Analysis and Design of a Flyback, Part 7, Testing the Transformer
Vložit
- čas přidán 17. 09. 2013
- In part 7, I test the flyback transformer and compare the measurement results with simulation results. I show the breadboard techniques that I use that allows easy installation of the transformer and Mosfet. I present the ringing in the drain due to the parasitic leakage ringing with the mosfet's drain to source capacitance. And finally, I show how to damp the ringing by using a snubber. For comments, questions can be posted here or can be emailed to
rbola35618@aol.com.
For TopSPICE files request, email me at rbola35618@aol.com. These files can run using the TopSPICE demo available at www.penzar.com
My your simulation always converge ;-)
Robert
I really admire your effort to provide such content. To be honest, i understood more about smps in few hours watching your videos than spending 1 year of power courses in the university :)
Thank you Bgdn!
yhanks brother.
Thank you Joo for the comment. I am glad that you have found these video useful.
Best regards Robert
Your videos are highly useful especially the lab videos give a feel for the subject. thanks for the crystal clear explanations!
hi i watched all the parts, amazed and learned a lot thank you for knowledgeable videos
Very nice tutorial! Thanks!
Thanks! Robert
Thanks for this video
I realized the values shown on the schematic in this video are different than that of what was calculated in the previous video for some reason.
Hi,
Thanks for the guidance and detailed video tutorial.I have a question regarding ZVS..As we use snubbers to remove the ringings that are caused due to Primary Leakage inductance and MOSFET output capacitance and that is clearly explained in the video. How can we achieve complete ZVS by increasing the reflected back voltage(N*Vo) that we say valleys to achieve complete ZVS.Do we need to add Inductance Lr in series with Lm.The high offset voltage makes the mosfet to switch at lowest drain voltage and results in reduced switch ON losses.
Regards
Dear Robert very nice video. Do you mind telling what kind of output capacitor you used for the 100kHz switching? Can we use electrolytic or do we need some special capacitor?
Hi Anonymous, yes you can use electrolytics. However, electrolytic tend to leak over time. I typically try to use tantalum capacitor or if you want maximum reliability, then I would use stack ceramic caps. That is what we use for space rated power supplyies. Hope that helps.
Thx for a wonderful video Robert. I have one question , actually i am trying to implement a flyback inverter. When i provide 20kHz switching waveform to my primary side switch , via an optocoupler the Vgs is a triangular waveform instead of a square waveform. The output of optocoupler is fine if not connected with the switch. What should i do in this case ?
Please answer
Hadeed Sher Typically, you cannot drive a mosfet with an optocoupler. The optocoupler would need to be followed by a CMOS driver such as the MIC4422. The capacitance of mosfet are large anywhere from 1nF to 5nf. To keep the switching loosed low, you want to swing fast and have a rise and fall times around 50ns. To do this, you need a device that can provide large amount of sourcing and sinking capabilities. i=(5nF/50ns)*12V=1.2A. So to drive a mosfet with Cgs of 5nf to 12V in 50ns requires 1.2A of current. I hope this helps.
🙏🙏😭😭😭😭
I use simplis to simulation, the result is that VDS peak to peak is about 174V and ringing frequency is 10MHz, so maybe simplis is more acurrate
this was a lot helpful thank you. How do I know the power rating of the startup resistor of the control IC I've the right ohmic value but it heats up so I can not operate it for long. I've also tried higher resistance value but the converter didn't startup.
Sorry Gaber, I missed your question. Do you have a schematic? You could use a Zener and a resistor to start the IC. Across the resistor, you could add a capacitor in parallel to the resistor so that when you start up the Vinput, the capacitor will supply a short charge current that might be enough to start the PWM. I hope this helps. You can email me at rbola35618@aol.com
Hi Mr Bolanos. I've designed a 12 to 24 flyback converter but the spike voltage on the drain doesn't change with the snubber circuit. I changed the resistor from 220 ohm to 47k but spike changed from 59 volts to 65 volts. Why doesn't it change the drain spike? Can you help me solve this? Thank you
Hi, Robert, It is very helpful of your video. I have on question, what is the cap with 16pF? In the video, you said it is parasitic cap, but i don't know where does it comes from. Hope it can be clarified. THANK YOU SO MUCH!
Hi Yu, That is a good question. The 16pF is the measured self resonant capacitance of the primary winding. The winding are wound on the bobbins and since the wires lay next to each other, they form a parasitic capacitors. So in the model , I included the parasitic capacitance in the model. Also keep in mind that you can only use the inductor up to the self resonance of the winding. Fresonace= 1(2*pi*SQRT(Lprimary*Cparasitic)). In closing, the transformer should be used below Fresonace. I hope this make sense. If not, you can email at rbola35618@aol.com
Thank you Robert.
