Announcements

Register now for the most anticipated Asia Pacific Power Seminar 2022

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

Gate Driver ICs Forum Discussions

mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Dear community,

we are using 6ed2230s12t gate driver IC for a three-phase motor drive, design is working smoothly, and now we have added over current protection of 22 amperes through shunt resistors. the value of shunt resistors is two 100 milliohms and one 470 milliohms in parallel, the total shunt resistance became 45 milliohms with a 10W power rating of each resistor (total 30W). for 22 amperes the voltage across shunt resistors is 1V which is then divided through a voltage divider which is  5.1K for each R1 and R2, so the total voltage across ITRIP becomes 0.50V. This is over-designed which is designed using the Infineon datasheet for the mentioned gate driver, pictures are also attached.

6ed.PNG

now the problem we are facing is the gate driver reset with a very small load even at the start or sometimes try to start but resetting, again and again, our team has tested this design for the previous batch of gate driver ICs, and it was working smoothly but with the new batch of ICs, we are facing the problem. there is a noise on the ITRIP pin but the same noise was on the previous gate driver which was working fine. The gate driver heats up after a day or even after an hour some time which damaged the IGBT module.

 

WhatsApp Image 2022-06-01 at 3.21.39 PM.jpeg             WhatsApp Image 2022-06-01 at 3.21.36 PM.jpeg

0 Likes
1 Solution
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Below are few of the additional observation which will help you to mitigate the problem.

1) From the schematic, C38/C39 should be between VCC and VSS. But two parts are farther from the pins of the gate driver IC.

2) A ceramic( 100 nF~ 1uF) capacitor between VCC and COM  and VCC and VSS (>4.7uF) is recommended. And the capacitors should be close to GD IC 6ED22.....

3) For sensing,  the resistor which you have used is radial type which has more ESL .Please try to use a SMD thick film resistor with low ESL to minimize the driver loop. 

4) We don't have any limit for minimum value of R0.It is up to level of current protection you need as per the application.

5) To further minimize the spike, cut the trace between COM of 6ED....GD IC and COM of module and add 2.2 ohm resistor between them.

AZIZ_HASSAN_0-1655292388301.png

I hope this will work and please let me know whether the issue is solved or not.

BR,

AZIZ HASSAN

View solution in original post

0 Likes
18 Replies
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Thank you for posting at Infineon community.

We really apologize that the gate driver IC is not working fine. In order to solve the problem for you, I request you to revert to following points.

1) The previous batch and current batch of driver  you are referring to has same part number or not and when did you buy your previous batch and current batch.

2) Can you please share the Vcc voltage waveform at starting when you see the fault at lower currents.

3) The resistors values (R0,R1,R2) which you are using for detection and fault clearing time (i.e. Rref and Cref) are same for previous and current IC lot.

4) Can you please share your driver schematic (including driver to controller connections)and PCB routing details to look for noise issues.

BR,

AZIZ HASSAN

0 Likes
mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Hi Aziz,  

Thank you for your support.  There are following answers to you point: 

1. Batch Number . 

Previous Batch numbers are H1908  and H2030. H1908 was bought on JUN 2021 from Chip1stop and H2030 was bought from Mouser on may 2021. 

The existing batch we recently bought  on April 2022 from Electronics Source Thailand and its batch number is H1919 which is not working. 

2.  VCC  voltage of the gate driver is attached. 

3. Yes. The resistors values (R0,R1,R2)  are same for all batches. 

4. Schematic and PCB layout are attached. 

we checked the previous batch, was working fine with load and reset according to our calculated ampere but with new batch, the gate driver even heat up without load. In some cases it is heat up after one or two trials but sometimes it work for one or two days.

 

 

 

 

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Thank you  for the details and I request you to answer  the following points.

1) May I know weather you have used same PCB for previous and current lot or you made some changes in PCB and then tested the current lot. If the PCB is different in both cases, then please test the current PCB with previous IC lot.

2) How many IC you have ordered and does all of them have same problem?

3) You mentioned that gate driver is heating without load. May I know which load you are referring to. Is it gate driver load( IGBT input capacitor) or the motor load.

4) Can you please share the serial number of couple of IC from previous and current batch each.

BR,

AZIZ HASSAN

 

0 Likes
mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Hi Aziz,

These are the answers to your requested questions.

1. PCB is same for previous and current batch.

2. We have ordered 1000 pieces for gate driver ICs, we didn't check all pieces but all the checked ICs have same problem, some ICs heat up and damage during QA/Testing, while some damage after two days of VFD installation at max.

3. Without load means, the load across IGBT module or VFD output, which is driven by gate driver. 

4. The batch number of gate driver ICs are:

Previous Batch numbers are H1908  and H2030. H1908 was bought on JUN 2021 from Chip1stop and H2030 was bought from Mouser on may 2021. 

The existing batch we recently bought  on April 2022 from Electronics Source Thailand and its batch number is H1919 which is not working. 

