Well what relation do you use to calculate max power possbile some ETD core (44 49 59 ) can give ?
What relation do you use to calculate primary turns ?
What about duty cycle ?
If we use that :
Lets say that 0,2 Tesla is maximum flux change at 100 khz that ETD ferrite core can take..
than we get Nmin ...
Lets presume that WALL AC voltage is min 210 V and max 250V
So for V max we use 250*sqrt(2)/2 for push pull HALF BRIDGE thats 175 V
So N1 min is calculated according upper relation !! using max voltage
MY question is secondary !!!!
secondary turn determines duty cycle !!!
if secondary is the closest to desired voltage than duty cycle is near 100% and maximum energy is got from FERRITE!!!
BUT we must make it litter lower (if voltage drops below 210V and if we calculate CORE LOSS and voltage drop on the transfomer itself: more energy transfered more loss and more voltage drop)
So what N2 turns should I use and what the worst duty cycle should I take into calculation
I hope my understanding is correct ?
What relation do you use to calculate primary turns ?
What about duty cycle ?
An externally hosted image should be here but it was not working when we last tested it.
If we use that :
Lets say that 0,2 Tesla is maximum flux change at 100 khz that ETD ferrite core can take..
than we get Nmin ...
Lets presume that WALL AC voltage is min 210 V and max 250V
So for V max we use 250*sqrt(2)/2 for push pull HALF BRIDGE thats 175 V
So N1 min is calculated according upper relation !! using max voltage
MY question is secondary !!!!
secondary turn determines duty cycle !!!
if secondary is the closest to desired voltage than duty cycle is near 100% and maximum energy is got from FERRITE!!!
BUT we must make it litter lower (if voltage drops below 210V and if we calculate CORE LOSS and voltage drop on the transfomer itself: more energy transfered more loss and more voltage drop)
So what N2 turns should I use and what the worst duty cycle should I take into calculation
I hope my understanding is correct ?
coreset output power
The factors for ouput power is core volume and frequency, The higher the frequency the more power delivered by the coreset(refer to pages 283-284 Switching Power Supply Design by Pressman) I would read the whole chapter its a good read.
If you use the simulation file set along with the software mentioned you can prove it yourself. If you don't want to do that I would refer you to the URL below and follow the tutorials for the SMPS.
chas1
http://www.ecircuitcenter.com/
The factors for ouput power is core volume and frequency, The higher the frequency the more power delivered by the coreset(refer to pages 283-284 Switching Power Supply Design by Pressman) I would read the whole chapter its a good read.
If you use the simulation file set along with the software mentioned you can prove it yourself. If you don't want to do that I would refer you to the URL below and follow the tutorials for the SMPS.
chas1
http://www.ecircuitcenter.com/
calculating the volume for the transformer plus number of turns
To answer your question, you will find many methods for calculating turns and core size. I will try to explain a easy one for me. You must remeber all methods are based on Faraday's law and just rearranged so you can find the unknowns.
Faraday's law:
E= N*AE *(dt/dt) * E-8
E is the voltage across the transformer winding, N is the number of turns on the winding, Ae is area of core in cm^2, DB is flux change (gauss) and dt time for flux to change (seconds)
Now we rearrange the terms and plug in some definitions and we can get the following for finding number of primary turns.This is for halfbridge. For Forward or Fullbridge you double this number because they have to support the rectified mains not just half.
V(in) * E8 / (4 * Fs * Bmax *Ae) = NT(pri)
Now for the core volume to support the power we need:
Lets set up a design for 500watts:
The primary current will be equal to :
Ip= 3 *P(out) /V(in)
V(in) will be calculated based on 110VAC line at it highest value:
130 VAC therefore with a voltage doubler we can expect about:
2(130 * 1.4) = 370 VDC after rectifiers
therefore Ip = 1500 /370 = about 5 Amps from this we calculate the input power which is 1500/ .85 for 85% eff which is about 1750 watts.
core size = .68 * P(out) * current density / Fs * Bmax
Choose a industry standard of 500 c.m. /A for current density and we get:
AcAe = .68 * 1500 * 500E3 / 100E3(Fs) * 1200 (bmax) =4.25 cm^4
If we check the ETD49 specs we find that its AcAe is about 5.7 cm^4 which is a good choice for this design
Look what happens when we raise the Fs to 200KHz:
AcAe = .68 * 1500 *500e3 / 200e3(Fs) *1200(bmax) =2.125 cm^4
so you can see the higher the frequency the smaller the core within reason (copper loss, core temp and other things come into play)
In this case we used a Bmax of 1200 and could have used up to 1600, recalucate using 1600 and then we get 3. 18 cm^4 which means we might be able to use an ETD44 coreset now for the turns calculation:
If we use the highest voltage we expect on our transformer windings when the supply turnon:
N(pri) = 370(V(inmax)E8 / (4 * 100E3 * 2400(B(max) *2.11(Ae) =
18Turns for ETD49 and for ETD44 = 23 Turns
Now a good guess for the secondary is Ns = VS/VP I want an output of 35 volts , so that means after the rectifiers I need twice that and to help with transits and load changes I will increase that by 5% or so make it 80 volts , I expect at low line to have about 110 volts after the main rectifiers using a 110VAC input so now my secondary turns will be :
18*(80/110) = 13 turns make it 14, After you wind and test your transformer you might find you need to add or take off a winding or two. And at high line you wil need 11 turns make that 12.
