Underurrent relay

[ptonsparky]

I bought a split core undercurrent relay and am having a hard time getting it to switch consistantly. It has a solid state N/O output that is being used as an input to a micrologix 1000. The relay is rated at 1 A 250 volt but I am not sure if the input on the plc is enough load to make it switch. I know I have had to load up prox sensors when used as inputs. Clear as mud even for me but the Droid is dieing sontje best inan do for now

 

[milldrone]

It has a solid state N/O output that is being used as an input to a micrologix 1000. The relay is rated at 1 A 250 volt but I am not sure if the input on the plc is enough load to make it switch.

Yes this is a common problem. The challenge is to find the right value of "burden resistor". In the past I have used some SSAC brand current detectors that were very bad at current leakage. These units were "two wire" (didn't use a neutral) and needed 3.3K ten watt resistors to "pull down" the output. The resistors got very warm so beware!

 

[gar]

110917-1339 EDT

ptonsparky:

Need more information on this current sensing device.

Is it two wire?
How is the contact rating defined? AC, or DC, or AC-DC?
Any minimum supply current requirement?

An assumption by me, it is AC only, and is a two wire device. For this device to work it has to have power from somewhere. The only possible source is the two wires. Thus, when the switch is off there must be current flowing thru the two wires.

Run this experiment:
Connect the device in series with a 25 W incandescent bulb. When cool, not bright light, the resistance may be about 4 * 10 or 40 ohms. Actual on samples -- 15 W 80 ohms, 25 W 44 ohms, and 40 W 26 ohms. Calculated at 120 V input 15 W 960 ohms, 25 W 576 ohms, 40 W 360 ohms.

Next you need a variable current to pass thru the current sensor in the device. Adjust this current up and down near the desired threshold point. Do you get reliable switching operation? When the device is activated, output contact closed, the 25 W bulb will be bright and about a 0.2 A load on the device. When the device is off the 25 W bulb will have little voltage drop across it. When the device is on, then virtually 120 V - whatever is required to make the electronics work.

.

 

[ptonsparky]

AccuAMP from AD. Two wire AC current switch. The LED that is supposed to flash during operation blinks quickly when the motor is starting but doesn't change in speed while setting the trip point.

It is a multi range split core. We are using the 40-200 range. Motor idles at 50 amp and bounces between 75 and 100 during operation. We need it to shut down at idle amp.

 

[cadpoint]

If the motor is idling at 50 amp, would you need a two stage starter to bring it back to 50 before your fire it to 75 or 100 amps?

Seems your could find a Ugly book and duplicate something from in there.

 

[ptonsparky]

5p amps means the coupling to the hammer mill broke or the tanks feeding tbe hammer mill ran out of product. Either way, something else plugs up or dumps many gallons of oil additive all over the place.

 

[milldrone]

AccuAMP from AD. Two wire AC current switch. The LED that is supposed to flash during operation blinks quickly when the motor is starting but doesn't change in speed while setting the trip point.

It is a multi range split core. We are using the 40-200 range. Motor idles at 50 amp and bounces between 75 and 100 during operation. We need it to shut down at idle amp.

I have not used the unit you are using but I have used an analog out version. One problem I had with the split core was if the core was under any mounting stress (like a wire pressing against the side of the coil). The latch would open up and it would drive the analog (4 to 20) very high. So my first recommendation is to check if the latch is completely closed. I ended up just needing a wire tie to keep it closed.

If your latch is in good shape then you might try a second wire pass / loop through the unit. It sounds like you are near the lower end of the range for this unit, and by passing another loop through the core you would double the reading. The problem with this scenario is that it will put you near the upper end of the range.

 

[ptonsparky]

Good on the split. Not enough room for double loop. I have grandkids here or I would have test stand set up. Just not enough time for everything

 

[ptonsparky]

Didn't matter if I ran it thru an ice cube relay or the plc input. Results are the same. At least for what I had at the shop to play with.

 

[gar]

110919-2344 EDT

ptonsparky:

Did you try a a 25 W light bulb as the series load on the current relay output?

To get a reasonable sensitivity for an easy variable current put 20 turns thru the current sensor. Now 5 A is equal to 100 A. Light bulbs and heaters can be used as a current sources. 250 W (2.08 A) would be equivalent to about 41.6 A. Adding another 100 W would bring you to about 58.2 A equivalent, etc.


.

 

[ptonsparky]

110919-2344 EDT

ptonsparky:

Did you try a a 25 W light bulb as the series load on the current relay output?

