| jamesrnz |
so i searched around with no avail.
i was thinking of using a mc33025 for a buck ps.
however to get the outputs together i thought about simple diode anodes on pin 14 and 11 then bringing the cathodes together,
any thoughts
i want to take 12-18 volts down to 6 to 8 volts.. |
|
|
| luka |
Hi
MC33025 will be great. I wanted too use it for buck too. You will need P-fets and save yourself a lot of trouble. If it is more powerfull you could use 2 fets, each on one output and drive same output inductor, or same it two phase one.
You will need good output diode, that will be with enough current and very fast
Maybe you will do something like this
 |
|
|
| JensRasmussen |
James
I think you should consider a part that supports active rectification (Synchronous buck).
This will give you the benefit that you can use N-channel mosfets and you will always run the converter in continuous conduction mode.
N-channel mosfets generally have lover Rds on and active rectification means that the regulation will be faster than a standard rectifier solution.
Maxim and TI have some nice parts that might support your needs without you having to think about gate boosters.
What output current are we looking at here?
\Jens |
|
|
| jamesrnz |
oh
somewhere around 16 to 18 volts at 50 amps.
i'll have to look into specific chips.
i was thinking about driving an n-ch with a transformer or ir21xx |
|
|
| Danko |
At 50 Amps sync. rectification gives you better performance, but increases circuit complexity.
Using a simple diode (maybe more paralelled) is very simple, but it gives little bit lower efficiency.
you can not use gate driver transformer, becouse the gate driving signal is not symmetric, and the core (the transformer's core) will saturate. As fas as I know ...
IR2113 is good.
50 Amps, oh my God! :-)
You will need as thick wires, as my finger :-D |
|
|
| luka |
Hi
But why even use IR or any kind, if you use P-channels, which aren't that bad, you will save money, number of components, circuit complexity. And efficiency can't be that bad, it won't be 95%, but higher then 85% for sure. Also what freq. will you be going for?
You can't use gate driver transformer as Danko said |
|
|
| jamesrnz |
oh 75 to 200 khz...........
any suggestions on p channels
i was going to feed the buck convertor with dc from batteries, so sync rect is not nesecsary and it is not isolated, simple straight through positive buss with the fet and inductor in serires......... |
|
|
| luka |
Hi
maybe irf9540n, FQA17P10, IRF4905 something like those |
|
|
| jamesrnz |
ok i ordered
samples of uc2573
and some fairchild p-channels samples
thanks:hot: |
|
|
| Danko |
Just an idea: You can put the switching devide into the negative rail. Then you can use N-channel MOSFET (better performance, cheaper device), adn you can use very simple and cheap MOSFET driving circuit.
I made a DC-DC buck regulator, like that, and worked well.
But in this case the feedback is a litthe bit complicated. You may need an optocoupler in this case, in the voltage feedback path. |
|
|
| JensRasmussen |
James,
Let me try to sum up :)
What I understand you would like is:
Vin = 16-18V
Iin= 50A
Pin = 18V*50A= 900W!
Vout=6-8V
Pout=Pin*0.85=900W*0.85=765W
Ploss=Pin-Pout=900W-765W=135W!!! :hot: :hot: :hot:
Iout=Pout/Vout=765W/8V=96A!!!
You will need a huge inductor (100A+), and possibly several converters running out of phase to arrive at this.... Is this intended as a welding system power supply?
Please tell us more
\Jens |
|
|
| jamesrnz |
Vin=10-13 vdc
Vout=17vdc
Iout=50A
not welding, just need power from some batteries
but welding sounds like fun |
|
|
| JensRasmussen |
| quote: | Originally posted by jamesrnz
Vin=10-13 vdc
Vout=17vdc
Iout=50A
not welding, just need power from some batteries
but welding sounds like fun |
In that case you will need a boost type regulator as a mimimum and active retification will be more difficult.
In you first post you talked about 6-8V output voltage from 12-18V in.....
\Jens |
|
|
|
