This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

PMP8740: Full Bridge Transformers

Part Number: PMP8740
Other Parts Discussed in Thread: UCC28950, TINA-TI, UCC28180, LM5023

Hello Dear Experts ,

- Can you share detailed information about Full Bridge main transformer T4 (PQ50-50 145400-A), Gate Drive transformers T2,T3 (56PR3362) and T7 transformer like Number of Turns (Np/Ns),

Primary and Secondary Inductance, Leakage Inductance, Wire Gauge of Primary and Secondary windings and Core Meterial of transformers ?

Thanks.

  • Hello Per,

    most of the magnetics are custom designs by magnetic vendors:

    PSFB xfmr T4 is made by KASCHKE
    pulse xfmr T2/T3 by VITEC
    shim ind T7 by CLICK

    Please get in contact w/ vendor for further info,

    Best regards, Bernd

  • Hello Bernd ,

    Firstly thanks for your reply,

    I contacted with KASCHKE . Their answer  " please note that we need a LOA from TI that we are allowed to send you samples or an offer.

    The reason is, that this is a customized part" .

    Actually , I have no idea about LOA.

    -Can you give me information about LOA ? 

     

  • Hello Per,

    head designer of PMP8740 will inform KASCHKE,
    so you will get the magnetics...

    Best regards, Bernd

  • Hello Per,

    The reason why Kaschke wanted to have our allowance, is because this reference design has been built for another customer.

    But this design is now public, so they can accept request of samples, or purchase for this design. This includes the main transformer T4 and the inductor L3.

    The shim inductor T7 has been built by hand. The construction details are visible at page 21 of the article I wrote in the recent Power Supply Design Seminars, available at the link:

    If still Kaschke is reluctant to provide samples, please let me know the email address of your contact at Kaschke and I will explain them the situation.

    Best regards,

    Roberto

  • Hello Roberto,

    Thanks for your reply.

    Yesterday.  I contacted with Kaschke . Kascke is still reluctant to provide samples.

    I contacted with Mr. Hagen Fuchs. (Key-account-Manager at Kaschke Component)

    E-mail address: H.Fuchs@kaschke.de

    If you can contact and explain the situation. I will be glad

    Thanks for your interest.

  • Hello Bilgin,

    I just wrote an email to Mr. Fuchs, explaining him that this design is not anymore restricted only to one customer.

    Let´s wait for the answer.

    Please let mw know if have still questions.

    Regards,

    Roberto

  • Ok. Thanks a lot Roberto,

  • OK Bilgin,

    Just let me know if you need further support.

    Best regards,

    Roberto

  • Hello Roberto,

    I still haven't received answer from Mr.Fuchs. 

    Today I will write an new email to Kaschke and I will wait for their answer.

    If Kaschke  reply to my email.I will inform you.

    Did you get an answer from  Mr. Fuchs  for your email ?

    Thanks.

  • Hello Bilgin,

    I also didn´t receive any answer from my email to Mr. Fuchs.

    Maybe due to COVID-19 situation some offices are working with limitations.

    If I get an aswer, I will inform you ASAP, no problem.

    Please let me know if you get one, or even if you don´t get any.

    Regards,

    Roberto

  • Hello Roberto,

    Unfortunately ,I didn't receive any answer from Kaschke. I decide to make own transformer. I read  past posts from the PMP8740 forum and I have some information about T4 transformer like  Np = 19 turns, 1.6mm diameter (LITZ wire); Ns1 = Ns2 = 2 turns 6mm diameter (LITZ wire). Also,  I found "PMP8740 UCC28950 Excel design tool " and T4 transformer technical datasheet . In the datasheet transformer core metarial is PQ50/50 / K2010 or equiv.

    Here I have some questions ,

    1) For  PQ50/50 / K2010 core , I didn't find K2010 core meterial , I found PQ50/50 N95 , PQ50/50 N87, PQ50/50 N97...etc.

        Which one is suitable for this design ?

        Transformer core material link:

        https://www.mouser.com.tr/Search/Refine?Keyword=PQ50%2f50

    2) Also, In excelsheet there are some informations about Primary Magnetizing Inductance=3,2mH, Transformer Primary DC Resistance=40mH,Transformer Secondary DC Resistance=0,8mH, Measured Transformer Primary Leakage Inductance=10uH.

