Calculation programs for transformers and inductors

aadf

New member
Hi,

I have a problem, the results are not showing well. see attach:

Sin título.jpg

What could be the problem??


Thanks
Andres
 

Silvio

Well-known member
Hi,

I have a problem, the results are not showing well. see attach:

View attachment 6980

What could be the problem??


Thanks
Andres

Hi Andres I do not know what is exactly your problem but it could be that the Excellent IT program got corrupted in some way. If your problem persists download it again and delete your previous version. However I calculated it for you with mine.

In the first calculation I used a switching frequency of 50Khz and an output voltage at 50v. The answer to this output voltage is not real if your smps is for audio purpose. remember that the smps is not regulated and the voltage will rise to the peak value. This voltage will rise to around 60v on the high input voltage, however you can check the peak by moving the mouse pointer at the output voltage where you will have 3 for a non regulated SMPS. I marked these with 3 red arrows. The output inductor if you are going to fit one is not really needed as the pulse width with the IR2153 runs at 50% duty. You can put a smaller one say around 5uH or so but this will create voltage drop. The higher the inductance the higher the voltage drop at load. The inductor will also limit a little the inrush current at start up charging the secondary capacitors. One other thing at a current density of 5A/mm² in the first calculation the winding fill factor of 0.3 is exceeded and the windings may not fit in the transformer bobbin, you can use 6A/mm² instead.

I suggest using a higher frequency say 60-68Khz. This will still use a copper wire dia of 0.5mm. Having a higher frequency the trafo will need less turns and gives a little more power. I made a calculation for you in the second screen shot at 60Khz and a lower voltage of around 40v, when you move the mouse pointer on the output voltages you will notice that the peak voltage is around 52 at the rated input voltage. Usually this voltage will be maintained during amplifier operation as music does not draw continuous power.

trfo calc.JPGtrafo calc 2.JPG
 

aadf

New member
Hi Silvio,

In my work computer, the IT(7300) have that problem. In another computer doesn't even start. It must be some incompatibility of regional configuration or something like that. Now i'm using the 4000 version, but it shows one less wire in the primary:

IT4000.jpg

Its my first smps with IR2153, based on NS450. And yes, is for audio. The inductor at the output is only 2.5uHy, but like you said, is not really needed.

I made a transformer with 36turns in the primary and 12+12 in the secondary (1/2 primary + Secondary + 1/2 primary). The output is 51vcc.

My final idea is to use a ER 44/22/15 for more power. What I don't know is how to calculate or obtain the correct frequency, since these cores are recycled and I don't know what material they are made of.

Thanks so much for your support!
Andres.
 

Silvio

Well-known member
Hi Andres, I also use recycled cores when I find them. I usually take the trouble and try to figure out the operating frequency they were working with by observing the type of chip and its timing resistor and capacitor. Observing the number of turns in the primary and trying to make out what flux density was used comparing with Excellent IT. At least I would have an idea more or less what type of material the core has.

As you may have already noticed that different core material take different number of turns per volt. In any case while you are experimenting you can try out what suits best . Keep in mind that ferrite cores tend to get a little hot in operation and do not be afraid of this. The temperature rise is normally calculated at 25 degrees above ambient so reaching 50 and 60 degrees is no big deal. This feels hot to touch but do not be mislead the temperature of the core must not exceed the 100 degree mark. Here is why I always say that adequate tapes must be used (Mylar tape) as other types tend to get hot and soften up or melt at times.

If the core material is unknown then use N27 as core material this will give you more turns for a given voltage. If the turns are more than necessary then the core will loose power, getting it right will get you the best result.

A good calculation you can adopt for half bridge, full bridge and push pull typology is :-

Nr. turns primary =155* 10^8 / 4 X freq in Hz X flux density in Guass (1200-2000) X core area in cm^2

The flux density will vary according to the frequency used. To give you a rough idea at 20Khz you can use 2000 guass for 50-60 Khz you can use 1500-1600 Guass and for 100Khz you can use 1100-1200 guass. You will see that the higher the frequency the lower the flux density used. This is because the higher the operating frequency the hotter the core will get due to losses.

Lastly with an ER44 transformer you will get around 500w and 700w peak at a medium frequency of 60 to 70Khz.

I suggest you read some blog posts that I posted on this website so that you learn more about transformer winding practices. you can also take a look at the video on youtube of me winding a small transformer for smps.

Link https://www.youtube.com/watch?v=_K3ixhcTYFg&t=1s

Regards, Silvio
 

Raza

New member
Hello sir starichok51
I downloaded your softwareT (7100) and tried to run it the first time but does not open in my computer. Can you pleae help me to solve the problem? Thanks
 

Raza

New member
Hello everyone!
I present to you my program for the calculation of transformers and inductors.
One of them ( ExcellentIT(6100) ) Dimonis you presented in your theme SMPS transformer design tool.
The program ExcellentIT(7100) was corrected English translation and added standard AWG wire and the added choice of the desired diameter wire.
selector is not signed due to lack of space, but it is a tool tip.
In other programs, this selector signed "Use the desired diameter of the wire"

I speak English poorly. I use Google translator.
I would be difficult to answer your questions, but I will try.

I'm not working, I'm retired. So I will be grateful for your financial support for my work on the development of these programs .