To get a reasonable sensitivity for an easy variable current put 20 turns thru the current sensor. Now 5 A is equal to 100 A. Light bulbs and heaters can be used as a current sources. 250 W (2.08 A) would be equivalent to about 41.6 A. Adding another 100 W would bring you to about 58.2 A equivalent, etc.


.

No, I did not. I have a VFD on a small Axial flow grain fan that we play with. Not a 25w lamp in the building. I set the trip point of the monitor then changed the speed of the fan to get it to trip. Worked on the low range with no loops at about 2.1 to 2.3 amps.

The axial fan draws more current when the inlet is blocked, so I tried setting the relay to trip, go undercurrent, when I would remove the blockage. Not enough difference in current flow or not enough sensitivity in the adjustment, depending on if you are selling or buying.

I did not try looping to see if it would work at the higher end of the range or allow me to jump up a range. Maybe today.

 

[jim dungar]

I am late to this discussion, but many split core CTs do not play well with the PWM output of VFD's.
In a past life I sold devices from NK Technologies, I believe they offered a specific device for drive outputs.

 

[gar]

110920-1143 EDT

Tom:

My first reference to a 25 W bulb was that this bulb and the current sensor output would be connected in series. In other words the bulb is what is being switched by the sensor. Using this type of load provides a low enough series resistance that enough power should get to the current sensor to power its electronics. It also serves the purpose of indicating the state of the sensor. You would also be able to sense half cycle cycling of the output Triac.

Then in the last post I was suggesting a way to simulate a high current thru the current sensor by a 20 turn coil and using light bulbs to generate the current. I would prefer to use a Variac to produce a variable load, but I assume most people won't have one.

jim dungar brings up a good point about high frequency noise on the current from a VFD causing inconsistent operation. But before this consideration I would want to know if the current sensor works as expected on a nearly pure sine wave.

.

 

[ptonsparky]

I suspected there may be some problems on the load side of the VFD so I have the current sensor on the line side. Hopefully it isn't quite as bad there.

The sensor may very well be working within its parameters, just not what I had hoped for.

 

[ptonsparky]

Tried a N/C output sensor instead of N/O and added a couple more loops. Stuck an Amp Clamp in to see the equivelant reading and to verify that I had the proper scale selected. Played around a bit and it works, at least in my shop. Proves undercurrent at about 52 amps and run current at 70 on the high range.

Any reason I should not just loop the existing 300:5 leads thru mine to get a bit higher up in the range?

 

[gar]

110921-2244 EDT

Tom:

I had not looked at a datasheet for your current sensor before my previopus comments. If I am looking at the correct data sheet, ACS200, then this device is a current sensor that gets its power to operate the electronics from the current flowing thru the sensor, rather than from the output side of the sensor. The output side can be viewed simply as a contact.

The output has very low leakage and thus probably requires very little load to obtain good switching at the PLC input. Maybe no shunt resistance.

You really need to know if the current sensor works correctly with a sine wave current thru the sensor.

The input current to the VFD is probably quite peaked from being the input to a capacitor input filter. The current flowing thru the sensor is what powers the sensor. Thus, getting the equivalent current seen by the sensors to be high should be advantageous. That you are doing by looping more turns thru the sensor.

You want a small hysteresis between on and off of the sensor. Whether you use a NO or NC output should not change how the input and threshold part of the sensor works.

The datasheet is at http://www.automationdirect.com/static/specs/acuampacs200.pdf this is a 5.8 MB .pdf file and takes an excessive amount of time to download. These essential specs could probably fit in 0.1 MB.

Hysteresis on the high range is listed as 3%. You should be able to set your threshold at about 70 A and get satisfactory detection of lost load. Input to the VFD should be the best place to make the measurement. What meter are you using to get the normal load and idle load current measurements?

.

 

[ptonsparky]

The motor this is actually going to work with is a 100hp across the line start. It has an amp meter with 200:5 CT. That is what we were using to determine the current there.

The vfd is on a 1.5 hp fan in my shop. Single to 3 phase conversion. I was using a true RMS amp clamp for my readings in the shop.

Thanks for the reference, looks like some good bedtime reading.

 

[ptonsparky]

Finally got back to this.

Installed the unit I tested with the vfd in my shop. Works great.

Only difference from original attempt is the N/C unit vs a N/O, which should mean nothing, and instead of installing it next to an existing CT, we moved it a few inches away on a different phase. That was mentioned in the trouble shooting hints I read somewhere along the way. (Distance not phase change).