12v 250w Car SMPS based off SG3525

wally7856

New member
“Looking into this datasheet of core material N87 from EPCOS, can you tell what is the Bmax? I can't figure it out.”

100khz is a lucky number. There is one dedicated chart just for that. Otherwise you would have to use the general purpose one.

I put a pic of both charts here with red markers. The charts have lines for 25 deg C and 100 deg C. Generally you run a transformer at 60 to 70Deg C, so i tried to place the red line at that point. It looks to me that is around 1100 to 1200G.

You should get the same gauss from reading either chart. The dedicated 100khz chart is much easer to read so take a few different points there and try to read them off of the general chart for practice.

“Regarding the EI type: In the datasheet of suppliers, for example, ferroxcube, the specification in those datasheet are only for one E half, right?”

Yes.

Now, the use of your method wally, can be applied to the I type? That part that goes on top of the E core? Where the I part enters in the calculations?”

Nothing changes in the calculation. The center leg is the same as it always was, length x width = CM squared. And window area, is window area, length x width = CM squared. Window area is the open area in the core. With a EE core the window area goes across two E cores. With a E and I core the window area is only as deep as one E core.
 

Attachments

  • 100khz curve only.jpg
    100khz curve only.jpg
    15.3 KB · Views: 29
  • 100khz curve Capture.jpg
    100khz curve Capture.jpg
    17.5 KB · Views: 26

wally7856

New member
“A: 33,07mm;
B: 23,75mm;
C: Can't take because there is only one halve;
D: 12,66mm;
E: 9,60mm,
G: Can't, for the same reason;”


D and E are either mixed up or wrong.
C is the thickness, we need that number.
G is the gap, if it has one. It may not. Put a straight edge across all 3 legs and look for a gap.
 

AMSA

New member
Hi Wally.

I have done like you have suggested: Put a straight edge across all 3 legs and look for a gap. and I didn't found any gap. So the core does not have a gap.

About the measures, again, I have done new measurements and the results are here:

http://imageshack.us/photo/my-images/526/measurementsy.jpg/

Nothing changes in the calculation. The center leg is the same as it always was, length x width = CM squared. And window area, is window area, length x width = CM squared. Window area is the open area in the core. With a EE core the window area goes across two E cores. With a E and I core the window area is only as deep as one E core.

Ok, I see. But the power that the EI core can handle is the same as if it is EE since the center leg is the one that determines the power that the core can handle right? The only difference will be the Ap value: Ap = Wa x Ae. What influence this will cause? Having a small Ap value? BTW, what is the purpose of Ap value?

About the Bmax, the 1100G to 1200G is for the 1st graph right? Comparing to the 2nd one, has you told me You should get the same gauss from reading either chart, seems to me that if we try to put the line in the 70ºC we can get that 1100/1200 Gauss.

“Regarding the EI type: In the datasheet of suppliers, for example, ferroxcube, the specification in those datasheet are only for one E half, right?”

Yes.


So that means that the length of the center leg doesn't matter for the question of the power he can handle. Because, we look at the cross-section area of the core and not to the length. Besides that, it's the cross-section that we use when we go to the pressman book. So the length, for power handle, doesn't matter, right?

Regards and thnak you very much, again, Wally, for you patient and availability to answer those question.

By the way, I found a vendor in ebay that ships to my country but the thing is that he only has N67 material type. Considering what I have read, that core material is something old, I mean, it was substituted by the N87. What can you tell me about that N67 material? Is suitable for the most applications? It can operate from 20kHz to 500kHz?
 
Last edited:

wally7856

New member
“Ok, I see. But the power that the EI core can handle is the same as if it is EE since the center leg is the one that determines the power that the core can handle right?”

NO, The center leg is the Ae value, used in the primary turns formula.

“The only difference will be the Ap value: Ap = Wa x Ae. What influence this will cause? Having a small Ap value? BTW, what is the purpose of Ap value?”

The Ap is the amount of power of a core.

“About the Bmax, the 1100G to 1200G is for the 1st graph right? Comparing to the 2nd one, has you told me You should get the same gauss from reading either chart, seems to me that if we try to put the line in the 70ºC we can get that 1100/1200 Gauss.”