     -Are these information true for T4 transfomers ?

    3)In the UCC28950 ic Design & Development website UCC28950 and UCC28951 TINA-TI Steady State Reference Design (Rev. A) simulation circuit is available , this circuit is for 12V 50A 600W  dc-dc converter.

      Now, I am trying to simulate your 2kW (PMP8740) dc-dc converter circuit in TINA-TI . I am using PMP8740 dc-dc converter schematic circuit identically except for Voltage and current loop Op-amps . I adjusted the output of dc dc converter 32V 1920W at full load because in the PMP8740 UCC28950 Excel design tool " output is 32V and 1920W. My purpose is determine  T4 transformers technical specs and understand the circuit.

     There are some differences between  Excelsheet results and PMP8740 dc-dc converter schematic circuit. For example, Redelab,Redelcd,Redlef resistors etc. 

    - Which one can I reference circuit for first step 32V 1920W simulation ?

    -Why there are some differences between  Excelsheet results and PMP8740 dc-dc converter schematic circuit ?

    I attached excelsheet and transformer technical datasheet below.

    Thanks.

    4113.PMP8740 UCC28950 Excel Design Tool_SLUC222C.xls145400-a_eng.pdf

  • Hello Bilgin,

    You can use the suggested data regarding the transformer and employ PQ50/50 core from EPCOS, with Np = 19 turns, 1.6mm diameter (LITZ wire); Ns1 = Ns2 = 2 turns 6mm diameter (LITZ wire). As core I would select N95 powder because, even though it has a bit higher core losses, with 19 turns you will reach ~ 3mH magnetizing inductance (for this material the Al is 8200 nH/T^2). For the point #2, I don´t know if these data are correct, but if you select the suggested LITZ wires, the copper losses will be for sure under control.

    I have never simulated my full bridge stage, I just relied on the excel calc-sheet, which is very accurate.

    Nevertheless, the delay times are, also for my prototype, starting points (and I didn´t change them finally). After you build your own, you may want to fine tune these parameters in order to optimize the efficiency. So, you can use the value I employed in my PMP8740, as starting point, without problems. 

    Best regards,

    Roberto

  • Hi Roberto 

    Thanks for your valuable informations.

  • Hi Roberto

    I have questions about T7  Resonant (shim) inductor and T4 transformer design. I need your help.

    1) In the shim inductor 10 turns (160 X 0.1mm) litz wire was used. This means ,

    From πr2  ,  1 pcs 0,1mm litz wire has "3.1416 X 0.05mm X 0.05mm = 0.007854 mmcross sectional area.

    160 pcs litz wires have " 160 X 0.007854=1.256  mm  cross sectional area in total.

    Instead of  1.256  mm litz wires. Can I use copper sheet winding with same cross sectional area ?

    For example, " 0,1mm X 12,56mm = 1.256  mm" copper sheet winding. Is it OK ?

    2) In the T4 transformer, primary side 19 turns 1,6mm diameter and secondary side 2 turns 6mm diameter  litz wires was used .

    Is it 1 pcs litz wire with 1.6mm diamater ? or Many  litz wires were used to create 1.6 mm diamater ?

    For example , Like shim inductor Could it made by  16 pcs and  0.1mm (16X0.1mm) litz wire in total  ?

    If more than one litzs wire was used to create 1.6mm diameter, How many litz wires are used and What is the diameter of these wires?

    Same  questions for secondary side.

    3) Can I use copper sheet winding in the T4 transformer instead of litz wire ? Is it suitible for T4 transformer?

     Thanks.

          

  • Hi Bilgin,

    Sorry for not replying earlier, but yesterday and last Friday were holidays in Germany.

    Let´s reply point by point:

    1) The calculation of cross section area is correct. In theory it´s equivalent to use a copper foil 12.56mm-width x 0.1mm-thickness, instead of 160 wires with 0.1mm diameter. The difference is that the interwinding capacitance is pretty much higher with copper foil, leading to multi-resonance frequencies. So, maybe it doesn´t affect the performance, but I would test both solutions, even just winding prototypes.

