It really depends... Side to side winding where you bunch the primary at one end and the secondary at the other leads to your highest leakage. Depending on the separation distance you could see anywhere from 0.7 to 0.9 depending on how you bunch things up.
Absolutely not recommended for any construction except EL ballast (resonant) type designs (low wattage). The efficiency is terrible. For a flyback all that power pent up in the leakage inductance gets released during the flyback phase and you end up having to snub that power out using very hefty snubbers. For a resonant converter you actually use that leakage for the resonant tank but because Kcoupling is so bad, your efficiency is still very bad.
Tony
Absolutely not recommended for any construction except EL ballast (resonant) type designs (low wattage). The efficiency is terrible. For a flyback all that power pent up in the leakage inductance gets released during the flyback phase and you end up having to snub that power out using very hefty snubbers. For a resonant converter you actually use that leakage for the resonant tank but because Kcoupling is so bad, your efficiency is still very bad.
Tony
You need a conversion factor to get the units to come out correctly. This isn't entirely clear from McLyman's book.
Pg=Ki*(10^4W/(((m^2)*(T^2))Hz)*WindowWidth*Lg*Fsw*(Bac^2)
Remember... this is an approximation a finite element analysis so treat any results from this with a certain skepticism.
Since this is the only equation I have ever found for this then I have to go with it since I don't have any other way of doing the hand calculation. I can tell you that this equation drives my magnetics friend nuts. It's strictly an empirically derived formula with not a lot of math around it to prove its validity. Or at least that is what he told me. So I asked him if he had anything better and he said that he didn't so that is what I use.
Tony
Pg=Ki*(10^4W/(((m^2)*(T^2))Hz)*WindowWidth*Lg*Fsw*(Bac^2)
Remember... this is an approximation a finite element analysis so treat any results from this with a certain skepticism.
Since this is the only equation I have ever found for this then I have to go with it since I don't have any other way of doing the hand calculation. I can tell you that this equation drives my magnetics friend nuts. It's strictly an empirically derived formula with not a lot of math around it to prove its validity. Or at least that is what he told me. So I asked him if he had anything better and he said that he didn't so that is what I use.
Tony
Yes, That is true. I found this in McLyman but there is only equetion. no example. using which I can evaluate my calculations.(even mathcad is giving me some absurd values.where as mathcad takes care of units.)
I very much wonder why is this losses not given significance no equation or work is found on this.
understanding that this equation would take me to approximate value, But it is of great help to check if the gap loss is acceptable in our design.
Pl. tell me if I have understood this correct.
we assume that transformer losses are half the system losses. so we say total loss is 2* transformerlosses = system losses.
so for calculation of DeltaT we use system losses. right ?
as explained above if the transformer losses is 7.5watts then system loses would be 15watts.
Where AT is sqmm.
so DeltaT = (15/AT)^0.833= Deg Celsius temp rise.
so the actual temp would be DeltaT + Ambienttemp = core temp.
is it correct?
Thanks..... Finally I think I got it correct.
What percent gap loss is ok? say if its 100W design.
I think at peak power a gap loss of 3% would be acceptable. But that is if I trust the number I am seeing. His gap loss calcs seem to work better for CCM operation than DCM. Since the Bac is much lower for CCM.
Tony
No... The DeltaT calculation is just for the dissipation on the transformer surface area to see if I might be in trouble. the Transformer_Wire losses are dissipated by the surface area of the transformer. If you have a fan or other means of cooling then this no longer applies.
Again this is just a tool to see how the design is progressing. Just like the gap loss model, this is an approximation. Though this model works much better than the gap loss equation.
You are correct on the rest (DeltaT+Tambient)=Core Temp.
Tony
Thanks Tony.
Its time to evaluate what I have learnt. Let me try changing values and see if i too can get any results.
I am talking to transformer winders. who will get me bobbins neatly wound on their machines. After talking to couple of winders, I see they are not equipped for multi strand winding. also after discussion they were reluctant for sandwich winding. I think I will have to personally monitor while they work.
Thank for the great help.
Its time to evaluate what I have learnt. Let me try changing values and see if i too can get any results.
I am talking to transformer winders. who will get me bobbins neatly wound on their machines. After talking to couple of winders, I see they are not equipped for multi strand winding. also after discussion they were reluctant for sandwich winding. I think I will have to personally monitor while they work.
Thank for the great help.
Usually I make the first couple by hand. Then I give them one unit as a sample so they can learn the construction technique. For gapping I have a set of daimond files (ebay) that I use to make the gap. I will file for a while then measure the primary inductance.
When you think you have the calculations the way you want then send the calcs to me and I will review it. Then we will talk about the physical construction.
Tony
As a first step, I collected data of EE core in single excel file. as I would need this in future I have calculated the other parameters. I am submitting it for your review.
Pls check calculation for any one core size. its just the same formula implemented for all.
Dark blue column are calculated values.
Sheet "Data" has Data table I have formulated using Reference sheet. and Sheet "Reference " has core data pasted from manufacturer.
Thanks.
Pls check calculation for any one core size. its just the same formula implemented for all.
Dark blue column are calculated values.
Sheet "Data" has Data table I have formulated using Reference sheet. and Sheet "Reference " has core data pasted from manufacturer.
