Phase sequence indicator

[sachin88]

Hello,
I am trying to build a phase sequence indicator. If the phase sequence is R,Y,B then the LED should light up indicating Sequence is correct. If sequence is other than R,Y,B then LED should not glow. I have attached my circuit below. I have done the simulation of this circuit. For any sequence LED glows. Could anyone please help me to troubleshoot the circuit.

Thank you

 

[crutschow]

Post your simulation results.
If you are using LTspice, post your .asc file also.

 

[sachin88]

I am using Proteus software to simulate the circuit. I am unable to upload the simulation file since its giving error "The uploaded file does not have an allowed extension" while uploading. If you could provide me your email ID i can mail you the simulation file. My email ID is ******

 

[rmn_tech]

not a good idea to put your email in open forum posts. To post the file try compressing it to a zip file and post that.

 

[sachin88]

Attached the Zip file with this post. Thanks for the suggestion about the email id. I'll take care about that here onwards.

 

[spec]

Hy sachin88,

Can you give details of the three phase signal: wave shape, amplitude, frequency.

spec

PS: where in India are you?

 

[sachin88]

Hello spec,
Three phase detail: Phase to phase 440V. Phase to neutral 230V. Frequency 50Hz. sine wave. ( its actually 3 phase alternator output).

I am from Karnataka (South India).

 

[spec]

Hello spec,
Three phase detail: Phase to phase 440V. Phase to neutral 230V. Frequency 50Hz. sine wave. ( its actually 3 phase alternator output).

I am from Karnataka (South India).

Thanks sachin88.

I will give your question some thought.

spec

PS: you can put your location on your user window on the left of your posts by adjusting 'Location' on your user page. Not only is it interesting to know where you are from but it also helps us to answer you question: component availability, mains voltage etc.

 

[crutschow]

What supply voltage do you have to power the circuit?

 

[MaxHeadRoom78]

Looks like this one, do you have it referenced to the neutral?
http://www.tradeofic.com/Circuit/5720-PHASE_SEQUENCE_DETECTOR.html
Yours is using the earth ground symbol?
Max.

 

[crutschow]

I couldn't get that circuit to work properly either, so I came up with my own take, using a D-FF as shown in the LTspice simulation below.
I added Schmidt triggers at the FF inputs to square-up the signal and avoid any false triggering.
You just need one FF to monitor two phases since there are only two possible phase sequences and those can be detected by just looking at two of the three phases.
I added a 5V line-powered supply to power the circuits. It uses the series capacitor C1 to limit the current to the regulator zener diode D8 without the high power dissipation that a resistor would have.

In the first simulation the phase sequence is a correct ABC. This generates a FF-D input high when the clock (CLK) goes high, so the FF Q output also goes high, turning on the LED (ID5) from the /Q output.

In the bottom simulation graph the phase sequence is reversed (CBA), giving a FF-D low input when the clock goes high. Thus the FF Q output is low (/Q output high) and the LED is off.

Note that the 5V power and ground must be connected to the ICs, which is not shown.
Also be sure and tie all unused inputs on both ICs to common to avoid anomalous behavior.

 

[crutschow]

The lack of a phase sequence detector cost my company dearly many years ago. I was working on one of the early FLIR (infrared) systems for the Army and it used a 3-phase, 400Hz powered helium cryogenic cooler for the focal-plane detectors. We had in working fine in one lab using 3-phase power from the wall-plug outlet supplied by the 400Hz rotary inverter in the building.
We then moved it to a larger lab in a different building and plugged it into the inverter outlet in that building. We turned in on and it seemed to be operating normally. Someone noticed that it sounded a little different but didn't think too much of it. They kept waiting for it to cool down, which usually took about 15-20 minutes at which point the IR picture starts to appear, but the display stayed dark. They started checking the unit and then realized that the cooling fans (also 3-phase) were rotating in the wrong direction, indicating incorrect phase. They immediately shut off the power, but it was too late. It no longer worked, even after they rewired the connector to correct the phase. They took the focal plane apart and found that the cooler rotating in the wrong direction acted to heat the detectors instead of cooling them, and they had gotten hot enough to ruin their detection ability.
Don't know what that cost to fix but it weren't cheap, as all 350 detectors were individually mounted and hand wired (it had a mechanical horizontal scan to generate the picture from a vertical array of detectors).
Of course after that, they closed the barn door, and installed a 3-phase sequence detector to remove the power to the cooler if the phase wasn't correct.

