David_
Advanced Member level 2
- Joined
- Dec 6, 2013
- Messages
- 573
- Helped
- 8
- Reputation
- 16
- Reaction score
- 8
- Trophy points
- 1,308
- Location
- Sweden
- Activity points
- 12,242
Hello.
I have designed this LTC2864 circuit which is a 40V to 3-33V step-down regulator:
But I have some trouble with some aspects and I really need some direction, lets start with the easiest stuff.
1, My max output current is 3A, but I don't quite get how these things function. Does I need to rate the rectifier for a 5A load since my current limit is set to somewhere there about?
I mean at the time when 3A is pulled at the output, is the inductor charged with up to 5A continuously or what?
Thats another big problem for me, the inductor. I have been since a year back maybe been looking at cores and I have spent time measuring BH curves but I find my self lost. I suppose I need to find a core that can handel 5A without saturation which I'll define as a 10% drop of inductance, but when looking at cores saturation is not given as a current but as Bmax.
2, How do I relate Bmax to a current value?
3, I have read things that makes me a little scared by secondary output filters, I om using this circuit as a pre-regulator for a lab supply and really do care about getting as fine a output as I can which is why I have chosen a big inductor and want to have a second LC filter in the form of a coil, then cap, then ferrite bead, then a cap again. App notes have told me to calculate the LC cut of to be 1/10 of the switching frequency and for the moment we'll assume that to be 350kHz alternately 100kHz.
Funny thing, I can't at this time remember why that filter could cause problems but I suppose it has to do with the feedback loop?
4, I've found a very alarming text telling a very convincing story about how the market has pushed out the thought that switchers are simple and that many here and there preach about things they does not really understand and as such very few people are even aware about NIR(Negative Input Resistance) which according to this text easily causes the input to oscillate which does not have to be noticed at the output but it increases power consumption.
And it says that the recommended input capacitor set up is all wrong and to low ESR will cause these problems, I have read many things about that I should get the absolutely lowest possible ESR caps for the input and output since the ESR is proportional to output ripple. And if true the use of cheap electrolytic's saves the day as they act like dampening with high ESR. Every app not says that ceramic is the way to go but apparently you need to have a resistor in series with the input ceramic since it has to low ESR... Though the math for calculating this resistor/cap was above me.
Anyone having anything to say about this?
I'll find the site and post it here later.
5, Loop compensation.
I have read the data sheet and other sources but I can't understand, can anybody explain what this is and why its needed in words I might understand.
My regulator is feed with 39.6Vdc, I have forgotten how the load regulation of my transformer is but it was quite good. But the output will be between 3 and 33V and the loop compensation is determined for specific output voltage. How do I manage this then?
6, I can recall briefly reading about compensating the winding capacitance of a toroidal core inductor with a second winding biased with a current that effectively canceled the capacitance.
In a buck-regulator the winding capacitance couples high frequency noise through to the output from the edges of the square switching voltage, the higher the capacitance the more high frequency content can propagate through this capacitance. Then its in my best interest to wind my coil to minimize this capacitance.
I assume there are some winding techniques that I can try to adder to in order to achieve this but would compensating it with a second winding be a possible solution in a switching regulator?
Well that's quite a lot, I would be very happy if there where any answers to these questions out there and all thoughts and opinions are very much welcome.
I have designed this LTC2864 circuit which is a 40V to 3-33V step-down regulator:
But I have some trouble with some aspects and I really need some direction, lets start with the easiest stuff.
1, My max output current is 3A, but I don't quite get how these things function. Does I need to rate the rectifier for a 5A load since my current limit is set to somewhere there about?
I mean at the time when 3A is pulled at the output, is the inductor charged with up to 5A continuously or what?
Thats another big problem for me, the inductor. I have been since a year back maybe been looking at cores and I have spent time measuring BH curves but I find my self lost. I suppose I need to find a core that can handel 5A without saturation which I'll define as a 10% drop of inductance, but when looking at cores saturation is not given as a current but as Bmax.
2, How do I relate Bmax to a current value?
3, I have read things that makes me a little scared by secondary output filters, I om using this circuit as a pre-regulator for a lab supply and really do care about getting as fine a output as I can which is why I have chosen a big inductor and want to have a second LC filter in the form of a coil, then cap, then ferrite bead, then a cap again. App notes have told me to calculate the LC cut of to be 1/10 of the switching frequency and for the moment we'll assume that to be 350kHz alternately 100kHz.
Funny thing, I can't at this time remember why that filter could cause problems but I suppose it has to do with the feedback loop?
4, I've found a very alarming text telling a very convincing story about how the market has pushed out the thought that switchers are simple and that many here and there preach about things they does not really understand and as such very few people are even aware about NIR(Negative Input Resistance) which according to this text easily causes the input to oscillate which does not have to be noticed at the output but it increases power consumption.
And it says that the recommended input capacitor set up is all wrong and to low ESR will cause these problems, I have read many things about that I should get the absolutely lowest possible ESR caps for the input and output since the ESR is proportional to output ripple. And if true the use of cheap electrolytic's saves the day as they act like dampening with high ESR. Every app not says that ceramic is the way to go but apparently you need to have a resistor in series with the input ceramic since it has to low ESR... Though the math for calculating this resistor/cap was above me.
Anyone having anything to say about this?
I'll find the site and post it here later.
5, Loop compensation.
I have read the data sheet and other sources but I can't understand, can anybody explain what this is and why its needed in words I might understand.
My regulator is feed with 39.6Vdc, I have forgotten how the load regulation of my transformer is but it was quite good. But the output will be between 3 and 33V and the loop compensation is determined for specific output voltage. How do I manage this then?
6, I can recall briefly reading about compensating the winding capacitance of a toroidal core inductor with a second winding biased with a current that effectively canceled the capacitance.
In a buck-regulator the winding capacitance couples high frequency noise through to the output from the edges of the square switching voltage, the higher the capacitance the more high frequency content can propagate through this capacitance. Then its in my best interest to wind my coil to minimize this capacitance.
I assume there are some winding techniques that I can try to adder to in order to achieve this but would compensating it with a second winding be a possible solution in a switching regulator?
Well that's quite a lot, I would be very happy if there where any answers to these questions out there and all thoughts and opinions are very much welcome.