It is like this already, i built it carefully.
Ground loops are small and bypass capacitors are close to every IC.
The ground is from LOW side mosfet source. The high side mosfet source is the floating one.
I dont know what happened before, probably strong oscillations somwhere damaging the IC.
Ground loops are small and bypass capacitors are close to every IC.
The ground is from LOW side mosfet source. The high side mosfet source is the floating one.
I dont know what happened before, probably strong oscillations somwhere damaging the IC.
I tried it and it was not feedback's fault.
When the duty cycle reaches around 65 - 70%, the oscillation occours. This oscillation can kill the IR2110 and SG3525.
I cant understand why this happens.
Maybe shoot-through?
Can i solve this?
When the duty cycle reaches around 65 - 70%, the oscillation occours. This oscillation can kill the IR2110 and SG3525.
I cant understand why this happens.
Maybe shoot-through?
Can i solve this?
Shoot through can occur, but only f ir2110 misbehave because of negative voltage spikes.
Probably it's time to check waveforms woth scope at variuos places. Ir2110 have shoot through protection , but it fails sometimes .
Probably it's time to check waveforms woth scope at variuos places. Ir2110 have shoot through protection , but it fails sometimes .
Dear OP, you've made a lot of fundamental mistakes:
In addition to the above, you're trying to make a power converter without an oscilliscope and a proper PCB. You seem to underestimate power conversion altogether ...
- Did you start by drawing up a clean and complete circuit diagram ? Cite the same while asking for help, as others need not know / remember what your circuit exactly is.
- Did you measure your pulses duty cycles before feeding them into the power converter ? I remember having said this twice, you dissed it off.
- Did you vary the pulse-width using a potentiometer and verify the same within the pulses ?
- Did you confirm the dead-time and absence of shoot-through ? If yes, how ?
- Did you verify the choke doesn't saturate / change value when load current is driven through it ?
- Did your converter run open loop before taking any feedback around it ?
In addition to the above, you're trying to make a power converter without an oscilliscope and a proper PCB. You seem to underestimate power conversion altogether ...
I do what i can... I cant have a scope for now.
When the duty cycle is raised over a certain limit, this strage behaviour happens.
Also the deadtime on SG3525 outputs is still present.
But i didnt understand why by combining the two outputs with two diodes makes the deadtime disappear. By logic is should not...
Without an oscilloscope it is basically impossible to verify stuff properly.
I will still try, somehow, with a lot of patience and calculus to solve this problem.
The XOR gates circuit is now present between the SG3525 and IR2110.
I should measure the waveform on the 470R resistor to see if the SG3525 with combined outputs is working as it should.
Then, measure at the input of IR2110 to see if the XOR gates circuit is doing what it should.
Then, measure the half bridge waveforms and inductor waveform.
Yes, but how without an oscilloscope?
Trial and error can work but in my experience its not worth it. It is really tiring and can result in tons of burnt components.
IR2110s are expensive at around €1.50 each. (infineon)
I am still a beginner...
After all, i'm just 19.
I still have experience to make.
When i obtain an oscilloscope i will finally be able to do much more.
When the duty cycle is raised over a certain limit, this strage behaviour happens.
Also the deadtime on SG3525 outputs is still present.
But i didnt understand why by combining the two outputs with two diodes makes the deadtime disappear. By logic is should not...
Without an oscilloscope it is basically impossible to verify stuff properly.
I will still try, somehow, with a lot of patience and calculus to solve this problem.
The XOR gates circuit is now present between the SG3525 and IR2110.
I should measure the waveform on the 470R resistor to see if the SG3525 with combined outputs is working as it should.
Then, measure at the input of IR2110 to see if the XOR gates circuit is doing what it should.
Then, measure the half bridge waveforms and inductor waveform.
Yes, but how without an oscilloscope?
Trial and error can work but in my experience its not worth it. It is really tiring and can result in tons of burnt components.
IR2110s are expensive at around €1.50 each. (infineon)
I am still a beginner...
After all, i'm just 19.
I still have experience to make.
When i obtain an oscilloscope i will finally be able to do much more.
XOR 4070 nemůže řídit 1n kondenzátor na 100 kHz, musíte snížit na max 100 pF. IR2110 nemůže řídit 17 nF na 100 kHz, musíte přidat gatedriver zůstávající ze 4 mosfetů na každou bránu zvlášť. Otherwise, damage will occur and explosion.
Last edited:
Hi,
I wish to design a synchronous buck converter using sg3525 and NCD57084DR2G gate driver IC. I have been instructed to use high frequency mosfet and also or gates for some reason. I am new to this and have got no clue as to how to design one and how to do the calculations. I have researched a bit and just know that my pwm ic will pass on a signal to the gate driver ic while will increase the switching frequency in order to drive the mosfet. However this is just a gist and am not able to proceed further in the designing part. Please help me out with this.
I wish to design a synchronous buck converter using sg3525 and NCD57084DR2G gate driver IC. I have been instructed to use high frequency mosfet and also or gates for some reason. I am new to this and have got no clue as to how to design one and how to do the calculations. I have researched a bit and just know that my pwm ic will pass on a signal to the gate driver ic while will increase the switching frequency in order to drive the mosfet. However this is just a gist and am not able to proceed further in the designing part. Please help me out with this.
OmV7 , Hi. You wanna use driver ic ,which is designed to drive IGBT , not N-Ch mosfet ,but it may work .For synchronous buck converter (step down voltage) this driver ic would control upper mosfet , but for controlling bottom mosfet you need to sense current and turn off mosfet when current stops flowing (same as diode stops conducting when there's no voltage difference for it to turn on) .This function can't be performed by sg3525 . You need synchonous buck converters controller ic , with either current sensing capability of bottom mosfet through a shunt ,or transformer , or alternatively by built in zero crossing comparator .Otherwise if you just use both separate outputs of 3525 , each output with its own driver ic , your converter will kinda work ,but not optimally at all, a current flow back will occur . Only in some load range that simplified method works . With low load or no load condition As alternative ,you can try to use a ic ,designed for flyback syncronous rectification ,to control bottom mosfet through driver .As example UCC24630 , i'm using it in my flyback designs, just you have to play with voltage dividers . In cooperation with mosfet it emulates the diode ,but with low heat . Another alternative is TEA1791T ic , to control bottom mosfet , but it has a big drawback , it controls mosfet with linear region , for faster turn-off , but then external driver can't be connected to it ,tea1791 must control external mosfet directly ,and that mosfet heats up more than with ucc24630 . So designing something is big challenge ,to make it work properly . Also , you didn't specify voltage and current ranges ,maybe there are already proven designs .