Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Open Circuit Protection - 3 phase AC

Status
Not open for further replies.
kappa_am

kappa_am

Full Member level 6
Joined
Jul 16, 2012
Messages
331
Helped
19
Reputation
38
Reaction score
19
Trophy points
1,298
Location
Vancouver
Activity points
3,859
Hi fellas,
I am looking for open circuit protection in 3-phase AC system. In other word, I would like to have an interrupt when one of the phases is disconnected at output of my inverter. Is there any simple way to implement this? except Park transformation.

Thank you for your help.
 

Do you have a neutral point that is not grounded? Would a disconnected phase cause neutral point to move away from 0V?
 

Most common 3-phase circuit breakers have three single phase circuit breakers that are coupled together with a plastic rod that forces all the three to trip together!!

You need to pass the three phases through three toroids that have the same (identical) number of turns wound on it. Each phase acts like a 1 turn winding and now you have three coils that represent the currents in the three conductors.
 

Motor inverters are often equipped with output phase loss diagnostic. I would expect that they only check the current magnitude, no need for park transformation. Output current measurement is required, however.

Phase loss detection must be disabled if the inverter is operated with low or asymmetric load.
 

reply to Mitra:
thank you for your reply.
What's the point?
suppose I do so, How a disconnected phase will be detected?
when a phase get disconnected the current of the other phases will be equal.

- - - Updated - - -

Hi FvM,
it's not that easy. output currents are AC; how the diagnosis system will discriminate between zero crossing and phase loss?
Although I have some ideas, I would be pleased if get others idea to sum it up to a perfect solution.

Thanks
 

it's not that easy. output currents are AC; how the diagnosis system will discriminate between zero crossing and phase loss?
Click to expand...
I don't see what AC current magnitude has to do with zero crossing. By nature, an AC current measurement is a kind of averaged quantity.
 

kappa_am said:
suppose I do so, How a disconnected phase will be detected?
when a phase get disconnected the current of the other phases will be equal.
Click to expand...

Sorry if I was not clear enough. Consider first one phase: this wire goes through a toroid that has (say) 100 turns of fine wire wound already. The fat central wire acts like a one turn coil. When some current goes through it, there will be some voltage induced in the secondary, that can be rectified and conditioned and used as signal for 'phase active'.

If all the three phases are active, there is nothing to be done. The load is reasonably balanced.

If all the three phases are inactive, there is nothing to be done. there is no load.

If one of the phase is active (or inactive) and the other two are inactive (or active) then there is some problem. Your attention is needed.
 

In analog measurement, like galvanometer, AC quantities measured as RMS value. But how about measuring with MCU?

- - - Updated - - -

Reply to mitra:
in 3-wire system, when a phase get lost, currents of two phases will be equal ( simple KCL). Therefore your recommended circuit cannot detect that. Because the result of the interaction these two phase is zero, As it is in symmetrical 3 phase system, where interaction of 3 phases are zero.

Thanks :)
 

Yes, current in only one phase cannot happen, but current unbalance respectively disconnection of single phase can be well detected.
 

kappa_am said:
in 3-wire system, when a phase get lost, currents of two phases will be equal ( simple KCL). Therefore your recommended circuit cannot detect that.
Click to expand...

Sorry if I was not clear enough! What I suggested and wrote that you need to process each phase individually and not together. This is the same principle that is used in the clamp on current meter.

You process each phase independently and get three outputs from three secondary coils.

All the three phases are detected together in case of ground fault interrupter. If for some reason, the sum of the currents in the three phases is not zero, that indicates that somewhere a leak is taking place. But we are not talking about that.
 

If you are attempting to protect a three phase induction motor, all that you require is to sense the current of two of the three phases.
If a preset limit is exceeded, you open an auxiliary contact that feeds the contactor's coil.

Think a little bit what happens to a 3-p motor when a phase it lost, and you'll realize how come it works.

This scheme has been in use for almost 100 years.
 
  • Like
Reactions: abc_de

    A

    abc_de

    Points: 2
    Helpful Answer Positive Rating
kappa_am didn't yet talk about the purpose of the intended phase loss diagnostic, in so far it's difficult to decide which solution is appropriate.