Yes, it looks like that to me also.

“Regarding the EI type: In the datasheet of suppliers, for example, ferroxcube, the specification in those datasheet are only for one E half, right?”

Yes.

“So that means that the length of the center leg doesn't matter for the question of the power he can handle. Because, we look at the cross-section area of the core and not to the length. Besides that, it's the cross-section that we use when we go to the pressman book. So the length, for power handle, doesn't matter, right?”

The length of the leg does matter. Power = Wa x Ae.

Regards and thnak you very much, again, Wally, for you patient and availability to answer those question.

“By the way, I found a vendor in ebay that ships to my country but the thing is that he only has N67 material type. Considering what I have read, that core material is something old, I mean, it was substituted by the N87. What can you tell me about that N67 material? Is suitable for the most applications? It can operate from 20kHz to 500kHz?”

N67, information starts on page 96.

http://www.thierry-lequeu.fr/data/SIFERRIT.pdf

Your transformer numbers.
Wa = 1.89CM x .713CM = 1.348CM squared.
Ae = .963CM x 1.266CM = 1.219CM squared.
WaAe = 1.348CM squared x 1.219CM squared = 1.643 CM cubbed, look for this in Pressmans book in the half bridge, full bridge, and i think push pull, section of charts. Pressman calls this AeAb CM cubbed. Also use the 96khz colum.

Page 312 of the book has a AeAb CM cubbed of 1.6 = 688 watts so core probably good for 700 watts.
 

AMSA

New member
Hi wally,

Your transformer numbers.
Wa = 1.89CM x .713CM = 1.348CM squared.
Ae = .963CM x 1.266CM = 1.219CM squared.
WaAe = 1.348CM squared x 1.219CM squared = 1.643 CM cubbed, look for this in Pressmans book in the half bridge, full bridge, and i think push pull, section of charts. Pressman calls this AeAb CM cubbed. Also use the 96khz colum.

Page 312 of the book has a AeAb CM cubbed of 1.6 = 688 watts so core probably good for 700 watts.


But that data was taken from table 7.2b, right? If so, is that correct? I mean, that information is for PQ core. The one that I have here with me is an E type. Doesn't matter? Besides that, accordingly to the table, that is for full-bridge and half-bridge. The topology that I want to design is the forward one. That means that we can not use the data from this table, since is for half and full-bridge, right?

So that means that the length of the center leg doesn't matter for the question of the power he can handle. Because, we look at the cross-section area of the core and not to the length. Besides that, it's the cross-section that we use when we go to the pressman book. So the length, for power handle, doesn't matter, right?”

The length of the leg does matter. Power = Wa x Ae.


You are right! That means if we use and EI type we get lower power when compared to the EE. Right?

Regards.
 

AMSA

New member
Let me take advantage to the fact that you haven't reply yet to my last post.

About the current rating spec-ed in circular mils or cm (...)

This is a guideline on how to use cm for transformer design. This gives you a cm per amp frame of reference.

500cm/A, for every amp you need 500cm. This number is for power supplies that have to put out a continues power and is considered a conservative design. For further reference 500cm = 23awg.

450cm/A, not as much copper used as above but when you can not quit fit enough wire this is acceptical most of the time.

350cm/A, for a power transformer this is considered poor quality and it will run hot.

275cm/A, considered a junk power transformer.

150cm/A, last and certainly least, is the “audio quality” transformer.

You can get buy with such a low cm/A because the average music level is so low compared to rated output.


I want to ask you one thing.

This current rating is the same as the current density J [A/m] that we see at pressman and McLyman books? Or doesn't have nothing to do?

That current rating translates the capability of the wire to carry out the current? I didn't understood in which situations we use 500cm/A, 350cm/A or 150cm/A? Which one is the best choice? It depends on the application? (For example, when you were calculating the codex transformer you used 150cm/A, why? What was the criteria?)

Regards!
 

wally7856

New member
AMSA, most of your questions i answered three times now and gave you examples. I can not make it any simpler or more complete than i already have. I do not know why you keep asking the same thing over and over or how you could possibly have trouble multiplying two numbers together. If it is a language problem then you will have to find someone who understands English better than you and explain it to you in your native language.
 