So there is your transformer calculations regardless of how you do it the frequency along with the Amount of flux will determine the output power and the core size along with the size of the winding area and wire size.
All this is more or less what I have found in appnotes,books and other articles plus a little experience mixed in. It is up to the designer or DIY to verify this thru research and reading as this forum is full of material covering this subject and I will not get in a contest with which is right or wrong, however if you find an error or have a question I will try to help if I can and I will correct the errors you find ... just remeber a transformer is just that and the law and priciples for their design is basic and well documented its just when you lump them into a design at high frequency that a lot of care must be taken for them to provide proper results.
chas1
To answer your question, you will find many methods for calculating turns and core size. I will try to explain a easy one for me. You must remeber all methods are based on Faraday's law and just rearranged so you can find the unknowns.
Faraday's law:
E= N*AE *(dt/dt) * E-8
E is the voltage across the transformer winding, N is the number of turns on the winding, Ae is area of core in cm^2, DB is flux change (gauss) and dt time for flux to change (seconds)
Now we rearrange the terms and plug in some definitions and we can get the following for finding number of primary turns.This is for halfbridge. For Forward or Fullbridge you double this number because they have to support the rectified mains not just half.
V(in) * E8 / (4 * Fs * Bmax *Ae) = NT(pri)
Now for the core volume to support the power we need:
Lets set up a design for 500watts:
The primary current will be equal to :
Ip= 3 *P(out) /V(in)
V(in) will be calculated based on 110VAC line at it highest value:
130 VAC therefore with a voltage doubler we can expect about:
2(130 * 1.4) = 370 VDC after rectifiers
therefore Ip = 1500 /370 = about 5 Amps from this we calculate the input power which is 1500/ .85 for 85% eff which is about 1750 watts.
core size = .68 * P(out) * current density / Fs * Bmax
Choose a industry standard of 500 c.m. /A for current density and we get:
AcAe = .68 * 1500 * 500E3 / 100E3(Fs) * 1200 (bmax) =4.25 cm^4
If we check the ETD49 specs we find that its AcAe is about 5.7 cm^4 which is a good choice for this design
Look what happens when we raise the Fs to 200KHz:
AcAe = .68 * 1500 *500e3 / 200e3(Fs) *1200(bmax) =2.125 cm^4
so you can see the higher the frequency the smaller the core within reason (copper loss, core temp and other things come into play)
In this case we used a Bmax of 1200 and could have used up to 1600, recalucate using 1600 and then we get 3. 18 cm^4 which means we might be able to use an ETD44 coreset now for the turns calculation:
If we use the highest voltage we expect on our transformer windings when the supply turnon:
N(pri) = 370(V(inmax)E8 / (4 * 100E3 * 2400(B(max) *2.11(Ae) =
18Turns for ETD49 and for ETD44 = 23 Turns
Now a good guess for the secondary is Ns = VS/VP I want an output of 35 volts , so that means after the rectifiers I need twice that and to help with transits and load changes I will increase that by 5% or so make it 80 volts , I expect at low line to have about 110 volts after the main rectifiers using a 110VAC input so now my secondary turns will be :
18*(80/110) = 13 turns make it 14, After you wind and test your transformer you might find you need to add or take off a winding or two. And at high line you wil need 11 turns make that 12.
So there is your transformer calculations regardless of how you do it the frequency along with the Amount of flux will determine the output power and the core size along with the size of the winding area and wire size.
All this is more or less what I have found in appnotes,books and other articles plus a little experience mixed in. It is up to the designer or DIY to verify this thru research and reading as this forum is full of material covering this subject and I will not get in a contest with which is right or wrong, however if you find an error or have a question I will try to help if I can and I will correct the errors you find ... just remeber a transformer is just that and the law and priciples for their design is basic and well documented its just when you lump them into a design at high frequency that a lot of care must be taken for them to provide proper results.
chas1
i dont understand alot in smps design but i have a question . are you designing a smps around a specific transformer that you had its data sheet or you just want to design the transformer as well .according to what i know there is alot of transformer specfication that already exist so we can design our smps on WE DONT NEED TO DESIGN THE TRANSFORMER .DO WE?