BR,

Awais

 

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

We are analyzing the failure case and for that I request you to answer the following queries.

1) Please share the application and complete schematics for reference. The schematic which you shared is not very clear.

2) Can you please share the failed IC's with local FAE so that we can locate the failure in the IC.

3) Please share the waveform of following pins.

VCC-VSS, ITRIP-VSS, REF-VSS, COM-VSS.

BR,

AZIZ HASSAN

0 Likes
mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Hi Aziz , 

I have uploaded the required waveform for you   

VCC-VSS, ITRIP-VSS, REF-VSS, COM-VSS.

. Please look at this. 

You can see there is signal at low side of the output of gate driver. Due to this low side signal , the low side IGBT turn onn and it   causes the IGBT module to burn  and also gate drive itself.  You can not at the waveform of HO and LO.  can you please check these waveform.  If there is a signal to ITRIP  above 0.5 V it must turn off the all output according to datasheet. 

Please share your email , I will share the complete schematic of the  application. 

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Thank you for sharing the waveforms. 

Please share Vcc-Vss pin waveform as well. I was verifying the value of the  shunt resistors and for that can you please share the load starting current peak value and normal operating trip current value . I strongly believe that the trip current value is much higher (either at starting or after sometime) which is causing the I trip pin voltage value to exceed 5V which is damaging the IC and due to this high current, the power rating of the resisters are getting violated leading to resistor damage . These cascading effects will lead to abnormal behavior of the IC and can damage the device by giving some fault pulse.

Along with the above details, can you please answer following queries.

1) May I know your organization name and country.

2) Please share the complete schematic here for our reference.

BR,

AZIZ HASSAN

0 Likes
mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Hi  , 

Thank you for your support and consideration on this issue. 

VCC to VSS wave is attached.   You are right for the resistor value. We checked the value of shunt resistor and it was 2.6K ohm it mean the shunt resistance burn internally and its value exceed from 0.2 ohm to 2.6 k ohm. We verified the resistance by opening the ceramic material.  The internal resistance components was burnt out.  

I have share the picture of shunt resistance . We are  using (0.475 ohm and 10 watt) power resistance in parallels of 4 resistor in combination.  So its value is 0. 156 ohm in parallels and power rating increases to 40 watt.  I think the problem is due to this type of resistor. They heated up and their contact burn after a  time and gate driver heats up and cause the other components to burn. 

There is one more issue, when we reduce the value of resistance below to 100 m ohm.  Gate driver heats up but at this time resistance value was okey.  

Our organization Name is SolDrive and country is Pakistan. 

Please check if we are using correct type of resistor for R0 and R1 and R2. 

Best regards 

Muhammad Awais.  

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Please provide the maximum fault current you are targeting to be sensed. It is difficult for me to answer the power rating of the resistors to be used unless I know the trip current level or the maximum current which you allow during fault.

Two points we need to consider when we design the resistors 

1) Fault current level 

2) The resistor power rating should be sufficient and the voltage at the Itrip should never exceed 5V. 

In case you are reducing the R0 value (below 100mohm), the current could be very high and consequently  the voltage at I trip could be high resulting in overheating of IC.

Please share your schematic to support you better.

BR,

AZIZ HASSAN

0 Likes
mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Hi Aziz,

 

In our designed application, we are allowing 16A peak current, exceeding this value will generate a 0.5V across itrip and it will reset, for this purpose our calculation are as follows.

1. Four power resistors of 10W, 0.47 ohm in parallel which gives a total 0.117 ohm and 40W, now for 16A, 0.117ohm, power dissipation will become 30W. divider R1 and R2 we are using are R1=1.3K ohm and R2=470 ohm which will gives 0.5V across ITRIP and the voltage across power resistor is 1.88V, with 16A peak current.

2. The voltage is not exceeding 5V across ITRIP.

3. I didn't get your last point about decreasing power resistor value below 100m ohm.

Also if you can suggest power shunt resistors for current sensing which is more ideal with gate driver and the value of divider.

complete schematic and circuit diagram is attached for our application, kindly go through it, if it can help you.

 

Best Regards,

Muhammad Awais

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

As per the trip current, the value of resistors given by you are ok. 

I have few suggestions from design point of view.

1) For sense resistors,  thick film type  having non inductive nature are preferred.

2) Please keep the filter capacitors much closer to I trip pin in order to minimize the noise issue.

BR,

AZIZ HASSAN

0 Likes
mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Hi, 

Can you please suggest any part number series of thik film type resistor.  Yes, we have place capacitor near to Itrip pin. 

 

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Any thick film type resistor is fine. Based upon the space and power rating of the resistor you can choose the package and make sure that the resistor has proper cooling.

Please let me know if I could help you more.

BR,

AZIZ HASSAN

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Below are few of the additional observation which will help you to mitigate the problem.

1) From the schematic, C38/C39 should be between VCC and VSS. But two parts are farther from the pins of the gate driver IC.