AMSA

New member
I'm sorry to bother you with those questions. Maybe I have already asked before and I don't remember. I will check again the previous posts.

Thank you once again Wally.
 

AMSA

New member
Hi there wally,

I was doing an exercise, following your method and I think that there is an error on one of your calculations:

In post #163, at the beginning, you wrote:

You want to build a 12vdc push pull smps and get around 250W with a +-35vdc output. So lets start gathering some data. First we will assume a poor 60% efficency because of no output inductors.

250W x 1.4 = 350W input power.

Assume 14.7 volts max input.

350W / 14.7vdc = 23.8A, max primary current.


I was wondering if, where you wrote 350W / 14.7vdc = 23.8A, max primary current. you meant 350W / 12vdc = 29.17A, max primary current?

I ask that because, if we want the maximum primary current, we will only get that if we have a lower voltage, right?

Can you confirm this?

If I find another error, while doing my exercise (I will post it here later in order to see if you agree with what I have done, ok?), I'll post.

Regards.
 

AMSA

New member
Hello wally.

I hope you can read this post in order to know if I am doing right this calculations and if it is on the right way accordingly to the method that you have suggested.

Like I said before, I want to design a forward converter, with one transistor only. Bellow you can see the specs:

SPECIFICATIONS:
Vin = 230Vac

Vin(nominal) = sqrt(2) * 230V = 325V
Vin(min) = 300V
Vin(max) = 350V

Vout = 30V
Iout = 5A

Pout = 30V * 5A = 150W

Considering 80% of efficiency:

Pin = 150 / 0,80 = 190W

Iin(max) = 190W / 300V = 0,634A

TRANSFORMER CALCULATIONS:
Bmax = 1600G
Ae = 1,21 cm2
f = 100000Hz

Primary turns = (350V * 100 000 000 ) / ( 4 * 100 000 * 1,21 * 1600 ) = 45T
Secondary turns = 45 * (30V + 1V) / 300V = 5T

WIRE SIZE IS NEEDED:
skin depth = 2837 / sqrt(100000) = 8,971 mil

Primary wire:

8,971mil = 0,228mm = 79,7cm = AWG31
0,643A * 150cm/A = 96,5cm

96,5cm / 79,7 = 1,21 -> Let's make it 2 strands

Secondary wire:

2 * 8,971mil = 17,973mil = 0,4558mm = 320,4cm = AWG25
5A * 150cm/A = 750cm

750/320 = 2,34 -> Let0s make it 3 strands

Wally, could you take a look to my calculations and procedure in order to see if it is ok?

Kind regards.
 

demykiko

Member
hi codex..i already did that project of rod eliot
and it really worked 4 yrs ago..i used IRF540
because its cheap and low Rds on...
next time i will give suggestion, im just busy right now

regards!

demy
 

codex653

New member
wow! I didn't realize how many comments that I had gotten on this thread while I have been gone! I'm going to try and get to working on the calculations for this transformer over Christmas break. I SHOULD have time because I will have 3 weeks of no classes, but I may have a job. I don't know yet. I really want to finish this project and get it working reasonably well. I'll be sure to make some more posts later after finals week
 

Redwire

New member
Alternate mosfets

Yes IRF3205 will work, a stronger mosfet. I have seen a pair used at 375W, so no need to run four. I would lower the gate drive resistor to 10 ohms. I think the IRF3206 looks good too.

IRLB3036: 60V, 1.9mohm, 195A, 380W, Ciss=11,210pF (logic-level) $4.68
IRFB3206: 60V, 2.4mohm, 120A, 300W, Ciss=6,450pF (looks good!) $4.74
IRF3205: 55V, 8.0mohm, 110A, 200W, Ciss=3,257pF (end of life) $3.14
MTP75N06: 50V, 9.5mohm, 75A, 150W, Ciss=2,600pF (obsolete)
 

Sierra

Valve Lover!
Very good friend !:angel:

Are you going to share with the PCB and the schematics? this looks very nice DIY project
 
Top