I agree that motor circuit breakers or overcurrent relays are the industry standard means to protect 3 phase motors. But it may be reasonable in some cases to implement protection that acts before an actual overcurrent can be determined. As said, motor inverters are often able to sense current imbalance and shut down operation in case of load phase loss. The protection is more or less for free because the inverter needs to sense the motor current already for other purposes.
 

But in 3-wire system a phase is lost, still the sum of the currents is zero. A phase that is disconnected =0, and two phases with equal current.

c_mitra said:
Sorry if I was not clear enough! What I suggested and wrote that you need to process each phase individually and not together. This is the same principle that is used in the clamp on current meter.

You process each phase independently and get three outputs from three secondary coils.

All the three phases are detected together in case of ground fault interrupter. If for some reason, the sum of the currents in the three phases is not zero, that indicates that somewhere a leak is taking place. But we are not talking about that.
Click to expand...

- - - Updated - - -

I need to prevent unbalance operation of motor because of harmonic generation and ripple torque.

Reply to Schmitt:
There may be a situation in which a phase is lost but the current isn't still higher than nominal current. Furthermore, the motor will work overloaded some times (for limited time).
 
  • Like
Reactions: abc_de

    A

    abc_de

    Points: 2
    Helpful Answer Positive Rating
But in 3-wire system a phase is lost, still the sum of the currents is zero. A phase that is disconnected =0, and two phases with equal current.
Click to expand...
Why "but", everyone agreed to this point as far as I see.

Can we conclude that the only remaining problem is that you don't know how to implement current imbalance detection?
In analog measurement, like galvanometer, AC quantities measured as RMS value. But how about measuring with MCU?
Click to expand...

I feel that there are about one thousand Edaboard threads discussing AC current measurement by a microcontroller, starting with the classical analog circuits (rectifier+averaging filter+ADC or rectifier+rms-circuit+ADC).
 
  • Like
Reactions: abc_de

    A

    abc_de

    Points: 2
    Helpful Answer Positive Rating
Can we conclude that the only remaining problem is that you don't know how to implement current imbalance detection?
Click to expand...
still beats me
How we can calculate system imbalance when the summation of the values is still zero?

I feel that there are about one thousand Edaboard threads discussing AC current measurement by a microcontroller, starting with the classical analog circuits (rectifier+averaging filter+ADC or rectifier+rms-circuit+ADC).
Click to expand...
Generally speaking, don't you think these method are rather slow? however in this application we don't need very fast diagnosis. It seems also expensive a transformer, bridge, ... .
However, I prefer to use Isoopamps. They measure AC values with a bias, dragging all the waveform to positive values.

It is a solution, It may help to who want to use the method explained by FvM, Don't forget to add a resistor to output of the rectifier. it may solve speed problem slightly.
 

kappa_am said:
But in 3-wire system a phase is lost, still the sum of the currents is zero.
Click to expand...

That is precisely the reason you must not sum the currents. You are simply detecting the condition when one of the phase is not having current but the two other phases are having current.

You have the following truth table:

Phase 1 Phase 2 Phase 3 Condition

Active Active Active Good
Inactive Inactive Inactive Good
-----------------------------------------------------------------------
Any other combination Fault
-----------------------------------------------------------------------
 

c_mitra said:
That is precisely the reason you must not sum the currents. You are simply detecting the condition when one of the phase is not having current but the two other phases are having current.

You have the following truth table:

Phase 1 Phase 2 Phase 3 Condition

Active Active Active Good
Inactive Inactive Inactive Good
-----------------------------------------------------------------------
Any other combination Fault
-----------------------------------------------------------------------
Click to expand...
Thank you,
I've got it. It is a similar method to what FvM discussed. we need RMS value of AC values. However, I will searching for more rapid and cheaper solution.
 

kappa_am said:
we need RMS value of AC values.
Click to expand...

We do not need the current values; we just want to know whether a given phase is carrying current or not. We set a threshold and every current value above the threshold is considered 1 (active).

We next do an XOR of the three signals coming from the three phases.
 

Suppose that the time one phase is crossing zero, the combination will be active active deactive--->fault. while the system is ok and just current of one of the phases is at zero crossing situation.

c_mitra said:
We do not need the current values; we just want to know whether a given phase is carrying current or not. We set a threshold and every current value above the threshold is considered 1 (active).

We next do an XOR of the three signals coming from the three phases.
Click to expand...
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top