IR2113 For VFD use, one driver keeps failing

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

cross mob
lock attach
Attachments are accessible only for community members.
User18812
Level 1
Level 1
Good day,

I am trying to use three IR2113 drivers for a VFD. I have followed the application guide as attached. I am using six STF13N95K3 n-channel MOSFETs whose datasheet is also attached.
Upon inspection of why the drive outputs an unbalanced voltage I found the output of the third driver was giving a constant high output for Ho and a constant low output for Lo.
The bootstrap capacitor furthermore also seemed to not fully charge and Vb-Vs was at 5.66 V whereas the others were at 15.2 V. I replaced that driver and both the two MOSFETs it was connected to and all seemed fine until it happened again.

My switching frequency is 25.6 kHz, using a sine-wave PWM scheme. The bootstrap capacitor calculation yielded 100 nF but it seems it might not be big enough as some of the pulses seems to switch off and then back on rapidly as though the capacitor was depleted.
What could I look at specifically to try and mitigate the problem as why one driver stops functioning and yet the others do not? I tried using a 450 nF bootstrap on all three drivers and it did not revive the one driver.

Kind regards
Ruan Kruger

0 Likes
1 Solution

Hi,

That's great that your problem is solved. We have seen this issue before in other IC as well. The main problem was because of the switching technique or with the capacitor calculation. It happened because of the layout parasitic inductances which perhaps oscillated with the capacitance and does not allow the bootstrap capacitor to charge (even if you keep the capacitor value higher).

Yes, these parasitic do have some impact based upon the switching frequency because, as per different frequency the parasitic impedance varies and so the charge and discharge rate.

BR,

AZIZ HASSAN

View solution in original post

0 Likes
10 Replies
Elisabeth_P_
Employee
Employee
Welcome! First reply posted
Hey Ruan,
thanks for that very interesting question!
Due to the fact that 2 out of three drivers are working as expected, I would say this is more an application problem.
Therefore it would be good to double check you application layout.

Here are some of my recommandations:
1. Is there any difference in position of the driver block? (like further away from switch or load than the others?)
2. Is there a difference in routing of the gate driver block? E.g. are there other functional blocks close, which are not there at the other?
3. Is wiring difference from the control signal the same as with the other 2 drivers? Make sure there is no delay of input signal to this third driver.

For further consulting I would need more information on the circuit layout e.g. a picture of the board and/or the oscilloscope pictures you are describing.

Hope that helped for the beginning!
Best regards,
Elisabeth
0 Likes
User18812
Level 1
Level 1
Hi Elisabeth,

I appreciate the advise. I have ordered a new PCB with a better layout, with more optimised routing of all the signal blocks. I hope this problem does not creep up again.
The old design had longer routing on the third block both on the logic side and the output driver side. If the problem persists on the new design, I would reach back with oscilloscope graphs and the new circuit layout.

Kind regards
Ruan Kruger
0 Likes
sujoy
Level 2
Level 2
10 replies posted 5 replies posted 5 sign-ins
Hi Ryan, Is your problem solved?
0 Likes
User18812
Level 1
Level 1
Hi Sujoy

I have not tested the new PCB with the new layout. At this point in time if anything fails it is going to be too late.
0 Likes
sujoy
Level 2
Level 2
10 replies posted 5 replies posted 5 sign-ins
The reason I asked is that I have a similar issue with IRF2007 for a BLDC application. BTW have you tried ditching the sinusiodal PWM and driving the motor with square waves? Just an Idea
0 Likes
User18812
Level 1
Level 1
I believe my problem helms from a poor layout and wrongfully calculated bootstrap capacitor sizing. My application requires a synthesized sine-wave unfortunately.
0 Likes
sujoy
Level 2
Level 2
10 replies posted 5 replies posted 5 sign-ins
Maybe I didn't explain it properly. You should first see if the system works with a low frequency Square wave PWM at rated load. Sinusiodal waves increase the switching losses and can generate high DV/DT pulses for the motors. I hope your motor wire is shorter than 5-10 ft. Above 20ft you'll need an output choke.
0 Likes
User18812
Level 1
Level 1
Good day, this thread can be closed. My problem has been resolved with the new layout.
0 Likes

Hi,

That's great that your problem is solved. We have seen this issue before in other IC as well. The main problem was because of the switching technique or with the capacitor calculation. It happened because of the layout parasitic inductances which perhaps oscillated with the capacitance and does not allow the bootstrap capacitor to charge (even if you keep the capacitor value higher).

Yes, these parasitic do have some impact based upon the switching frequency because, as per different frequency the parasitic impedance varies and so the charge and discharge rate.

BR,

AZIZ HASSAN

0 Likes
electricuwe
Moderator
Moderator
Moderator
10 sign-ins First question asked 5 sign-ins
Hi Ruan,

my recommendation is to look for another power device first. The body diode of HV Si-Mosfets generally has very poor reverse recovery behaviour. But his is mandatory for any application where an inductive current is to be commutated. Change to fast IGBT with a co-packaged diode or an SiC-Mosfet.


Best regards,


electricuwe
0 Likes