High current through low side mosfet while charging bootstrap capacitors

Thread Starter

MSemz

Joined Jan 19, 2021
14
I'm applying the following circuit:
Untitled.png
The only way I found for charging the bootstrap capacitor is through M1. And when M1 is opened I found high current goes through it for first cycle as follows.
Untitled1.png
This high current is for charging bootstrap capacitor. But what to do to reduce this current ? I though about finding another path to charge the capacitor but I don't know what is the right solution.
 

MisterBill2

Joined Jan 23, 2018
18,167
One obvious option will be to add a resistance in series with the drain of M1. Try 100 ohms as a starting value.
Also, check the circuit timing, since if ever M1 and M2 are conducting at the same time there is nothing in the circuit to limit the current delivered from V1
 

Thread Starter

MSemz

Joined Jan 19, 2021
14
One obvious option will be to add a resistance in series with the drain of M1. Try 100 ohms as a starting value.
Also, check the circuit timing, since if ever M1 and M2 are conducting at the same time there is nothing in the circuit to limit the current delivered from V1
I'm having a deadtime of 200ns and in the graph the high current is on the first period only.
I though of the resistor option but it will has a high voltage drop on it so I'm asking if there is another option.

I though also to have transistor between the bootstrap capacitor and the ground with series resistor and enable it at first time only to charge the capacitor but I don't know if it will work or not. What do you think about this option ?
 

crutschow

Joined Mar 14, 2008
34,281
I would not be concerned. That power MOSFET can readily handle the short surge current.
But if you need to, you could limit it with a small resistor in series with C1 (say 1 ohm).
That will have no significant effect on the drive pulse to M2.
 

MisterBill2

Joined Jan 23, 2018
18,167
Certainly a series resistor will have a high voltage drop across it for a short time. That is the whole purpose of having a series resistor. You can see that after only 20 microseconds the current is down to 5 amps, while at the peak it is 45 amps. Another option would be to remove C1, because that seems to be where the high current goes.
And while I am not familiar with the specifications of transistor M1, that is a very large current spike.
 

Thread Starter

MSemz

Joined Jan 19, 2021
14
I would not be concerned. That power MOSFET can readily handle the short surge current.
But if you need to, you could limit it with a small resistor in series with C1 (say 1 ohm).
That will have no significant effect on the drive pulse to M2.
OK I will try if the mosfet will handle this current. Thank you.

@MisterBill2 thank you.
 

shortbus

Joined Sep 30, 2009
10,045
Does C1 need to be such a high value? Usually such capacitors are 100nF to 1uF.
The TS should use the link Maxheadroom always gives - https://www.silabs.com/support/isolation-bootstrap-calculator That calculator is for the size of cap that is needed.

If the the capis way too big it would be the reason for high current on just the first cycle. On all of the ongoing cycles the cap may be partially charged and not requiring the high current to top it off.
 

MisterBill2

Joined Jan 23, 2018
18,167
Sure shows you don't know what or how bootstraps work with highside gate drivers.
It does not seem that a mosfet needs anywhere near that much current for the gate drive. And given that the circuit shown does not seem to have any useful purpose, why would M2 need such a huge current source? The 100n cap should be able to provide adequate drive by it's self.
 

crutschow

Joined Mar 14, 2008
34,281
The 100n cap should be able to provide adequate drive by it's self.
That's probably a little marginal, especially if the supply voltage is only 10V.
The total gate charge for that MOSFET is 250nC max. which would cause the cap volt to drop a couple volts or so to provide that charge.
Thus the cap value should probably be at least a few hundred nF.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
The link for the cap that Shortbus posted plus this blog info is what I use and no problems so far, I have used 5v & 15v also 15v for both for VCC & VDD, I keep the LO side on and switch the Hi side.
Max.
 

shortbus

Joined Sep 30, 2009
10,045
It does not seem that a mosfet needs anywhere near that much current for the gate drive. And given that the circuit shown does not seem to have any useful purpose, why would M2 need such a huge current source? The 100n cap should be able to provide adequate drive by it's self.
I stand by my earlier post-
Sure shows you don't know what or how bootstraps work with highside gate drivers.
And no I won't explain it to you.
 

Thread Starter

MSemz

Joined Jan 19, 2021
14
I added a resistor of 4 ohms as follows
1612763773675.png

and the current through it is as follows
1612763838457.png

The question is Can I use a 0.25W resistor as it's only one spike happens in the beginning only ? or it will be damaged immediately and I should use 5W resistor ?
 

crutschow

Joined Mar 14, 2008
34,281
Can I use a 0.25W resistor as it's only one spike happens in the beginning only ?
No problem.
A quarter watt resistor is more than adequate.
The energy dissipated in the single spike is very small, which is readily absorbed by that size resistor.
 

MisterBill2

Joined Jan 23, 2018
18,167
Why not put the resistor in series with the drain connection of M1, since then it will have no effect on the discharge path of the capacitor..
 

crutschow

Joined Mar 14, 2008
34,281
Why not put the resistor in series with the drain connection of M1, since then it will have no effect on the discharge path of the capacitor..
Obviously that will increase the on-resistance of M1.
Whether that's a problem depends upon the circuit output load.

The 4 ohm resistor in series with the capacitor just adds to R4 in series with the the gate of M2 when M2 is turned on.
That will have negligible effect on M2's switching time.
 
Last edited:

shortbus

Joined Sep 30, 2009
10,045
The most obvious and workable answer is to pick the correct boot cap. The one in the link of post #10, the one Max uses. It takes into account both the needed gate voltage/capacitance and the gate on time, no "extra resistors" needed.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
When the IR2110 first came out, IR handed out some simple PCT design kits based around the IR2110, I still have them.
A very simple basic circuit.
Max.
 
Top