Do we need to design a transformer
You might have a point, but in most cases one size does not fit all. Thats why companies that design and wind trasnformers are around and there are quite a few; however I will continue to look for your point. IF YOU HAVE SOMETHING OF VALUE TO CONTRIBUTE THEN DO SO OTHER THAN THAT READ THE WHOLE THREAD TO GAIN IN SIGHT TO THE WHY'S OF CERTAIN POSTS.
Have a nice day
chas1
You might have a point, but in most cases one size does not fit all. Thats why companies that design and wind trasnformers are around and there are quite a few; however I will continue to look for your point. IF YOU HAVE SOMETHING OF VALUE TO CONTRIBUTE THEN DO SO OTHER THAN THAT READ THE WHOLE THREAD TO GAIN IN SIGHT TO THE WHY'S OF CERTAIN POSTS.
Have a nice day
chas1
Thanks chas1 , I hoped you would examine my calculation, hehe now I have to examine yours ..
But the calculation is I presume the same, MUST be cause I used the same faraday law...
Ill post my comments soon ...
so whats max delta flux for etd cores at 100 khz
it seems 0,2T according to datasheet ?!
but I cant find and good datashet with gauss/temperature/current saturation etc. graphs
But the calculation is I presume the same, MUST be cause I used the same faraday law...
Ill post my comments soon ...
so whats max delta flux for etd cores at 100 khz
it seems 0,2T according to datasheet ?!
but I cant find and good datashet with gauss/temperature/current saturation etc. graphs
Transformers
Yeah , Kinda long winded but if you are really interested read pages 284-288 of Switching Power Supply Design by Pressman, all info is there also take a look at the charts he provides on pages 289 - 292 a real eye opener. By the way the equations were not derived by me I only use them and others as cross checks.
chas1
Yeah , Kinda long winded but if you are really interested read pages 284-288 of Switching Power Supply Design by Pressman, all info is there also take a look at the charts he provides on pages 289 - 292 a real eye opener. By the way the equations were not derived by me I only use them and others as cross checks.
chas1
WEALSPEAKER
I am sorry about my remarks if they offended you.
Of course you can ask questions thats what this forum is about helping others gain knowledge to enable them to complete their projects and learn at the same time and yes I think you have more to contribute than you think, just by asking questions you are contributing but when you use caps in your post it is the same as shouting to me. Please use lower case letters unless you are trying to stress an important point even then I don't think its necessary and of course ask all the questions you like I am sure that the members of this forum will respond but please do a little research first on you subject this shows you really have an interest and are not just looking for someone to design your project for you.
chas1
I am sorry about my remarks if they offended you.
Of course you can ask questions thats what this forum is about helping others gain knowledge to enable them to complete their projects and learn at the same time and yes I think you have more to contribute than you think, just by asking questions you are contributing but when you use caps in your post it is the same as shouting to me. Please use lower case letters unless you are trying to stress an important point even then I don't think its necessary and of course ask all the questions you like I am sure that the members of this forum will respond but please do a little research first on you subject this shows you really have an interest and are not just looking for someone to design your project for you.
chas1
E8 , E-8
Scientific notation E3 is 1000 as well E-3 is .001 and E8 would be 100000000 and E-8 would then be .000000001. You might want to research the internet for info on designing halfbridge converters all the info you need is easily found . I would start with TI's website. You can find the formula for the primary current along with how it is derived on the internet. I just apply them as needed. Look for " A 300W 300KHz Current- Mode Half-bridge Converter with Multiple Outputs Using Couple Inductors " by Roger Adair It covers everything just about.
chas1
Scientific notation E3 is 1000 as well E-3 is .001 and E8 would be 100000000 and E-8 would then be .000000001. You might want to research the internet for info on designing halfbridge converters all the info you need is easily found . I would start with TI's website. You can find the formula for the primary current along with how it is derived on the internet. I just apply them as needed. Look for " A 300W 300KHz Current- Mode Half-bridge Converter with Multiple Outputs Using Couple Inductors " by Roger Adair It covers everything just about.
chas1
Keith's website
In post #705 the url is for Keith's website where you will find a lot of info about classd amps and SMPS. The two switch forward converter is covered in detail. Keith has helped a lot with my projects. The reason for this post is this is not my website and I post it again for those who missed it.
Keith's website:
http://www.genomerics.talktalk.net/
Thanks Keith....
chas1
In post #705 the url is for Keith's website where you will find a lot of info about classd amps and SMPS. The two switch forward converter is covered in detail. Keith has helped a lot with my projects. The reason for this post is this is not my website and I post it again for those who missed it.
Keith's website:
http://www.genomerics.talktalk.net/
Thanks Keith....
chas1
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.