0
\$\begingroup\$

Background:

There is a 3 phase motor running by an inverter. motor specification is about 1.8KW power and 230/400V and 1.13A. I want to drive another DC motor synchronized with that AC motor so the DC motor goes ON and OFF with AC motor. there is no Neutral connection to motor so I can't use a relay or contactor, also inverter is connected to PLC through Modbus connection and I don't want to modify inverter or PLC.

What I have done and my questions:

I thought about using a current sensor (or hall effect sensor) to monitor AC current going to motor, and I found too many different kinds of it. I also tried 3 phase monitoring relay, when I connect it to main 3 phase supply it works correctly and a relay activated inside of it but when I connect it to motor/inverter it doesn't work in 6 different connection type I tried. why?

  1. Is it right to use current sensor for this application? for example allegro current sensor IC ACS758 is it the right choice for this application? as the inverter make different frequencies, doesn't it cause to make buzzing at sensor or op-amp outputs? should I connect 3 different sensor for each phases and SUM their output into an Op-amp?
  2. or is it better to use a non-invasive clamp current sensor (this) around all 3 phases wire? and/or one of the phases?
  3. why the phase control relay not works when I connect it to motor/inverter ?

my problem solved:

I use inverter discrete I/Os terminals to detect motor ON/OFF state.

*inverter model is: Schneider ATV12H018M2 *phase control relay is: Schneider RM17TG20

\$\endgroup\$
  • \$\begingroup\$ How about a relay connected between two phases - no need for a Neutral then. \$\endgroup\$ – Peter Bennett Oct 28 '17 at 15:42
4
\$\begingroup\$

For most purposes, it is probably best to use a signal from the variable frequency drive (VFD or "inverter"). Most VFDs have analog signals, discrete signals or relay contact indications for external use. An output from the PLC would also be a good alternative.

You can probably get a variety of indications including: motor running, motor running above X speed, motor torque and others. If the VFD has a Modbus connection, it is probably capable of providing the best indication of most motor operational conditions while the motor is under the control of the VFD. Many of those indications should be available separately from the PLC if you want them that way. A VFD with Modbus likely has some customer configurable discrete and analog I/O points built in for external use. Study the VFD manual to se what is available and don't hesitate to use features intended for customer use. You should not need to buy anything else unless you need to know what is happening while the motor is not under the control of the VFD.

\$\endgroup\$
  • \$\begingroup\$ I checked vfd functions before there is a relay inside of it which it get energized at standby mode, it doesn't really help me because I want to detect motor running state, I don't know how to do it with vfd? \$\endgroup\$ – rocknroll Oct 30 '17 at 17:23
  • 1
    \$\begingroup\$ @rocknroll Get the manual of the VFD and change the function of the relay if it is not already in use. \$\endgroup\$ – Jeroen3 Oct 31 '17 at 11:45
  • \$\begingroup\$ @Jeroen3 I read the manual of VFD and found that there are many discrete and analogue I/Os terminal, and simply one of the logic outputs indicate motor running state with +24V signal. thanks for you information, it is my first approach to VFD. I think the VFDs are a very flexible and intelligent devices to control motors. \$\endgroup\$ – rocknroll Nov 7 '17 at 14:13
1
\$\begingroup\$

Is it right to use current sensor for this application? for example allegro current sensor IC ACS758 is it the right choice for this application? as the inverter make different frequencies, doesn't it cause to make buzzing at sensor or op-amp outputs? should I connect 3 different sensor for each phases and SUM their output into an Op-amp?

It is a way since for a motoring electrical machine would have current in the stator (not always, see [3]. That particular part is marked not for future use but there are newer, equivalent parts. These current sensors have an internal filter around 120kHz so it will track current ripple due to PWM.

NO NOT sum their outputs with an OPAMP. For a balanced 3phase load (and an induction machine, fault free will appear as such) the current summation will be zero. In practice there will be some "current" due to sensor errors (not 4% error in measurement) and capacitive paths.

or is it better to use a non-invasive clamp current sensor (this) around all 3 phases wire? and/or one of the phases?

This isn't really a 2nd question, just another way to sense current and yes this method could work it really comes down to what you are to interface with. The hall effect sensor method will require a PCB of sorts and the main stator current to pass through this PCB. This clamping CT doesn't have this restriction. NOTE: CT's do not sense DC current but for this application that might not be a problem (I know induction machines do not have DC current... but stall cases and equally fault cases might be an alternative).

LEM's are a non-contact current sensor you might want to investigate as well

or is there any better way of doing this?

You state you are interested in when the motor is rotating correct? does this also include free-wheeling/coasting? IE you only want to act when the rotor has stopped? you could you a rotary encoder, even a simple slotted disc on the shaft and a slotted opto would give you a pulse once per revolution. A monostable could then be used to indicate when it has been stopped to trigger your higher logic.

Basically there are many ways to solve this and the optimum one will depend on how quickly you need something, how fast you need something, how "future proof" you want something

enter image description here

enter image description here

I have in the past ( as recently as 3months ago), used such a scheme to provide an absolute position marker while working with motor-drive. The disk I use has one slot as I am interested in once per revolution for a fixed mechanical datum, but the concept is exactly the same. if you see pulses, you are rotating. If you were to connect such an opto to an edge-triggered mono stable (to deal with the cases it parks on a slot) you would have a nice indication of rotation which also covers the case when the inverter is disabled and the electrical machine is coasting to a stop (no stator current)

\$\endgroup\$
0
\$\begingroup\$

The sum of the three currents is obviously always zero, otherwise Kirchhoff's law would be void.

schematic

simulate this circuit – Schematic created using CircuitLab

This is generally the same if you used three current transformers instead and summed up the resulting values. You may have additional artifacts from low-pass filtering then, so don't do that.


You have to weight the currents as they are measured so the sum of two can never outweight the third. This can be easily done by having 1, 2, 4 windings through the common current transformer (or multiples of that for better sensitivity.)

schematic

simulate this circuit

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.