Voltage rating of bootstrap cap for IR2111

[jsteev126]

Hi all,
I am putting together a half bridge motor driver using a IR211 and two IRF640s.
The voltage going thru the MOSFETS is about 100V. My concerns are; I think the cap connected between VS and VB (C1) should be rated at about 200V. Is this correct? Can anyone see any problems with the values of the caps, resistors and diode? I've spent hours searching the net for info on this; there doesn't seem to be much around. Any help, deeply appreciated.

 

[Tahmid]

Hi,
Wrong, the capacitors need to be rated at atleast 2 x VCC = 2 x 12 = 24vdc, not 2 x vpower = 200vdc
You can use 200v caps, but just a waste.
Use low ESR caps.
You've used a 4.7u cap across VB and VS. Use one across VCC and COM as well.
Connect a 1k resistor from across the gate to source of both MOSFETs to prevent accidental turn-on.
Besides this, everything seems to be fine. I don't know your frequency of the PWM, but I don't think it's very low.
If your frequency's above 500Hz, everything's fine. Else, increase the value of the bootstrap caps.
Hope it helped.
Tahmid.

 

[karesz]

Hi,
I think you must check the top limit of working vltage as possibly low.
Because if you have capacitors as Y5U and same types_ their has much capacitance loss (i.e. in worst case up to 30%) if working voltage is very low to nominal voltage!
It can be slight different between producer, but datasheets diagrams are to study.
K.

 

[jsteev126]

Many thanks for the advice. You have helped a lot.

 

[karesz]

Hi jsteev126,
Tnx!
You are welcome!
K.

P.s.:
The discussed capacity loss in relationship working to nominal voltage is only par % i.e. by X7R & negligible by X5R types too...

 

[FvM]

I think it's not exactly negligible. A 16V X7R capacitor has about 20% capacitance loss at 12V, at typical gate driver voltage.
This makes me prefer 25V capacitors for this application.

 

[jsteev126]

Thanks for your input everyone. I have posted the revised schematic. I can not get hold of low ESR caps here, so I have to make do with the components I have. One more question, the diode I plan to use is a SB340, 40V 3A. Do you think this is ok?

 

[Tahmid]

Hi,
SB340 should be enough. I used BYV26C as that is what I had in hand. It was 600V 1A.
Everything seems fine, but you should also place a 0.1uf ceramic cap across the 4.7uf one between Vcc and Com.
You could, for safety, use an inductor at the input to prevent excessive inrush current as you have a large bulk capacitor (120uF) which could pull a great amount of current during charging.
Tahmid.

 

[jsteev126]

Tahmid, Thanks heaps for your help. I'll add that cap as suggested. I'm not sure where to put the inductor. Would you be able to elaborate? ie. where to put it and a suggested value.

 

[Tahmid]

Hi,
You have the 90v power supply which you connect to H-bridge and cap positive, right? Instead now, connect the 90v to the inductor and the other end of the inductor to the H-bridge and the cap positive. Thus the 90v will flow through the inductor and then to the power stage.
Anything around I guess 100 - 300uH should do. This does not have to be a strict value just an inductor needs to be present. 100uH should suffice. And also add a fuse for safety.
Hope this helped.
Tahmid.

 

[jsteev126]

Hi Tahmind. I think I'm almost there, thanks to you. There is a lot of current going thru that inductor, so is this https://www.marutsu.co.jp/user/shohin.php?p=2700 type suitable? I don't have much experience with inductors, so I'm not sure.

 

[Tahmid]

Hi,
Looks pretty good. Rated at 5A. So it should be safe for loads of upto 300W. Maybe, you can go higher but I wouldn't go over that for sake of safety.
Tahmid.

 

[karesz]

Dear FvM,
I have wroten it only in relation to Y5U and some similar types, ok it was not so clear wroten_sorry!
K.
I found it yet (randomly)...

 

[jsteev126]

Many thanks to all who helped me with this. I have posted what I think will be the the final schematic. Added the inductor and upgraded the diode (the diode maybe overkill, but just found it in my junk-box)

 

[Tahmid]

Hi jsteev126,
It seems fine now.
Hope it'll work fine.
Tahmid.

 

[FvM]

Dear karesz,

I didn't intend to critisize your statement, just mention, that there's still a considerable voltage dependency with "good" ceramics.

Regarding the suggested 100µH + 120µF bus voltage filter, I doubt if it's generally a good idea. It has a resonance frequency of about
1500 Hz and can produce severe oscillations with respective load steps. Usually, we try to eliminate all supply bus inductances.

 

[karesz]

Lieber FvM,
Kritik wäre mir auch kein Problem, denn keine ist ja unirrbar nur die Konjuktionen sind mir so problematisch, dass ich mich kaum verständlich ausdrücken kann.....
I have wished only to tell, that (as you told it yet too) even X7R has voltage dependent capacitance_other subject, that the same dielectrica from different producer is (in these relation)NOT exactly the SAME; for exact characteristics we must know the concrete types from the concrete producer...
OK, these is not "every days" problem & not for every body, but in my opinion is better if more developer are thinking on such "side effects" too.
K.

 

[jsteev126]

Hi FvM,
Would you be able to suggest an alternative to the LC arrangement or point me to a website that can provide some good tuition about it?
jsteev126

 

[FvM]

If the 120 uF is the only bus capacitor and supplied directly from the input rectifier, an inductance can be
meaninful as a current limiter. But it would be placed usually on the AC side of the rectifier.

 

[Tahmid]

Hi,
I've used inductors on the DC bus to limit inrush current as 120uF on a HV DC bus will take a somewhat great amount of current while charging, and limiting this current is necessary.
Tahmid.