2) A ceramic( 100 nF~ 1uF) capacitor between VCC and COM  and VCC and VSS (>4.7uF) is recommended. And the capacitors should be close to GD IC 6ED22.....

3) For sensing,  the resistor which you have used is radial type which has more ESL .Please try to use a SMD thick film resistor with low ESL to minimize the driver loop. 

4) We don't have any limit for minimum value of R0.It is up to level of current protection you need as per the application.

5) To further minimize the spike, cut the trace between COM of 6ED....GD IC and COM of module and add 2.2 ohm resistor between them.

AZIZ_HASSAN_0-1655292388301.png

I hope this will work and please let me know whether the issue is solved or not.

BR,

AZIZ HASSAN

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

Any update on the issue.

BR,

AZIZ HASSAN

0 Likes
mawais
Level 1
10 sign-ins 5 replies posted 5 sign-ins
Level 1

Hi Aziz,

Thank you for your support and sorry for late response as we were working on your suggestion.

We have used a 10 milli ohm and 5Watt resistor, as the drop is low, so we have used the divider through filter circuit and amlifier the signal through gain, when we have used one resistor, it work smoothly but when increased the number of resistor in parallel to increased the power, now the gate driver is working but due to high spike across the resistor when connected to an inductive load three phase supply, it reset the gate driver again and again as a result the output PWM generated by gate driver is not working.

also as you suggested 2.2 ohm resistor on COM, we have used that resistor after adding resistor, the  gate driver is not generating output PWM.

attached is the link of resistor we have used and current spike picture which passed through this resistor on COM.

https://eu.mouser.com/datasheet/2/219/BPR-944283.pdf

 

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

We had a look  at the issue again and we strongly believe that the 2.2 ohm resistor addition between COM and COM of module( emitter of IGBT) would have helped in reducing the noise between COM and VSS as your waveform shows voltage between COM and VSS above spec level. Actually, its like adding a extra resistor in gate path. I cant understand why the PWM has stopped.

For your reference, I am again sending the recommendations for your PCB because the PCB layout doesn't seems to be good to us and I request you to improve it.

  1. Rshunt: (feedback from the customer), three resistor parallel 100mΩ //100mΩ//470mΩ= 45mΩ, so the over current protection is 22.2A with resistor divider (5.1k/5.1k at ITRIP pin). It is ok for me. But It is better to use low ESL SMD resistor for the shunt resistors.
  2. R1 = R2 = 5.1kΩ ,C3= 330 pF ~ 1nF, R2 and C3 could be close to ITRIP and VSS pins.
  3. Decoupling capacitor between VCC and VSS: C1 = 4.7 uF~ 10 uF , Ceramic capacitor. C1 is close to VCC and VSS pins, it is very important.
  4. Decoupling capacitor between VCC and COM: C1 = 100 nF ~ 1 uF , Ceramic capacitor. C2 is close to VCC and COM pins, it is very important.
  5. R2 = 1Ω~2.2Ω which is between COM of 6ED2230S12T and the emitter of IGBTs, in order to filter the noise between COM and VSS.
  6. I have attached a reference diagram for current sense and component placement.AZIZ_HASSAN_1-1656399305813.png

    PFA of a application note for PCB current sense configuration, PCB layout and other recommendations. I hope this will help you more.

     

0 Likes
AZIZ_HASSAN
Moderator
Moderator 100 sign-ins 50 solutions authored 10 likes received
Moderator

Hi,

We had a look  at the issue again and we strongly believe that the 2.2 ohm resistor addition between COM and COM of module( emitter of IGBT) would have helped in reducing the noise between COM and VSS as your waveform shows voltage between COM and VSS above spec level. Actually, its like adding a extra resistor in gate path. I cant understand why the PWM has stopped.

For your reference, I am again sending the recommendations for your PCB because the PCB layout doesn't seems to be good to us and I request you to improve it.

  1. Rshunt: (feedback from the customer), three resistor parallel 100mΩ //100mΩ//470mΩ= 45mΩ, so the over current protection is 22.2A with resistor divider (5.1k/5.1k at ITRIP pin). It is ok for me. But It is better to use low ESL SMD resistor for the shunt resistors.
  2. R1 = R2 = 5.1kΩ ,C3= 330 pF ~ 1nF, R2 and C3 could be close to ITRIP and VSS pins.
  3. Decoupling capacitor between VCC and VSS: C1 = 4.7 uF~ 10 uF , Ceramic capacitor. C1 is close to VCC and VSS pins, it is very important.
  4. Decoupling capacitor between VCC and COM: C1 = 100 nF ~ 1 uF , Ceramic capacitor. C2 is close to VCC and COM pins, it is very important.
  5. R2 = 1Ω~2.2Ω which is between COM of 6ED2230S12T and the emitter of IGBTs, in order to filter the noise between COM and VSS.
  6. PFA of reference design which will act as user guide for PCB design and recommendations for current sense.
0 Likes