SMPS Transformer Primary Turns Calculations

L

lorylaci

New member
Microsim, you are saying stranded wire wrong!
Wire with 1mm is equal to 4 strand of 0,5mm diameter, 25 of 0,2mm diameter etc.. because area is what matters.
 
MicrosiM

MicrosiM

Administrator
Staff member
Solid Wires

Yes!

Solid wire…

Is typically used for transformers operating at 100 kHz and less, and RMS current levels under
20 amps. The main advantage of solid wire is low cost, both in material cost, and time to wind. The main disadvantage
is losses. Solid wire can generate excessive AC winding resistance at SMPS frequencies.


Hope that helps! ;)
 
N

_nlc_

New member
Microsim, I used your formula, and I also think the freq must be in Hz and not in Khz !? Because I obtain too much turn for the primary :)
Vmin = 42VDC (I will be in push pull topology thus the result will be for 1/2 primary)
Freq = 70KHz
Bmax = 1200 Gauss (N87 material)
Ae = 3.68cm² (ETD59 core)
Thus we have 42*100000000/(4*70*1200*3.68) = 3396 turns.

With the frequency in Hertz the problem is solved : 42*100000000/(4*70000*1200*3.68) = 3.396 turn.

Regards
 
MicrosiM

MicrosiM

Administrator
Staff member
_nlc_ said:
Microsim, I used your formula, and I also think the freq must be in Hz and not in Khz !? Because I obtain too much turn for the primary :)
Vmin = 42VDC (I will be in push pull topology thus the result will be for 1/2 primary)
Freq = 70KHz
Bmax = 1200 Gauss (N87 material)
Ae = 3.68cm² (ETD59 core)
Thus we have 42*100000000/(4*70*1200*3.68) = 3396 turns.

With the frequency in Hertz the problem is solved : 42*100000000/(4*70000*1200*3.68) = 3.396 turn.

Regards
Click to expand...


It is in Hz, that is true!
 
N

_nlc_

New member
Ok so I calculated my transformer, need your opinion !
My need is a 48VDC -> 240VDC 3000W converter, but I want to use multiple transformers, primary in parallel and secondary in serie. I will try to obtain 3000W with 2 ETD59 transformer, 3C90 material, and full bridge topology.

With the formula Np = Vp x 10^8 / (4 B A f), I use Vp = 54V (max input voltage), B=1200G, A=3.68cm² (etd59), and f=100Khz.
I obtain Np = 54x10^8/(4x1200x3.68x100000) = 3.05 turns.
If I take 3 turns, Bmax become 54*10^8/(4x3x3.68x100000) = 1223G, I think it's good too.
And the input voltage will be at 54V for a very short period of time, the battery will go the the 48V nominal voltage very fast, thus Bmax will be 48*10^8/(4x3x3.68x100000) = 1087G.

For the secondary turn, with the 2 transformers secondary in series, I need 240+10% / 2 = 132V at the output of one transformer, thus with the minimal input value I obtain 132/42 = 3.14 ratio. Thus 3 primary turn x 3.14 = 9.42 turns = 10 turns for the secondary.

If I verify, 42V at the input (minimum input value) divided by 3 primary turns = 14, multiplied by 10 secondary turn = 140V, multiplied by 2 transformers = 280V. Thus normally it's OK, I will be able to regulate 240V with the duty cycle.

Now I need to see what I need for primary and secondary wire. The primary RMS current max for one transformer will be full power of this transformer with minimal input voltage, thus (3000W/2) / 42V = 36A on the primary (thus ~40A), and ~13A at the secondary.

For primary and secondary I plan to use copper foil, and wind the 3 turns primary first to have the minimum length of copper because of the 40A, and the 10 turns secondary on the primary.
But the question is, where to buy copper foil, I don't find it anywhere :(

Regards,
Cyril
 
MicrosiM

MicrosiM

Administrator
Staff member
!

_nlc_ said:
Ok so I calculated my transformer, need your opinion !
My need is a 48VDC -> 240VDC 3000W converter, but I want to use multiple transformers, primary in parallel and secondary in serie. I will try to obtain 3000W with 2 ETD59 transformer, 3C90 material, and full bridge topology.

With the formula Np = Vp x 10^8 / (4 B A f), I use Vp = 54V (max input voltage), B=1200G, A=3.68cm² (etd59), and f=100Khz.
I obtain Np = 54x10^8/(4x1200x3.68x100000) = 3.05 turns.
If I take 3 turns, Bmax become 54*10^8/(4x3x3.68x100000) = 1223G, I think it's good too.
And the input voltage will be at 54V for a very short period of time, the battery will go the the 48V nominal voltage very fast, thus Bmax will be 48*10^8/(4x3x3.68x100000) = 1087G.

For the secondary turn, with the 2 transformers secondary in series, I need 240+10% / 2 = 132V at the output of one transformer, thus with the minimal input value I obtain 132/42 = 3.14 ratio. Thus 3 primary turn x 3.14 = 9.42 turns = 10 turns for the secondary.

If I verify, 42V at the input (minimum input value) divided by 3 primary turns = 14, multiplied by 10 secondary turn = 140V, multiplied by 2 transformers = 280V. Thus normally it's OK, I will be able to regulate 240V with the duty cycle.

Now I need to see what I need for primary and secondary wire. The primary RMS current max for one transformer will be full power of this transformer with minimal input voltage, thus (3000W/2) / 42V = 36A on the primary (thus ~40A), and ~13A at the secondary.