 

[MaxHeadRoom78]

They are a must for elevators and remote pumping station installations.
Max.

 

[spec]

The lack of a phase sequence detector cost my company dearly many years ago. I was working on one of the early FLIR (infrared) systems for the Army and it used a 3-phase, 400Hz powered helium cryogenic cooler for the focal-plane detectors. We had in working fine in one lab using 3-phase power from the wall-plug outlet supplied by the 400Hz rotary inverter in the building.
We then moved it to a larger lab in a different building and plugged it into the inverter outlet in that building. We turned in on and it seemed to be operating normally. Someone noticed that it sounded a little different but didn't think too much of it. They kept waiting for it to cool down, which usually took about 15-20 minutes at which point the IR picture starts to appear, but the display stayed dark. They started checking the unit and then realized that the cooling fans (also 3-phase) were rotating in the wrong direction, indicating incorrect phase. They immediately shut off the power, but it was too late. It no longer worked, even after they rewired the connector to correct the phase. They took the focal plane apart and found that the cooler rotating in the wrong direction acted to heat the detectors instead of cooling them, and they had gotten hot enough to ruin their detection ability.
Don't know what that cost to fix but it weren't cheap, as all 350 detectors were individually mounted and hand wired (it had a mechanical horizontal scan to generate the picture from a vertical array of detectors).
Of course after that, they closed the barn door, and installed a 3-phase sequence detector to remove the power to the cooler if the phase wasn't correct.

You would have thought they would have a temperature sensor on the IR array as it is so complex, fragile, and expensive.

spec

 

[spec]

Often it is handy to establish what the phase order is rather than just know that it is one particular order. You can indicate this with three LEDS which continuously illuminate on and off in sequence.

spec

 

[crutschow]

You would have thought they would have a temperature sensor on the IR array as it is so complex, fragile, and expensive.

There was a sensor but it was just to control the temperature once it cooled to the desired point.
There was no particular need to know the temperature otherwise.

Often it is handy to establish what the phase order is rather than just know that it is one particular order. You can indicate this with three LEDS which continuously illuminate on and off in sequence.

There are only two possible phase sequences for a 3-phase circuit so a single light is sufficient.

 

[sachin88]

Thank you for the simulation and the circuit Crutschow. I appreciate your help. I'll build the hardware and update the progress.

 

[spec]

There was a sensor but it was just to control the temperature once it cooled to the desired point.
There was no particular need to know the temperature otherwise.

I was thinking of a temperature sensor to warn in the case of a sensor overheat. Ideally in that situation it would power the system down to save the sensors.

There are only two possible phase sequences for a 3-phase circuit so a single light is sufficient.

ARRRAGH

spec

 

[schmitt trigger]

I simulated the circuit shown by the OP. It works perfectly,
and it will even detect a phase loss.

With the sequence ABCABC..... where A is the top phase and C the bottom one, the output of the last NAND is a negative going, 60 electrical-degrees-wide pulse.
With the sequence ACBACB...........the NAND output is always high.
Likewise when you lose any phase, the NAND output is always high.

With the PNP transistor and LED connected as shown, a good sequence will dimly light the LED, and be dark with all other conditions.
If one desires a brighter LED, a retriggerable monostable could be used.

 

[crutschow]

I simulated the circuit shown by the OP. It works perfectly,
and it will even detect a phase loss.
.........................
With the PNP transistor and LED connected as shown, a good sequence will dimly light the LED, and be dark with all other conditions.
If one desires a brighter LED, a retriggerable monostable could be used.

That was my observation also. It didn't keep the LED lit steadily so I considered that not acceptable.
But you brought a good point about stretching the pulse with a monostable.
Below is the circuit using an RC pulse stretcher with an added diode and capacitor at the NAND gate output, if you want a circuit that will also monitor for a phase loss.

The first plot is for the correct ABC phase sequence which gives a 10mA steady current through the LED (with a slight ripple due to the power supply ripple which will not be visible).
The second plot is for the incorrect CBA phase sequence which give zero current through the LED.