    2) As far as I rememeber (I don´t have this data, because the transformer was already built) the primary side wire was a single LITZ wire, while on secondary side it was a pack of several LITZ twisted together, but I don´t know the exact amount of wires.

    3) As for point 1, you can use copper foil instead of LITZ. In this case it makes more sense and I would suggest to do that, because we have only a limited amount of turns, which reduces the total amount of interwinding capacitance.

    Thanks and best regards,

    Roberto

  • Hi Roberto 

    Thanks for your help .

    I want to make L1 pfc inductor (in pfc circuit) and L3 inductor (in dc-dc converter circuit) by hand and I have some questions about these inductors. 

    For Pfc Inductor:

    Which core can I use for L1 pfc inductor? 

    Which wire can I use for L1 inductor and what is your suggestion for wire diameter ? 

    For L3 Inductor:

    From the datasheet ,  L3 has  HS-400 /26  core material but I haven't  HS-400 /26 core. 

    Which core can I use to be alternative for HS-400 core ?

    Thanks.

  • Can you also share the technical details of T4?
    I would like to build a 240Vdc to 24Vdc battery charger, using a planar transformer. So a different ratio but I really would like to know what you have used for your 400Vdc to 24V converter

    Thanks in advance,
    Mark

  • Hi Bilgin,

    For the PFC inductor it is possible to use the one specified in the BOM or wind it manually.

    In my case, I made it by hand with two cores in parallel from NKL, part number CM400125 D0522-00.

    One single core has external diameter = 40mm and thickness 15mm. By putting these two cores in parallel, the thickness will be 30mm, and the Al value double (The total inductance is 210uH).
    I wounded 35 turns of single enamelled copper wire, with a diameter of 1.6mm.
    I added isolation tape before winding; in theory is not necessary, because the core is already coated, but it increases the isolation, and it´s easier to slide the turns in and out.

    The inductor L3 is more complicated, because this core is unique.

    You might replace this inductor with some off-the-shelf parts, like the 6.8uH:

    Or the 10uH:

    Depending on your nominal current; if > 60A I suggest 6.8uH, while for Iout < 60A, please use 10uH.

    Best regards,

    Roberto

  • Hi Mark,

    The details of the transformer T4 are described in my reply on 04-07-2020.

    I attach anyway the datasheet of this transformer:

    7220.145400-a_eng.pdf

    In order to calculate the new transfomer, I suggest to use the excel calc-sheet of the UCC28950, available at the following link:

    Here you can enter the main specifications regarding Vin, Vout, output current etc. The tool gives as result the magnetizing inductance and all currents, primary and secondary side.

    Best regards,

    Roberto

  • Hi Roberto 

    Thanks a lot for your help .

    For Pfc inductor, The total inductance is 210uH and number of turns=35 and the total Al value is 171.43 nH /N^2 . One core has 85.714 nH / N^2 Al value.      I didn't find same NKL  CM400125 D0522-00 part number core. I find micrometals powder core  Al= 80nH /N^2 and u=60. When I used 2 core from micrometals, the total Al is 160nH/N^2  and to get 210uH inductance , I should wound 36 turns (207.3uH total inductance approx). I attached datasheet of the micrometals core. Is this core suitable for my PFC inductor ?   

    Also I have question about T2 and T3 full bridge gate drive transformers . Can you suggest any transformer with part number for T2 and T3 transformers?

    Thanks.

    PFC Core.pdf

  • Hi Bilgin,

    I verified that the core you selected fits for the PFC inductor, so there is no problem using it.

    Regarding T2 and T3, you can also employ the part P0584NL, which is also used in the phase shift full bridge design (350W) downloadable at the link:

    www.ti.com/.../PMP5568 FULL FEATURE&tisearch=Search-EN-everything

    The datasheet of the transformer is attached.

    P0584NL.pdf

    This transformer has lower magnetizing inductance, compared to the one in the BOM, but I verified that the magnetizing current isn´t too high and is manageable by the gate drivers. Also the volt per usec product is ok, so no saturation will occur.

