1) You'd need to drive them in such a way that the voltage is evenly divided. Any slight asymmetry (timing error) and your FETs are toast.
2) Higher VDSmax goes with higher Ron: http://en.wikipedia.org/wiki/Power_MOSFET#Breakdown_voltage.2Fon-state_resistance_trade-off
I have!
Hi to everybody,
I have builded so kind of a SMPS with ETD49 with IR2113 and another circuit I used driver transformer. My ETD49 magnetics P material and it wound on only 24sp primary and 2*8sp sec for +-37V 8A output. I am working on the new updated version of the SPS's pcb. But ý would like to say the circuit uses driver transformer is more efficient then the gate driver version. I am enclosing the picture of circuits and scope results of my circuit.
Best regards
medogrizli said:
Hi to everybody,
I have builded so kind of a SMPS with ETD49 with IR2113 and another circuit I used driver transformer. My ETD49 magnetics P material and it wound on only 24sp primary and 2*8sp sec for +-37V 8A output. I am working on the new updated version of the SPS's pcb. But ý would like to say the circuit uses driver transformer is more efficient then the gate driver version. I am enclosing the picture of circuits and scope results of my circuit.
Best regards
I take a look at the SG3525.PDF and there stays: Low power transformer can bi DIRECTLY driven from sg3525: So why extra transistors, isnt SG3525 enough "strong# to drive this ferit that is used for driving mosfets?!!
what about DEADTIME when using FERIT for driving mosfets (half bridge method) : is it possible in that case BOTH mosfets are open ?!!
How to prevent that?
thanks!!
can anyone send BOOTSTRAP method for driving upper mosfet(that has not gate at ground)
thanks
what about DEADTIME when using FERIT for driving mosfets (half bridge method) : is it possible in that case BOTH mosfets are open ?!!
How to prevent that?
thanks!!
can anyone send BOOTSTRAP method for driving upper mosfet(that has not gate at ground)
thanks
Drive, He Said
Medo-
From the ONSemi datasheet, the SG3525AN can drive, steady-state, +/-100mA, peak +/-400mA. If you know both the gate capacitance and the reverse transfer capacitance of the MOSFETs you want to drive, then a simple equation (sorry, I don't have it in front of me) will enable you to calculate if the SG3525 you are using will provide enough drive for the MOSFET.
In cases where either of these two specs are high, then an external PNP-NPN buffer driver may be necessary to prevent overheating of the PWM chip, and ensure fast switching times.
The same thing holds true for directly driving those "low power" driver transformers. Just a simple plug-n-chug procedure will help determine whether your PWM chip will be able to drive the xfmr or need some help from the aforementioned buffer driver.
A SMPS I recently did used four MTY100N10Es (100V, 100A) MOSFETS. Their gate capacitance was something like 11,000pF EACH. You bet I used a buffer driver! Switching frequency was 40kHz (clock 80kHz), and switching wavefroms to the MOSFET gates were fast, nice, and clean. I'm pretty sure I wouldn't have seen such nice wavefroms without the buffer. The NPN-PNP pair I used were the MJE243-253 complimentary pair (100V, 4A).
As for driving two IRFP460s in half-bridge fashion, wine&dine is right, and here's why: a perfectly symmetrical drive to both the lower and upper MOSFETs is absolutely necessary so the main transformer core won't eventually "walk" toward saturation from an uneven magnetization of the core (as illustrated by an asymmetrical curve on the B-H graph). Driving both MOSFETs from the same driver transformer will ensure this.
Hope this little diatribe helps make a better understanding about driving MOSFETs from PWM chips.
Steve
Steve
Medo-
From the ONSemi datasheet, the SG3525AN can drive, steady-state, +/-100mA, peak +/-400mA. If you know both the gate capacitance and the reverse transfer capacitance of the MOSFETs you want to drive, then a simple equation (sorry, I don't have it in front of me) will enable you to calculate if the SG3525 you are using will provide enough drive for the MOSFET.
In cases where either of these two specs are high, then an external PNP-NPN buffer driver may be necessary to prevent overheating of the PWM chip, and ensure fast switching times.
The same thing holds true for directly driving those "low power" driver transformers. Just a simple plug-n-chug procedure will help determine whether your PWM chip will be able to drive the xfmr or need some help from the aforementioned buffer driver.
A SMPS I recently did used four MTY100N10Es (100V, 100A) MOSFETS. Their gate capacitance was something like 11,000pF EACH. You bet I used a buffer driver! Switching frequency was 40kHz (clock 80kHz), and switching wavefroms to the MOSFET gates were fast, nice, and clean. I'm pretty sure I wouldn't have seen such nice wavefroms without the buffer. The NPN-PNP pair I used were the MJE243-253 complimentary pair (100V, 4A).
As for driving two IRFP460s in half-bridge fashion, wine&dine is right, and here's why: a perfectly symmetrical drive to both the lower and upper MOSFETs is absolutely necessary so the main transformer core won't eventually "walk" toward saturation from an uneven magnetization of the core (as illustrated by an asymmetrical curve on the B-H graph). Driving both MOSFETs from the same driver transformer will ensure this.
Hope this little diatribe helps make a better understanding about driving MOSFETs from PWM chips.
Steve
Steve
Re: Drive, He Said
ok thanks for answering
what about L6384 : can it be usead instead of "driving transformer" and can SG3525 outputs be connected to this L6384 cause it has only one "input"
when uasing driving transofermr how to ensure that both fets are NOT ON in the same time: When using push pull (without driving transformet) than deadtime can bi directly set by resistance on SG3525 pin...
But pulse transfer through driving transifrmer i a little bit different:
Can I when changing this resistance on sg3525 for deadtime when using riving transformer method CHANGE deadtime as wel as when using ONLY sg without driving transforemr in push pull mode
thanks?
N-Channel said:
ok thanks for answering
what about L6384 : can it be usead instead of "driving transformer" and can SG3525 outputs be connected to this L6384 cause it has only one "input"
when uasing driving transofermr how to ensure that both fets are NOT ON in the same time: When using push pull (without driving transformet) than deadtime can bi directly set by resistance on SG3525 pin...
But pulse transfer through driving transifrmer i a little bit different:
Can I when changing this resistance on sg3525 for deadtime when using riving transformer method CHANGE deadtime as wel as when using ONLY sg without driving transforemr in push pull mode
thanks?
Danko said:
Hi DANKO and the others,
1.Because "driver tranformer version" has a different shematic.
2.555 and 393 : 393 for erors against to over temperature and over current and it drives 555. 555 tirgered by 393 and stanbys the 3525 chip for 3-5 second for erors.
3. I will redraw the driver transformer version's shematic and post it as soon as possible.( it is hand drawn by me so may cause to be misunderstanding)
4. output inductors from an old pc PSU and rewound. I am not shure but they must be micrometals T-106-26 core. The circuit my old desing and I can not remember the munber of turns for the output inductors but I will look my circuit and measure the inductor value and post all of them as soon as possible.
Please let me a bit time to post.
By the time I can answer your questions.