always, "beak before make"?which would be ideal.

Is the "operator" two contact capable?
 
100314-1628 EST

ELA:

I think I see where my confusion was. I do not see the full picture on the screen as I would on a piece of paper.

If you look at SG's part number W388ACPSRX-30 and I originally found this on page 4-17. If you go to this page it is classified as a true off-delay timer, and has only one circuit drawing on p 4-17. This unit has no control input. So I am back to my original analysis that energy is stored to use to unlatch the magnetically latched relay.

This relay should drop out after its time delay following the trailing edge of the AC applied to terminals A and B. Probably this needs to be a somewhat abrupt turn off. But there seems to be a problem from what SG said about abrupt reduction of input voltage.

.
 
SG-1 said:
There is no point 8 to wire on this device. That whole center row is empty.

W388ACPSRX-30
Click to expand...
Hmmm... that's most interesting. Your's is a TDR-OFF unit??? Most have some external control terminal. I suppose it don't need one if internal circuitry is capable of keeping the coil energized with power removed from the input terminals... but it is contrary to what they show (pictured below) in the pdf gar linked above. The model number of the one pictured isn't even available on the page, perhaps the entire catalogue, and from browsing their website, the 388 series must have been discontinued.... :confused:
 
Status
Not open for further replies.
Top