    Best regards,

    Roberto

  • Hi Roberto 

    Thanks a lot  for your help.

    I have a question about gate drive transformers .In the  datasheet there is  also P0584ANL part number transformer. Do you suggest this transformer instead of P0584NL ?  and Are there important differences between  P0584NL and P0584ANL transformers? 

    My another question about PFC Heat sinks. In PFC schematic HS1 for bridge rectifier and part number SK 481-50 and length of this heat sink  50mm. It is Ok.    HS2 and HS3 has SK 481-100 part number and length of this heat sinks are  100mm . If HS2 and HS3 are used together , total  heat sink length  should be 200mm. But in the PFC layout total length is 100mm.  Are HS2 and HS3 part number  correct ?  and Why do you use two seperate heat sink to obtain 100mm lengt.? Could one heat sink to be enough? 

    My last question about AC voltage sensing transformer T2 PA3000NL in PFC schematic. I didn't find  this transformer with same part number. Can  you suggest a transformer instead of PA3000NL ?

    Thanks.

  • Hi Bilgin,

    The difference between P0584NL abd P0584ANL is the the version with "A" is 4250Vrms rated, while the non-"A" version is 3000Vrms. So depends on the specification about the isolation voltage in your application you may want to use one of them. From my perspective, since both are reinforced, you may want to employ the one at 3000Vrms, because it has lower leakage inductance.

    You are right, the heat sink of the bridge is 50mm, and also HS2 and HS3 (placed in a row) are 50mm (at least as I made my first prototype).

    Then I replaced HS2 and HS3 with a single 100mm (according to the BOM), but I forgot to remove one heat sink from the BOM.

    So, please, use only one heat sink HS2 or HS3, 100mm.

    AC voltage sensing transformer: I am not sure we need here this exact part number because I used it as current transformer and not voltage transformer (like it is). As you can see from the schematic, the load of this transformer is a "virtual" short circuit (done with an op-amp) therefore the magnetizing current doesn not play any role. This way you should get the same result by using a "normal" 50/60 Hz small transformer, or even better an audio transformer, like the one at the link below:

    Best regards,

    Roberto

  • Hi Roberto ,

    Thanks a lot .

    Ok. For gate drive transformer .I will use  P0584NL  transformer and for heat sink I will use one heat sink 100mm lenght .

    For AC voltage sensing transformer : In audio transformer you suggested me , turns ratio Np/Ns 1:1 and 2:1. Which one can I use for same voltage sensing circuit in the schematic. 1:1 or 2:1.  In PA3000NL datasheet turns ratio 1:1.

  • Hi Bilgin,

    I suggest to use the 1:1; this way you don´t have to modify any parameter in the AC voltage sensing section.

    Just let me know whenever you need more support.

    Thanks,

    Roberto

  • Hi Roberto,

    Thanks a lot.

    Ok for voltage sensing transformer. 

    I have another questions

    In Pfc schematic one pin of bridge rectifier heat sink (HS1) connected to the GND-P and one pin of HS3 (for boost diodes and mosfets ) connected to 1000pF capacitor.  Are these connections  important ?  

    Also , in pfc schematic there is a BAS21 (D100) diode connected to the pin 3 of ucc28180 ic in rev. E schematic but in PFC pcb layout I didn't find D100 diode. Why did you add this diode in revision E schematic.? and  Where can I place this diode in pcb layout ?

    My last question : In Pfc schematic ,   "PCB Rule " was written above the some connections. What does it mean?  Are there  important points here for layout?

    I attached related file.

  • Hi Bilgin,

    Normally, when a component has to be isolated and put on a heatsink, the heatsink can be left floating.

    In order to avoid building voltages on the heatsink (which can be dangerous by touching), it´s better to ground it.

    If, instead, the component to be dissipated is switching, then it´s important to connect the heatsink to ground, with or without a capacitor in series.

    This avoids the heatsink acting as an antenna and radiating noise.

    The reason why I prefer to add a small capacitor in series is: first, a small capacitor is a short circuit for the high frequency band of the noise (typically due to high dV/dT), therefore it´s like it´s shunting the most of the noise to grund. Second, if the capacitor fails, and shorts, there is still the isolation pad (Sil-Pad) between the FET and the heatsink. In case the Sil-Pad fails and shorts the FET drain to the heatsink, we have still the small capacitor isolating the heatsink to ground.

