Brushless AC-Motor Controller

Oh, dear. That looks a bit horrible.

If you take a close look on smoke IGBT picture you will find a heat larger heat sink exactly equal to the size of the whole PCB and unfortunately i didnot take the complete picture to show you completely...

Secondly... i did not agreed with your guess, i try to explain in more practical way...

I have use this type of controller in our machines which produced socks, and we have divide all the machines catogrically like LINE A,B,C,D,E,F and each lines consist of 25 Machines that is equal to 25 controllers (PCB) each IRT-Drive is responsible to control one machine and all the requirements like "Improper sizing of driver to motor or load" temperature issues, defective driver will cover from the manufacturer and done properly i.e. upto the requirement of the machine.

What we notice is that LINE A and LINE B has a good reputation in respect of accidents means the ratio of these type of occurence is much much lesser as compare to LINE C,D,E,F keep in view that all the parameters like machine capacity, model, power, voltage rating and load are similar...

what i show to you is the INPUT side in terms of voltage readings (describe in first post) because we have find only difference in each lines which leads to this sort of casualty is phase to phase minor voltage varaition but how i can practically prove that this is because of this voltage varaiation if that is the case.. OR you guys help me to work out how i solve this problem...

Iam available to provide any information you want to know about this sort of problem...

Help out...

Dear Sir, a very productive and observable reply received at your end, with reference to your feedback i think two things are much similar as what i observed.

1) Over Voltage from line spikes

2) Drain over current

lets talk about point-1 that how we observe or analyse or detect Over Voltage from line spikes...

and how we measure drain over current...

Thanks in advance...

waiting for your favorable reply...

Put good low pass emi filtering on supply and find mov or similar device to go across the supply phases. Find out what the nominal voltage of phases to ground. Use properly sized mov or vdr to snub the spikes.

Are the lines sourced to a central supply separately or taken from one line to another.

Does the "a" line supply and other subsequent lines daisy chain each line.(not a good technique.)

Separate each line and use emi and mov's on each line/machine.

I can cite many instances of dirty power causing these kind of issues.

Remove any added on accessory devices and run the on separate power source.

Are other devices hooked into the power for these machines or feed back to the same source?

Now let's get to specifics. From the photo you provided I think I can see a circuit deficiency. The W output is directly connected to output and a "turn" does not pass through a core like Y and V, which I must assume is for current feedback to the controller. On normal conditions this scheme is OK since the third branch carries the algebric sum of the other two currents, but this is not the case for shooot-through current (in W branch) or if for some reason W suffers a short with power rails. By pure coincidence, W driving IGBT pair has given the biggest bang, and the high side IGBT the worst. This is also consistent with a W short to (-) but I would investigate the shoot through scenario first. Can you give the gate driver types and gate resistor values? (both low and high sides) S.M.

What are the motor and drive ratings? How many amps was the motor drawing?

Dear Sir,

Motor ratings are listed Below:

RCV Turin Italy

Lyra 28

Nm30-Kw1.25-w42-V202-A4.5

and Driver Rating (R,S,T)=220V

Your voltage readings in the table were all around 385, yet the motor and controller are rated at 202 and 220 volts. Was the motor accidentally wired in delta versus star?

Sir the voltage readings i have shown you it basically came out from five stages..

1) Output from Generation Unit (400V)

2) Becomes input to the Distribution Side (400V)

3) Then Distribution Output will become the input of the Voltage Stabilizers of each lines (A,B,C,D,E and F)

4) Finally the output of the Voltage Stabilizer will become the Input (380V which fetch as input of the machine Transformer) of each machines which are connect to this stabilizer.

5) and Hence the output of Transformer i.e. 230V will become the Input of Drive (R,S,T)

Ok, the input voltage is for sure not the issue.

To give you good explanations from remote is close to impossible -From the photo it is hard to judge what happened. But I think I can notice "shoot out craters" close to the right pins of the upper two IGBTs. This the area where the bond wires between chip and pin where molten and exploded due to the exessive heat. This is a sever damage and happens when a lot of current can flow

VFD need in most cases special fuses super fast blowing in the supply line - to prevent such damage - are they in place and have you checked withthe supplier of the VFD?

Have you checked that the rating of the motor and and VFD fit each other and the VFD control unit is not overloaded . Especially consider instant high load changes on the motor

Consider possible shorts in the supply cables to the Motor due to movement of the cables or irregular current follow due to damage of the cables.

other reason i could imagine that the isolation between the IGBT and the heatsink is not vibration resistant and after some time will cause a short. It would be interesting to see the burn marks on the Heatsink side.

Discuss the burns withthe vendor of the VFD or the maschine it was build into.

Best regards

Sir here is the image of heat sink for better understanding....

Thanks for the picture. I see that there is heatsink grease and persumably also heat conducting rubber plate below the transistors.I understand that the cabinet is alo used as a heat sink, So clearly the shoot through goes also through the rubber foil - that is might be an indication that the material does not withstand the voltage levels which occure inside the circuit.

And annother issue makes me wonder. Clearly to the left of the left transistor group there are two spots where a direct break through from the PCB! to the heatsink sheet metal is visible. I suspect that the PCB has not enough stand off to the heatsink and you get arking there. The area of the left transistor group looks pretty sotfree.

Naturally also an exploding capacitor can be the reason. But I would check for the standoff issue at first.

Difficult to give a hint but I would try to really find out what material the original manufacturer is using - the material is widly used but there different materials which look the same but have different properties. In a PC where most of the stuff is used you find only low voltages 3.3V - but here you need material with 1KV standing power at least. There are power electronic grade materials out there.

Ok what is happening is a severe disaster but the severe burns I would not expect. Can it be that ol and fibers cause the severe burning?

