I've had time to tinker for the first time in a long time. I have been working on an amplifier in the style of the linked patents, which has a linear error correcting amplifier working in tandem with a switching amplifier.
US Patent 4,523,152
US Patent 5,905,407
I implemented the IR2010 gate driver after the IRAUDAMP9 gate driver example and ran into some unexpected behavior. When the switching part of the amplifier (disconnected from everything else) has no logic power supply, but does have its main supply, its output was going high or periodically oscillating to high (brief blips at 50Hz or 120Hz intervals). After much debugging, the only thing that fixed it was to eliminate the bootstrap charging resistors which charge the high side floating power supply before the logic supply takes over via the bootstrap diode.
Is anyone familiar with this problem of self turn on in the absence of a logic supply? I can't tell from the datasheet whether the IR2010 outputs are specified to be high, low, or open in the absence of its logic power supply.
In my design I have a 10kOhm resistor to floating return on the output of the transistor buffer, but I omitted any resistor to return on the input to the buffer. I thought that might have been an oversight, and soldered in a resistor manually and still ran into problems with self turn on. This happened with that stage loaded or unloaded.
The switching part of the amplifier is now happily working without the bootstrap charging resistors, and the switching waveforms do not have excessive ringing on the oscilloscope. I'd like to try to fix this though, because now there is a 100uS long large voltage blip after turning on the amplifier due to the bootstrap being uncharged, which wasn't there earlier.
The gate driver circuit is attached, the resistor in question is RB57, which appears to causes unwanted turn on in the absence of the logic supply when the main supply is active.
US Patent 4,523,152
US Patent 5,905,407
I implemented the IR2010 gate driver after the IRAUDAMP9 gate driver example and ran into some unexpected behavior. When the switching part of the amplifier (disconnected from everything else) has no logic power supply, but does have its main supply, its output was going high or periodically oscillating to high (brief blips at 50Hz or 120Hz intervals). After much debugging, the only thing that fixed it was to eliminate the bootstrap charging resistors which charge the high side floating power supply before the logic supply takes over via the bootstrap diode.
Is anyone familiar with this problem of self turn on in the absence of a logic supply? I can't tell from the datasheet whether the IR2010 outputs are specified to be high, low, or open in the absence of its logic power supply.
In my design I have a 10kOhm resistor to floating return on the output of the transistor buffer, but I omitted any resistor to return on the input to the buffer. I thought that might have been an oversight, and soldered in a resistor manually and still ran into problems with self turn on. This happened with that stage loaded or unloaded.
The switching part of the amplifier is now happily working without the bootstrap charging resistors, and the switching waveforms do not have excessive ringing on the oscilloscope. I'd like to try to fix this though, because now there is a 100uS long large voltage blip after turning on the amplifier due to the bootstrap being uncharged, which wasn't there earlier.
The gate driver circuit is attached, the resistor in question is RB57, which appears to causes unwanted turn on in the absence of the logic supply when the main supply is active.
Attachments
Last edited:
In products with this device that we developed some time ago, that resistor was never necessary.
If I recall, the SD can be used for startup. Check their old app notes.
http://www.farnell.com/datasheets/2254418.pdf
http://web.mit.edu/6.131/www/document/float_drive.pdf
If I recall, the SD can be used for startup. Check their old app notes.
http://www.farnell.com/datasheets/2254418.pdf
http://web.mit.edu/6.131/www/document/float_drive.pdf
Last edited:
Thank you for the references. Indeed, the pre-charging resistor is nowhere to be found.
I have the shutdown pin going to a pull-up resistor to hold it off until commanded to turn on. This mechanism is effective while the chip logic supply is powered.
I did another experiment because I thought maybe I have too low a quiescent current to the bootstrap supply and that is causing it to be indeterminant. I changed the pre-charging resistor to increase the current above the quiescent threshold in all circumstances. This only held the MOSFET output up better. In this arrangement, the measured MOSFET output voltage was fixed at 10 volts lower than the high voltage supply.
So I am currently omitting the pre-charging resistor, and there is not unwanted voltage on the output before turning on. However, right now, I am living dangerously without an output relay and there is a small transient for the period that the bootstrap is charging. I might use a switchable current source to only charge the bootstrap when the logic supply is present.
I have the shutdown pin going to a pull-up resistor to hold it off until commanded to turn on. This mechanism is effective while the chip logic supply is powered.
I did another experiment because I thought maybe I have too low a quiescent current to the bootstrap supply and that is causing it to be indeterminant. I changed the pre-charging resistor to increase the current above the quiescent threshold in all circumstances. This only held the MOSFET output up better. In this arrangement, the measured MOSFET output voltage was fixed at 10 volts lower than the high voltage supply.
So I am currently omitting the pre-charging resistor, and there is not unwanted voltage on the output before turning on. However, right now, I am living dangerously without an output relay and there is a small transient for the period that the bootstrap is charging. I might use a switchable current source to only charge the bootstrap when the logic supply is present.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.