    Without the small series capacitor, so by shorting the heatsink to ground, if the Sil-Pad fails, the FET will be damaged, because its drain will be shorted to ground.

    In the case of the bridge (dissipated by HS1), since the housing is plastic and it´s not switching, HS1 can be directly connected to ground.

    I added in the later revision a diode BAS21 (it has reference > 99, therefore it´s not in the actual board layout but must be added later) according to the 8.3.14 paragraph of the UCC28180 datasheet, see below:

    In theory this diode is not necessary in this design because we have here inrush current limit, but I would always add it, in case you want to derive from this project a lower power version, where you may want to remove the inrush current limit, maybe done by means of a simple NTC.

    The location of this diode is not critical, so you can put it where you have space.

    The PCB Rules you see on the schematic have been added to route the board layout easier. For example here this rule concerns an isoaltion from the net to the rest. By doing that, when I place a copper pour close to this net, the CAD software creates automatically the isolation. 

    Best regards,

    Roberto

  •  Hi Roberto ,

    Thanks a lot for your valuable informations .

    I have another questions  about resonant( shim ) inductor and its calculation.

    -From the resonant inductor pdf  PQ20/20 N97, gap=0.77mm Ue= 57 Epcos core are recommended for shim inductor  but  the datasheet say that the Epcos N97 core has not  gap , this core is ungapped core. My question is here, the total winding lenght of N97 core for wire or copper foil is 7.15mm X2=14.3mm. When I add 0.77mm air gap the total lenght will be 15.07mm without tolerances. Is it true? So I  add gap later  for N97 core or Can I select core with gap ?

    -My another question about calculation of shim inductor .From the formula of shim inductor calculation  :

    Coss(Qa,Avg)=336pF, Vin(Fb,nom)=400V, Load(min)=0.15, IPP=7.526A,ILout=12A, Nps(fb)=9.5 , Llk=3.5uH the shim inductor(Ls) value  is  118uH  approx. (If we want %15 ZVS and we use IPP65R190CFD mosfet). For example If I want to adjust the minimum load current for ZVS operation to 30%  Ls is 26.9 uH. approx. Is it true ?

    -If I adjust ZVS operation to  30% , Should I make any changes in the  the full bridge circuit ?

    -Also, there is a leakage inductance Llk=3.5uH from the shim inductor formula. What is this leakge inductance? . Does it come from full bridge main transformer primary leakage inductance?

    I attached N97 core datasheet and Shim inductor calculation pdf.

    Thanks.

    pq_20_20.pdfResonant Inductor.pdf

  • Hi Bilgin,

    Sorry for the late reply, but yesterday it was holiday in Germany.

    I am not sure I understand your question about the winding lenght for the copper: do you mean the window area? From my understanding, the width available to the winding is 11.7mm, while the height is 17.2mm (external diameter of the bobbin) minus 10.9mm (internal diameter) divided by two = (17.2 - 10.9)/2 = 3.15mm. This height does not change when you add the air gap (simply the core will rise a bit, but the bobbin is fixed). If you multiply the two dimensions of the winding area (3.15mm x 11.7mm) you get 36.85mm2. The datasheet specifies it as An (which is for PQ20/20 = 35mm2, here a small margin is taken).

    You are right to say that there is no gapped core available, infact I added it with the help of spacers. Please consider that adding gap on all legs is equivalent to have a single gap in the  middle leg, but twice the thickness.

    So, you can add manually the gap, or select a different part, with orderable gapped half-cores.

    Shim inductor:

    I think I simply used very low Coss MOSFET at the beginning of the development, and then I calculated the shim inductor. For the calculations you described, the new value of the shim inductor is correct.

    I suggest you to have ZVS operation down to 30% and use 26.9uH inductance, which is the result of the formula of the slide, minus the leakage inductance of the transformer.

    In case I didn´t yet send you this link, here are the whole calculations for a UCC28950 PSFB topology:

    Best regards,

    Roberto

  • Hi Roberto

    Thanks. 