Could it be that by accident someting can press the heatsink sheet metal in direction to the PCB and create some shorts on the board?

What is the purpose of the two threaded parts blow the the transistors?

Sorry not to be able to give you a good hint regarding the issues at hand.

Best regards

Thanks for your keen consideration... i am responding in bold to your comments.

Thanks for the picture. I see that there is heatsink grease and persumably also heat conducting rubber plate below the transistors.I understand that the cabinet is alo used as a heat sink, So clearly the shoot through goes also through the rubber foil - that is might be an indication that the material does not withstand the voltage levels which occure inside the circuit.

Since this sort of incidents occuring instantly so i observed that sometimes on some DRIVES heat found little greasy and sometimes on some DRIVES heat sink found lush shinny... so i think black smokes are very much common in all the cases of incidents.

And annother issue makes me wonder. Clearly to the left of the left transistor group there are two spots where a direct break through from the PCB! to the heatsink sheet metal is visible. I suspect that the PCB has not enough stand off to the heatsink and you get arking there. The area of the left transistor group looks pretty sotfree.

Naturally also an exploding capacitor can be the reason. But I would check for the standoff issue at first.

With this sort of incidents only GATE DRIVER CIRCUIT are mostly damage and during repairing i only focus on the gate driver section and none of any case i have found any CAPACITOR has been found exploded.

Ok what is happening is a severe disaster but the severe burns I would not expect. Can it be that ol and fibers cause the severe burning?

It is true that these drives are install in the area where fly waste (yarn) are present since the panels are close properly but still after several weeks i have perform preventive overhauling so it might be possible fly waste are helping to burn more severely.

Could it be that by accident someting can press the heatsink sheet metal in direction to the PCB and create some shorts on the board?

I did not get this point will you please elaborate more....

What is the purpose of the two threaded parts blow the the transistors?

One thread region is responsible to hold the PCB by tightening the small screw and other is responsible to screw the base of TEMPERATURE SENSOR (RT4) of DRIVE.

Sorry not to be able to give you a good hint regarding the issues at hand.

Atleast with discussion i am sure i can figure out this problem...

I came back here now that I noticed that I was granted a GA with my earlier answers. I now see the heat sink contact area and the circuit board provides access for the six mounting screws that secure the six power transistors to the heat sink. The lack of smoke marks directly between transistor and heat sink implies that they were in contact with the heat sink but I see something now that really bothers me. Two of your transistors have marks that align with small black circles in your heat sink. It looks like somebody drilled through your heat sink into these two transistors. The holes in the transistors are similar if not identical in size and circular. I would expect an arc through to produce a much more irregular shape. Drilling into a power transistor will destroy many things beyond just the transistor. You have a bigger problem than just this driver.

Good Luck

Suggest you analyze system controls for:

1. IGBT gate control PLL timing errors. If the IGBT gates are triggered at the wrong point in time the result is a "phase-to-phase" short through the associated IGBT units and the result will be as ilustrated in your pictures. If unstable check for induced voltage spikes, noise, etc from other components/equipment.

2. If the PLL signal and control is stable and of correct timing; Check your filter network as any filter component failure can allow regeneration voltage spikes to occur that far exceed the rating of the circuit board components/devices. (open or shorted capacitor(s), diode(s), resistor(s).

3. Check for atmospheric contaminents that are causing short-circuit faults. (moisture, chemicals, lint, etc.)

4. Install a voltage recorder on the primary power supply side and monitor for high transient voltage spikes then identify the source of any issues found.

5. Verify the IGBT's are original equipment design compliant as the gate transition timing, reverse voltage impedance , and all other design factors are critical in facilitating proper/correct firing sequence timing events. (OEM specification sheets)

6. Look for abnormal operating events such as multiple pieces of equipment being shutdown/stopped or started at the same time causing supply voltage disturbance.

7. Check your primary power for induced harmonics 5th, 7th, 11th, 13th that can be (are) extremely destructful. The filter usually will have default/standard correction for this however in some cases, rogue harmonics can be present. A high quality oscilliscope can reveal a lot of information in a very short time if used properly.

Good luck, Jim

Dear Sir,

Your feedback is very much effective please let me know few things ...

1. How i analyse PLL timing Errors...

2. In my DRIVE there is a section of internal power supply which consist of FLYBACK transformer, LM2575 ADJ,VIPER 50SP and Banks of Capacitor SMD type and Inductors... but mostly with these sort of cases transistors BSR14 and BSR16 are also damage which leads towards explosion.. also let me know what is the best way to test the SMD mounted Capacitor(s) open or shorted ???

3. Temperature environment is hotter mostly in my production unit and yes fluffy environment is also there but there, but there are other machines lines (Like Line A and Line B which i describe in my first post) which also covered in these condition but on these lines this sort of incidents are almost tends to zero..

4. Please suggest any best Voltage Recorder for my case...

5. OEM specification sheet of GATE DRIVER CIRCUIT already displayed you can also take a review of it...

Waiting for your favorable reply

Not to insult your capability or intelligence but I have no way of knowing your level of knowledge of troubleshooting experience/expertise.

Therefore in the interest of safety I must caution you that the risk and exposure you or whomever will be operating the test equipment and performing the analysis is very significant especially when testing on an operating and/or energized control circuit of this type.

Please use utmost care and if you are not absolutely sure of what you are doing, you must not perform any procedure or measurement on an energized system.

This conversation and the procedures are way too complex and lengthy to post in this forum.

CR4 Admin - email address removed

From the CR4 Rules: Do not post phone numbers or email addresses. The CR4 Admin will delete all phone numbers and email addresses posted in threads or comments. You can share this information via the CR4 internal messaging system.

If this is not acceptable, my suggestion is to contact the OEM and have their service department go to you location for support and repair.

Jim