Ideas for SMPS 14v 150amp in a design

[j_caracas]

Good night,
I need to mount a source that provides 14v and 150A, that has to be switched and high frequency. I was thinking in a forward or flyback. the primary source is from 180 to 220AC, then have to be with pfc. Any suggestions? I already did some simulations using spice, forward with two keys, but no pfc, and peak current of the rectifier is around 60A.

 

[svhb]

Re: smps 14v 150amp.

what rectifier do you mean? The output of the smps, or at the input of the mains.

For the input:

14V*150A = 2100W, efficiency estimate dat ~90% --> input power = 2100/0.9 = ~2330W.

Max input current = 2330/180 = 13A rms. (sinus regime, with PFC).

But when you use a rectifier and filter capacitor, the capacitor loads in peaks, and it is not so easy to predict the peak current.

For the totpology, I think a forward will be a better choice than a flyback for this range of power. Personally, I should use a bridge topology for this kind of power.

There is an interesting Unitrode (now TI) design note on the current-doubler rectifier for the output stage. (https://focus.ti.com/lit/an/slua121/slua121.pdf)

Stefaan

 

[j_caracas]

Re: smps 14v 150amp.

I was speaking of the input rectifier, the forward without PFC peaks reached 60A with a RMS current of 30A. is there a controller CI for the bridge topology? control of the bridge would not be complicated?

 

[j_caracas]

Re: smps 14v 150amp.

Which would be the advantages of using the converter bridge? I think that would complicate the control circuit, but not greatly improve the losses.

 

[svhb]

Re: smps 14v 150amp.

With a bridge design, you can use the transformer better:

- for a single switch forward converter, you have only energy transfer when the switch is on, the off time is used to reset the transformer. You can simply not go for a 100% duty cycle.

- for a bridge design, at maximum duty cycle, there is a continuous energy transfer to the output (correction, nearly continuous, because there is a certain dead-time needed for switching the FETs).


On this link you can see an overview of power-levels and the topology that can be used: https://en.wikipedia.org/wiki/Switched-mode_power_supply

But you are right, a full bridge smps has a higher complexity mainly due to the high side switches.

Some components from TI:
**broken link removed**

 

[powerpratik]

smps 14v 150amp.

Flyback topology is only for 10A capacity.


And for these rating, i preffered full bridge topology....

 

[j_caracas]

Re: smps 14v 150amp.

Ok, I'll change my project for a full bridge ... but I confess that with this topology i'm lost, so if someone can help me. How to control the full bridge PFC? there is some specific CI with which you have worked?

I will do the transformer design using ORCAD, for a switching frequency of 50Khz. so I've done, I post here for comments.

is better to use a boost for the PFC? or to integrate the PFC ?

Do you have any scheme of a practical power converter with full bridge?

Thanks

 

[svhb]

Re: smps 14v 150amp.

There are some other questions that can be asked:
- what stability do you need at the output?
- is this a one-off project, or do you need to produce a lot?
- is the size and weight of the converter importand?
- is the efficiency the biggest concern?
- is PFC an absolute requirement

I ask this, because depending on the application, maybe a simple mains transformer and rectifier can be used?
Eventually, you can set the output of the transformer/rectifier to a bit above 14V at the minimum AC input voltage, and add a simple unisolated buck converter to stabilize the voltage on 14V.

I used this strategy a few years ago to replace a full switching application to a much more robust "mains-transformer + post regulator" design.

 

[Tahmid]

smps 14v 150amp.

Hi j_caracas,
Some suggestions:
1. At such high power I'd go with full-bridge. Safe limits for each topology is:
Half-bridge - 1000W
Full-bridge - No limit
Forward - 500W
Flyback - 200W
Push-pull - 1000W
2. Use an input line filter for two purposes:
a) Filter
b) Reduce inrush current
3. Use proper rectifier and bulk capacitor for a clean DC supply. You can use GBPC3510 or GBPC08 bridge rectifier. And use about 100uF input cpacitor.
4. Use high/side low side drivers for the bridge MOSFETs. eg. IR2110 / IR2113 / IR2181 / L6388 / IR2112 / L6385 / L6387 and many more.
5. Use proper ferrite cores and proper windings.

 

[j_caracas]

Re: smps 14v 150amp.

Hi, thank you for the suggestions and all of them:

1 - This converter will be connected to a battery bank, then the question of stability is simple.

2 - The idea is to series production, then the low cost is very important.

3 - Yes the size and weight matter! this is why I do not want to use a single 60hz transformer and a rectifier

4 - the efficiency does not worry me much, but I can not dissipate much heat, because this equipment must operate at maximum at about 30 degrees above the ambient temperature. Someone here has an idea of thermal resistance of an aluminum case used in these drives (eg Xantrex)

5 - PFC is only one way to improve the current peaks in the order of 60A that I got when I made a simple rectification diodes, such peak would disarm the circuit.

Does anyone have a schematic of a source fullbridge with similar characteristics?

Why the forward is limited to 500w? I can not see this limitation, I did the project in ORCAD magnetic part and got the following features:

I build a transformer , but with a different nucleus, which also is designed by ORCAD, (I don't have the characteristics of this project right now) but it looked like this:

I know that it is very "full", the diameter of the nucleus is to small, I'll do some testing next week using a forward with two keys. and hope this nucleus does not saturate at least the project says it will not.

 

[svhb]

Re: smps 14v 150amp.

There is actually no theroretical limit of how much power you can get with any topology, most of the limitations are practical reasons.

Yes you can make a forward for this application, and yes it will probably work. But for a lot of reasons it becomes more economical / smaller / more efficient / more cost-effective, .... with a bridge topology.

BTW, for looking at saturation for your core you have to use the maximum input voltage for calculations.

Happy 2010

 

[Tahmid]

smps 14v 150amp.

Hi,
Like shvb said, the reasons are more practical, and whatever he just stated is correct.
For power electronics, I would not trust simulation, as many factors are not considered in simulation such as core winding technique(transformer and inductor), capacitor ESR (VERY IMPORTANT AT THIS FREQUENCY AND POWER), wire resistance, PCB trace layout (may lead to parasitic capacitance), and many others. This is SMPS we're talking about, that too in the order of Kilowatts. I had to work with SMPS for more than a year before being confident about SMPS construction.

 

[j_caracas]

Re: smps 14v 150amp.

I'm have a question doubt, any topology that I do will get a inductor in the output, for example how would be one 10uH inductor for a current of 150A? whats the best way to do it?
Air inductor ?

Thanks

 

[svhb]

Re: smps 14v 150amp.

If you can do transformers, you should be able to do inductors as well.

You definitely don't want an air inductor. There is nothing special about 150Amps, it just gets a bit bigger .
With a chosen core, you calculate the maximum number of turns to not saturate the core. If the induction in this configuration is too small, go to the next bigger core, and repeat.
You can also put solenoids in series.

Warning: for saturation of the inductor, use the maximum value of the triangular shaped current.

 

[FvM]

smps 14v 150amp.

Viewing at the transformer photo, I wonder what's the used core type. It looks like an ungapped ferrite or tapewound
core. In this case, it won't be suited for flyback converter. Or is it a powder core?

As another point, did you consider proximity effects? Some twisting of the wire bundles is required to reduce it.

How about primary to secondary safe insulation?

 

[svhb]

Re: smps 14v 150amp.

FvM, in the download below the picture of the transformer, it looks like it is "designed" for a forward converter. If my understanding of "component type : Forward-transformer double" is right. I'm not sure what "double" means however...

But you are correct, it should be better to use some proper made litz wire for this power level. And the safety isolation can be done better.

Stefaan