    Sorry for my poor explanation. Ok I understood  your explanations. I wanted to say the height of one core for winding is 7.15mm  when I use two core, the total height will be 14.3 mm without gap. With 0.77mm gap , the total height will be 15.07 mm . I attached  my  related png file below. I drew my explanation  in the photo . I hope it will be more understable. 

    For shim inductor,

    Here , I studied  whole calculations for a UCC28950 PSFB topology pdf which you added. My purpose is to use same full bridge circuit in the PMP8740 design. In the calculation   COSS _ QA _ SPEC   and  COSS _ QA _ AVG values available. Using the COSS _ QA _ SPEC value from the datasheet , COSS _ QA _ AVG value is calculated and this value was used for shim inductor calculation but you was suggested "effective output capacitance, time related  for the calculation of the ZVS times."  I didn't use this calculation and I used effective output capacitance, time related value from the datasheet.       Are there imporatant differences between COSS _ QA _ AVG  and effective output capacitance, time related value

    As I understood from the calculations for 30% ZVS I don't need change the full bridge circuit . What do you think ? Is it true?

    If I think wrong  , Can you suggest proper mosfet for %15 ZVS circuit  because I dont want to change full bridge circuit for my desing. I want to use  same circuit. Also , I found IPP50R199CP mosfet with Coss=160pF time related from the PMP8740 UCC28950 excel file. Maybe you used this mosfet for first prototype.  Using this mosfet Ls=54uH approx.  

    My last question about full birdge transformer , If I adjust the full bridge main transformer leakage inductance to  high enough  for example 54uH  . My calculation is also 54uH required  for shim inductor . Result will be zero and I don't need shim inductor . Is it right ? Are there any problem here?

  • Hi Bilgin,

    OK, now I understood your point: you want to use the core without the bobbin, therefore the total width of the winding will be 14.3mm, while I always consider to use the bobbin, which has smaller width. BTW, I am not sure you can make the winding without bobbin, because, if done in free air, the wire might get in contact with the core, which probably doesn´t widthstand 400V isolation voltage. Anyway, only for prototyping, and arranging a sort of isolation, it should be fine.

    Regarding the Coss of the MOSFETs, probably I didn´t use the right numers at the beginning, and also the old FET IPP50R199CP, even with lower Coss, is not suitable when ZVS is lost, because of its slow Trr of the internal body diode. The "CFD" MOSFET series are better suitable for this.

    Which Coss should be taken into account for the calculation of shim inductor?

    In the MOSFET datasheet, there are two numbers:

    The "energy related" should be considered when calculating switching losses (for hard-switched converters), while the "time related" should be used for the calculation of the time delay between switching the MOSFETs A and B in the same leg. This capacitance takes into account its variation during the swing of the voltage on the Vds from 0V to 400V.

    So, I suggest to use this capacitance, and employ the CFD series MOSFETs (the above is the section of the MOSFET I used in the latest revision of PMP8740.

    If you want to reduce the value of the shim inductor, you can do it freely: the power stage will not change at all.

    As rule of thumb, we can say:

    1) Higher shim inductance --> lower effective duty cycle, slightly higher peak currents, higher conduction loss at full load, higher efficiency at light load; the efficiency curve tends to flatten at broder load currents

    2) Smaller shim inductance --> higher effective duty cycle, lower peak currents, lower conduction loss at full load, lower efficiency at light load; the efficiency curve tends to be more likely a "mountain", with the highest efficiency at 3/4 to full load

    Finally, I don´t suggest to remove the shim inductor and use only the leakage inductance Llk of the main transformer, because with external shim inductor and very low Llk, the diodes D7 and D10 are clamping the voltage on primary side of the transformer; this clamping is also translated to the secondary side by reducing the spikes on the rectifiers (or sync-rectification FETs).

    So, at the end of the day, you can select proper shim inductor (I suggest to go with 30% minimum load for ZVS) and use the schematic as is, without any modification.

    Best regards,

    Roberto

  • Hi Roberto ,

    Thanks a lot for your detailed informations.