Hello sir starichok51
I downloaded your softwareT (7100) and tried to run it the first time but does not open in my computer. Can you pleae help me to solve the problem? Thanks
 

hesamavr

New member
Hello sir starichok51
I downloaded your softwareT (7100) and tried to run it the first time but does not open in my computer. Can you pleae help me to solve the problem? Thanks

I have the same problem, downloaded 7100, but it doe not run on my PC. Even I tried to run by compatibility mode, no success. W10-x64
 

Silvio

Well-known member
open the file and double click on the colored icon

Excellent IT icon.png

I am running windows 10 64bit and the version 7300 works fine

regards Silvio
 
Hi, I tried Forward and I can't input any core dimensions. When I click on "adding to database (input dimensions) nothing happens.
 

liviu

New member
Used one of your programs for calculate a flyback trafo
In fact Flyback ZCS 2000.
OK, for EI33 3C90 result 80 turns 0.4mm

Lp=1.47mH

1.2 choosed gap.

In fact i tried to make trafo and after finish L=20mH!!!
E+I from an ATC source (recovered)
Tried use gap and 0.65 mm spacers (trat mean 2x0.65=1.3mm total gap) for decrease Lp and Lp=1,2mH but with Llose (short-circuit on secondary coils) =80uH..so big.
I made a mistake?
 

Attachments

  • _SMPS Invertor EI 33,0x24,0x12,7x9,7 3C90 Ferroxcube 1,2mm, 38kHz, 250V, (110V), 24Vx1,5A, 17V...zip
    618 bytes · Views: 46

liviu

New member
Yes, is OK. UniT UT58 series and work well. Measured known inductances and my instrument work well

I tried make another trafo on ER35/16/11 !!! (Yes, atipic dimensions) (from ATX too) and without gap 9mH.
maybe another material? Not 3C90? Remember, China cheap ATX.
 

Silvio

Well-known member
Yes, is OK. UniT UT58 series and work well. Measured known inductances and my instrument work well

I tried make another trafo on ER35/16/11 !!! (Yes, atipic dimensions) (from ATX too) and without gap 9mH.
maybe another material? Not 3C90? Remember, China cheap ATX.
Well I got curious about your inductance readings, I measured an EI 33 from an old ATX and for 155v at about 20 khz because that used to be the average frequency they used to work with. I got and inductance of around 5.6 mH for the primary winding. These ATX having a HB topology do not have any gap in the core. These usually have around 30 to 40 turns in the primary if I remember correctly. This should give you an idea of what inductance readings you should expect. The number of turns will vary with frequency, however in flyback the pulse width duty cycle, frequency , input voltage and gap size will all alter the number of turns, hence the final inductance of the primary winding.
About the program itself ( Excellent IT ) I never used it for flyback topology though But only for HB and FB which gave me very good results.
 

liviu

New member
Thank you for your answer.
In post #133 i wrote "FlybackZVC" used.
L is depending with N^2, right?
So, if you for 40 turns =>5.6mH will have:

100/40=2.5 times my turns versus ATX turns.
2.5*2.5=6.25 times my L big like ATX L
My L=5.6 * 6.25 = 35mH (my L is only 25 mH, so i not make mistakes, only must very large gap). I reffer here at EI33.

For ER35/16/11 not measured primary winding before.
 

Silvio

Well-known member
In a flyback topology firstly you are dealing with a wide voltage range between 85 and 250vac. That makes the calculation more complicated due to the fact that the secondary output still have to be maintained at the minimum input voltage. For this reason the pulse will vary between 10% and may be as high as 85-90%. If the load is not constant there are times when the flyback is running with no load and only on the minimum load resistor and near the maximum input voltage the duty cycle will be at its minimum not to say it may run on DCM (discontinuous conduction mode)
Yet again in the design frequency of operation and the maximum design pulse width all play an important part on the primary inductance.
Just to make it clear for you it is not so easy to design a flyback smps without good knowledge. If the inductance is too large the power transfer to the secondary will be poor while if it is less than needed then there is a chance of saturation of the transformer which will eventually blow out the switching transistor.

To give you an example I dismantled an EE20 transformer this having only around 125 turns in the primary considering the small center core being around 5 or 6mm square. The frequency of operation was 100khz and the gap was very small around 0.15mm, the primary inductance was 2.4mH. This was a small auxiliary smps which I made giving about 10w of power.
To make it clear for you I did not design the transformer turns but dismantled and counted the turns. I altered the secondary winding a little so that instead of having 12v at the output I needed 17v and thus I added 2 turns to the secondary to achieve that. I came to this figure by simple proportion. However the primary winding still had the same number of turns.

Please note that in flyback topology a core will give around 0.3 the power it gives in HB and FB topology. The reason for this is because only one quadrant of the magnetic flux is used while with other topologies the four quadrants are used. So to make more clear an EI33 will give around 300 watts at 20 Khz in HB while in flyback it will only give around 80w to 100w.
If you want to learn how to calculate a flyback then you can follow Robert Balanos on youtube in the video Design of a flyback.

Regards Silvio
 

liviu

New member
In Flyback 6100 for RCD clamper having two resistor values:
For "slow" diode (FR207) and for fast diode.
FR207 having trr=500ns

UF4007 or BYV26 E having 75 ns.

Where is the limit between "slow" and "fast" here?

B.e. having 10nF and 23.6 kOhm / 1.2W for slow diode and 10nF and 5.71 kOhm / 4.55W

If use fast diode (BYV26E) i can't use 23K? Why i want use 23(24K or 27K)/2W - because is smaller like 5,6K /5W


Thank you
 
Last edited:
Top