Thanks.
Dear Tony,
I have lot of questions.
but two of them are
1) What is the procedure to test a transformer, step by step, and the significance of each step.
2) during the construction of transformer you have mentioned margin, creepage. Is this to maintain the voltage isolation required for certifications. what if we don't bother much about the margin tape much, just to check if the calculations are correct?
Thank you
I have lot of questions
. but two of them are
1) What is the procedure to test a transformer, step by step, and the significance of each step.
2) during the construction of transformer you have mentioned margin, creepage. Is this to maintain the voltage isolation required for certifications. what if we don't bother much about the margin tape much, just to check if the calculations are correct?
Thank you
Dear Tony,
I have lot of questions
but two of them are
1) What is the procedure to test a transformer, step by step, and the significance of each step.
2) during the construction of transformer you have mentioned margin, creepage. Is this to maintain the voltage isolation required for certifications. what if we don't bother much about the margin tape much, just to check if the calculations are correct?
Thank you
1) Test the design parameters (Leakage, Primary side inductance, turns ratio, Q-determines the transformer efficiency, Self Resonance-determines what the inner winding capacitance is so when you do EMI you have an idea where to tlook for issues, and DCR). Next run you transformer in your design at full load,, lowest input voltage for at least an our so it hits thermal equilibrium then look at the primary side current through the transformer. You are looking to see iff there are any signs of saturation.
2) Margin is actually important. If I am going to spend thousands of dollars on my audio ezuipment, I do not want a lighting strike somewhere taking it out. Now if you are doind a DC/DC converter then you probably can get away with very rudimentary insulation and not bother so much with the creapage distance. I wll tell you that for anything I sell commercially, I always make sure that i follow the requisite safety guidelines. They were put there for a reason and the last thing I would every do is tell someone to ignore them.
I hope that helps.
Tony
Thanks Thanks... Thanks Tony for being here again.
For Point 1, I ll try to find the test procedure in McLyman for the parameters you suggested. I would also request you to suggest any document available.
point 2, I don't wish to compromise on safety. I would never forget how important margin is. The reason why I asked here was, I wanted to understand that if that is required for proper working (as specifications) of transformer or is introduced for safety.
Yes your answer always helps.. Thanks again.
For Point 1, I ll try to find the test procedure in McLyman for the parameters you suggested. I would also request you to suggest any document available.
point 2, I don't wish to compromise on safety. I would never forget how important margin is. The reason why I asked here was, I wanted to understand that if that is required for proper working (as specifications) of transformer or is introduced for safety.
Yes your answer always helps.. Thanks again.
For Number 1) A LCR meter usually has most of what you need. Lp is easy try and test at your CCM frequency, for a QR that would be your full load frequency. Q should also be on the meter (sometimes). Leakage is done by measuring the primary winding and shorting all the other windings together. DCR is fairly obvious. Inner winding capacitance is a bit trickier. For the capacitance measure the primary side winding inductance with the other windings open. Keep increasing the frequency until the inductance peaks and then goes negative. The first peak is the SRF. That and Lp give you the inner winding capacitance. In general lower is better except when I have an EMI issue. Then I will add overlapping C shields and sandwich layers to add capacitance at strategic frequencies to damp the EMI fields. These are tough to tune and usually take several days of different iterations to achieve the correct balance and to make the reproducible across a production run.
Point 2) Both creapage and isolation are important. For example if I am using straight margin tape for creapage, then teflon tubing on the exit wires where it crosses the margin tape all the way to where it solders to the transformer pin. Also do not forget that the core itself has to be declared as either primary side or secondary side references. Then you have to add appropriate tape on the core to insulate it from the coponenst on the other side. For exammle if I reference the core as primary side then I need 5-6mm creapage between it and any conductive surface on the secondary side. If I wrap the core in 2 layers of insulating tape then I can snug the parts up close to the transformer on the secondary side. The same goes for inside the bobbin. If I have secondary side wire overlaying the primary side then I need at least 2 layers of insulating tape between them. If I am using triple insulated wire then I don not have the tape requirement inside the transformer. The above is what is required for 60950 compliance for reinforced insulation systems. If you are already isolate somewhere else (as in DC/DC converters) you may get buy with only basic insulation.
Tony
Point 2) Both creapage and isolation are important. For example if I am using straight margin tape for creapage, then teflon tubing on the exit wires where it crosses the margin tape all the way to where it solders to the transformer pin. Also do not forget that the core itself has to be declared as either primary side or secondary side references. Then you have to add appropriate tape on the core to insulate it from the coponenst on the other side. For exammle if I reference the core as primary side then I need 5-6mm creapage between it and any conductive surface on the secondary side. If I wrap the core in 2 layers of insulating tape then I can snug the parts up close to the transformer on the secondary side. The same goes for inside the bobbin. If I have secondary side wire overlaying the primary side then I need at least 2 layers of insulating tape between them. If I am using triple insulated wire then I don not have the tape requirement inside the transformer. The above is what is required for 60950 compliance for reinforced insulation systems. If you are already isolate somewhere else (as in DC/DC converters) you may get buy with only basic insulation.
Tony
this may help you to design ferrite core transformer
Ferrite transformer turns calculation with example
Ferrite transformer turns calculation with example
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.