    Ok. I will use 30% minimum load for ZVS and  I will use IPP65R190CFD mosfet as you used the lastest version of PMP8740. As a result , my shim inductor value will be Ls= 26.9 uH approx. and I will use full bridge circuit schematic without any modification. 

    I understood the differences between higher and lower shim inductance .Also, the full bridge transformer should have low leakage inductance and Shim inductor should be used to reduce the spikes. It is ok for me.

    Dear Roberto , You was attached a file , but I didn't open it.and I didn't download it. Can you attached this file again? 

  • Hi Bilgin,

    The missing section of the datasheet was referring to the MOSFET I used in the last revision of PMP8740 (and which you are also considering to use), downloadable from the manufacturer website:

    Then in page 6, table 7, of this datasheet you can find the parameters I was talking about.

    Best regards,

    Roberto

  • Hi Roberto ,

    Thanks.

    Now, I am trying to make my full bridge transformer. I have some questions. 

    I have shim inductor technical spefications for winding . For example , 10 turns litz wire (160X0.1mm) . For full bridge transformer .  I had asked you before and You said primary side of transformer 19 turns 1.6mm single litz wire (one litz wire). Secondary 1 ,and Secondary2 =  2 turns 6mm litz wires (pack of several litz wires together for secondary side).

    If I use single 1.6mm litz wire for primary side, I think it will be not good for skin effects. Because of this reason , I want to use copper foil instead of litz wire and  You was suggested this method for full bridge transformer .Also, using the copper foil the leakage of transformer will be more less.  

    Normally , the amount of current flows  from the full bridge transformer primary side should be equal to the the amonut of current flows the shim inductor. So , same magnitude of current flows the shim inductor and transformer primary  side . From this point , Can I use (160X0.1mm ) litz wire instead of 1.6mm single litz wire for full bridge transformer primary side ? Am I correct?

    Also , I had calculated 160X0.1mm  litz wires area is equal to the 1.256mm2  cross sectional area in total. (For shim inductor). If my idea is correct , I will use 1.256mm2  copper foil instead of single 1.6mm litz wire for full bridge transformer primary side .  

    Thanks .

  • Hi Bilgin,

     

    Sorry for the late reply, but Thursday last week was holiday in Germany and I took one day off on Friday.

    Probably I didn´t correctly explain the LITZ wire selection: as 1.6mm LITZ wire I meant a LITZ wire, composed typically by 0.1mm single wire, built with so many strands to form 1.6mm total diameter.

    I attached a list of these wires, which I use for my reference (it´s in German, but should not be too difficult to derive the right parameters). The "Aussendurchemesser mit Umspinnung" refers to the external diamter with the twisted isolation inlcuded. If you consider the "Einzeldraht Nenndurchmesser" (single wire nominal diameter), the one I use normally, as said, is 0.1mm because it´s mainly used for the frequency I work with high power, like 100KHz...200KHz. From this table, it´s possible to see that for a maximum 1.6mm diameter, we can select 120 x 0.1, so I suggest to use this one. This wire has a "Gesamtleiterquerschnitt" (total cross section) of 0.96133 mm2 and has typical resistance of 18.5 Ohm / km (or 18.5 milli-Ohm / m).

    So my suggestion is, since we know that the RMS current through the primary side of the transfomer is the same as the current through the shim inductor,  to use this wire in both windings, shim inductor and primary winding.

    If you can find thelitzentabelle.pdf big LITZ wire for the secondary side, you can use it, or employ copper foil.

    Best regards,

    Roberto

  • Hi Roberto ,

    Thanks  a lot for your clear explanation. I have no doubt about full bridge transformer .

    Now, I want to change auxliary power supply circuit . In this design , aux power supply output is 10W but I  want to change this output to the 24W (12V ,2A).     My first question is about some calculations in the white paper .For example , Vrefleceted voltage  is important for transformer turns ratio. According to white paper    Vin(min)=120V , Vin(max)=400V + overshoot ~ 420V approx.  Vreflected < Vin(pfc,peak)(reduced by 20% percent) was used to choose Vreflected value.  Here I have a doubt because If Vin(pfc,peak) is equal to 420V . 20% of Vin(pfc,peak) is equal to 84V or  20% reduced by 420V is equal to 336V approx. Vreflected <336 V is a large margin because Vreflected =100V was chose . How did you choose the Vreflected value ? 