For primary and secondary I plan to use copper foil, and wind the 3 turns primary first to have the minimum length of copper because of the 40A, and the 10 turns secondary on the primary.
But the question is, where to buy copper foil, I don't find it anywhere :(

Regards,
Cyril
Click to expand...


Thats good choice.

Copper foil might be a problem to get, depends on suppliers. you can look at RS electronics, or if you can make some samples of that transformer in china, and you only pay shipping!
 
N

_nlc_

New member
Thank you for your answer Microsim !
I look for the copper foil at RS or Farnell.
Just a question, with pull pull topology, the frequency transformer is the same frequency as the mosfets/controller, right ?
Regards
 
MicrosiM

MicrosiM

Administrator
Staff member
_nlc_ said:
Thank you for your answer Microsim !
I look for the copper foil at RS or Farnell.
Just a question, with pull pull topology, the frequency transformer is the same frequency as the mosfets/controller, right ?
Regards
Click to expand...


I am afraid that I dont understand.

Controller internal frequency could be 100KHz, but output of each will be 50KHz, SG3525 example
 
N

_nlc_

New member
Okay, so to have a 100Khz seen by the transformer (I used this frequency for primary turns calculation), I need to have 200Khz at the controller.

Edit : In my previous I made a mistake, I was talking about full bridge topology
 
MicrosiM

MicrosiM

Administrator
Staff member
usually

In formula, you will put the frequency at the output of switching devices, not the internal frequency of the main oscillator.
 
N

_nlc_

New member
Yes I understand, because in fact which is important it's the frequency transformer, same as switching device.
 
TO-247

TO-247

New member
Hi there

Microsim ,
Thanks for transformer calculation methods.
I'm going to build a smps for giving +/-40V from 220V AC Line

Result based on your calculation Formula :

http://www.img9.org/uploads/26406a2131fd51958d44b0721f7b2b52.jpg

therefore , 16 Turns for pri. and 5+5 Turns for sec , but my question is what size of wire diameter ? (80v-8.75A Output it's means 700w rms )

can you help me for calculating wires diameter ?
"sorry for my English"
 
MicrosiM

MicrosiM

Administrator
Staff member
FASTER_2011 said:
Hi there

Microsim ,
Thanks for transformer calculation methods.
I'm going to build a smps for giving +/-40V from 220V AC Line

Result based on your calculation Formula :



therefore , 16 Turns for pri. and 5+5 Turns for sec , but my question is what size of wire diameter ? (80v-8.75A Output it's means 700w rms )

can you help me for calculating wires diameter ?
"sorry for my English"
Click to expand...

For best results, at 700W RMS, I would recommend a 1.2MM Litz wire, check this project for more information's!


http://www.diysmps.com/forums/showthread.php?79-1kW-smps-project-(based-on-MicrosiM-design)


It looks to me you are using ETD59 @ 50KHz??? why?

ETD49 @ 80KHZ will give you 1200W nicely
 
TO-247

TO-247

New member
They use ETD59 Due low switching frequency. You can use ETD59 or ETD49

it depends on your application. I will provide you with any help u need
Click to expand...

my application is an audio amplifier with +/-40V supply and (400w output power X 1.5) = 600w max power consumption at 7.5A per rail.

i think etd49 OK(after your help) , but i don't know suitable switching frequency :(

first , I'm going to buy an etd49 core , because i don't know its material before buying it.

it's my amplifier based on Nmos200 :



http://sites.google.com/site/quasisdiyaudiosite/nmos-series/nmos200-1

Thanks MicrosiM for helping me :)
 
MicrosiM

MicrosiM

Administrator
Staff member
That looks cool!

have you checked the thread I told you about? it shows the complete SMPS.

You will need ETD49 with N27 Core material

All help is here
 
TO-247

TO-247

New member
Hi MicrosiM , Sorry for delay !

Tel us more about the sound of that mosfet amplifier please
Click to expand...

You can hear it's sound here :

http://www.youtube.com/watch?v=CRdn6Qz410Q

but it's not mine , you can find about it here :

http://maxhobby.altervista.org/quasi/quasi.html

i can't tell more about quality of nmos200 because it's my first P.A. and i'm not listening another diy amplifier for comparison the sound , quality , bass , etc.
but for me , it's nice and i love it.

you can find more powerful amps in quasi's site.(Nmos350 , Nmos500 ,..)

you can also try Apex Amplifier with good quality and high power :

http://www.diyaudio.com/forums/solid-state/162408-900w-h-class-pa-amp-limiter.html

http://www.diyaudio.com/forums/solid-state/164208-500w-pa-amplifier-limiter.html

http://www.diyaudio.com/forums/solid-state/164093-100w-ultimate-fidelity-amplifier.html

http://www.diyaudio.com/forums/solid-state/164630-50w-mosfet-amplifier-irf540-irf9540.html

http://www.diyaudio.com/forums/solid-state/162043-mosfet-amplifier-irfp240-irfp9240.html

________________________________________

after few days , i bought the ETD49.but it's not N27 material , it's like N87 (TRIDELTA MF102) :(

please download the datasheet from the link below :

http://www.tridelta-weichferrite.de/images/49-25-16.pdf

so the etd49 with MF102 material is here :) :



but at core manufacture's site , if you click on open material comparison chart , you can see the chart :



i found at the table the TRIDELTA MF102 = EPCOS N87

You will need ETD49 with N27 Core material
Click to expand...

so what can i do ? is the N87 suitable for my project (1200 Gausses)?
 
You must log in or register to reply here.
Top