    -Also , what is the maximum duty cycle for this flyback converter? and  What do you suggest?

    My another question is about LC output filter of flyback converter. As I understood , Lf value is equal to 2.2uH and Cf=1000uF. When  I change output is to 12V 2A , Can  Lf and Cf values change or not ? and How did you calculate these values ? 

    My last question is about Snubber circuit , In this converter RCD clamper circuit is available but RCD snubber circuit for resonance(in parallel to Q6 mosfet ) is not available . Do you prefer this circuit  adding to schematic?   

  • Hi Bilgin,

    Please consider that the voltage applied to the auxiliary power supply (called Vin(aux), according to the white paper) has different values according to two states:

    1) At startup this voltage is equal the output of the PFC stage, which is coincident to Vin(pfc,peak) , due to the fact that the Boost rectifier charges the PFC output capacitor to input AC mains peak. Since the PFC stage has not yet started, this peak voltage does not have so much ripple, and the 20% we considered is just for margin.

    2) When the PFC stage started  and is delivering the nominal output (400V), the auxiliary power supply will be supplied with this 400V; we also consider a maximum overshoot of 420V, due to the intrinsecally slow bandwidth of the PFC

    Now Vin(pfc,peak) = 85VAC * SQR(2) = 120.2V; by subtracting 20% from this voltage for margining, we end up the minimum Vin for the aux. supply, which is 96.16V (which is rounded to 100V, because the 20% is just suggested value).

    Differently from other quasi resonant controllers we have, the duty cycle is not restricted to 50%, but we usually consider this value for the LM5023 as working point at minimum Vin, in order to minimize the reflected voltage on primary side (which is key for selecting the MOSFET), reduce the spikes, therefore a proper selection of the turns ratio of the transformer is necessary.

    Regarding the values of Lf and Cf, these components depend on the specifications about output ripple voltage and transient response requirements. Since I didn´t have any specific number, I kept the typical values, like 0.5% ripple annd 1...2% Vout deviation when the load is switched from 50% to 100%. I believe you can just double Cf and keep the same inductance value but capable of carrying more current. There is plenty of documentation about how to select filters and output capacitors, but I think you don´t need to select specific values here. Please let me know if you need further documentation about how to close the loop, and select those components.

    Snubber: this is a low power auxiliary converter, and the noise generated from it was pretty small, so I didn´t add RC snubbing network. Since you are doing a new transformer, I suggest to add as placeholders the RC snubber, which you can fine-tune later, when you test the board. As first calculation, I would use the Power Stage Designer tool, available in TI web site.

    Best regards,

    Roberto

  • Hi Roberto , 

    Thanks .

    As first calculation  for LM5023 ,I used power stage designer tool in TI website. There are some input parameters for calculation. For example Vin(min) RMS, Vin(max) RMS , Vout, Iout . Here my Vin(min)=120V DC and Vin(max)=420V DC (with PFC). So  Which input parameter can I select for first calculation?. For example , Vin(min)=85V RMS , Vin(max)=265 Vrms . If I select this one maximum input DC voltage will be 265 X sqrt(2)= 374V DC approx. but this one is not enough I think because our circuit has PFC stage.  If Vin(max)=300 Vrms  maximum input dc voltage will be 420 volt . Which one is suitable for this desing ?  Because when maximum input voltage changes . Some components change in circuit.

    Also, Can  you share documantation about close loop and selection of this components . It will be good for me .

    Thanks.

  • LM5023 85VAC-300VAC to 12V-2A - WBDesign.pdfHi Bilgin,

    I agree to use 300VAC for the maximum Vin, because this makes the peak voltage = 424VDC.

    The Power Stage Designer is a very good tool, but also you may want to have a look to Webench.

    I entered the data for you; if I remember it correctly, you need to specify a power supply 85VAC to 300VAC, 12Vout at 2A, correct?

    If that is right, I attached what comes out from Webench. You may want to input different values: in this case, please, just access the product page of LM5023 and enter the data in the Webench window.

    Best regards,

    Roberto