Hi Robert , I have designed PUSH pull converter having same serious problem of this ringing which will cause Mosfet to blow off(instantaneous exploading ) as it exceeds the max voltage capacity of that , How to avoid that issue ? The only solution for this is Snubber ? How this Leakage inductance can be avoided ?
Hey Robert, I've designed a similar design on my bench except its 12W. My Vin(min) was 25V, and I set the duty cycle to 0.4. The transformer has been design correctly as I've confirmed with my RCL meter. My circuit works very well in simulation. But when I turn my circuit on, it's working for a second but then current slowly starts to build up (like thermal runaway) and destroys the Mosfet and maybe even the output rectifier diode. And they're both well over rated for this power/current level. It's all open loop for now. What could be causing this thermal runaway ?
Please send schematic, scope plots and any other information to rbola35618@aol.com
Hi Robert,
What about the low frequency ringing before the Mosfet turns on? How can that be reduced?
This is a good question. You use the same procedure to snub the low frequency that I used to snub the high frequency. Keep in mind that the high frequency is Fhi = 1/(2*pie*sqrt(Leakage_inductance*Cmosfet)). The low frequency is therefore Flow = 1/(2*pie*sqrt(Lprimary_inductance*Cmosfet)). Since the Lprimary >> Leaksge_inductacne, the Flow will have a much lower frequency. Therefore, to snub the low frequency ringing, you place a second RC in parallel to the primary of the flyback transformer. The Csnubber = Cmosfet, while the Rsnubber =sqrt(Lprimary/Cmosfet). If you have any more questions, you can email me at rbola35618@aol.com
Look for video on leakage inductance
Then after parallel RC， peak to peak voltage of VDS is 86.1V
here you have used D=0.46, but then how you got 12 V. shouldn't it be lesser around 10 V as per the voltage transfer ratio??
It is very useful. I have some doubt..1. We calculated Snubber Cap using Ringing period and total inductance across primary. Shouldn't we add leakage inductance as well during calculation? 2. In general Snubber circuit should be connect across MOSFET. Why we are connecting this across primary? It might create problem of resonance? (I think we are adding Cap and Res in series to avoid this but it will add cost)
3. Why u not added resistor across primary? Can we avoid if it is too small?
4. Snubber circuit design what u explained is common for all topology of SMPS. Can I used the same for Full Bridge DC to DC converter as well.
SAMEER GARG Hi Sameer, excellent questions. There are two ringing frequencies. One is the leakage inductance(higher frequency) and the second is with the primary inductance (lower frequency). The leakage inductance is what usually gives you the large leading spike which can give problems if the MOSFET voltage is not large enough (Vds rating). So to damp this leakage spike, you only use the leakage inductance. Some people chose to also damp the ringing associated with the primary inductance. If you notice that the primary inductance will produce a much smaller frequency. That is because the Lprimary is much greater that L_Leakage. So when calculated a snubbber for the Lprimary ringing, you can add the L_Leakage to the Lprimary but it is so small that it is not worth adding it. I hope this makes sense
Hi Robert, may I know where to buy those pins?
Year after - I would like to know the same. Thx
Hi sir how to design pi filter for flyback converter
Can this circuit be tested in open loop, just to observe the ringing waveforms?
Yes it can be tested open loop so that you can see the drain voltages and primary current.
I agree but this is low power so I can get away with it. Most of the leakage comes from the transformer itself as evidence by the coupling coefficient that I calculated of 0.9898. Typically, we can get it to 0.9998. The leakage inductance from the pin and leads are probably about 20 nH per inch of wire so I probably have about 80 nH in the primary. All transformers have leakage inductance and lowering will increase the ringing frequency and lower the characteristic impedance of the flyback.
i could not find a video on flyback snubbers design, can any body help plz
Sudhhi India look at video 17. Spice files are available at rbola35618@aol.com
You have a strange breadboard, can you show me where to buy it?
i have a problem i measure lk = 70 uH and lp = 175 uH so K = 0.77 is this very bad?
Yes the leakage to very high
how can i fix it?@@RobertBolanos
Generally, you want to fill the layers to 80 to 90 percent converage. In other others, lets say you primay on the 1st layer is 5 turns of 1 strand of wire, and if would amounts to 40 percent of the first layer, then what you can do is wind it a two strands to that it will cover 80 percent of the 1st layer on the bobbin. You do the same with the rest. Please details of your transfomrer calculator and any other information to rbola35618@aol.com
sir excellent video but i have designed a fly back but i am getting ringing frequency more,long length ringing, sir i want to take help from you.if possible,could you inform your email id?
www.danyk.cz/reg60v_en.html
Sir visit the link I make this forward converter I face sum problem like igbt heat and burning problem please help me 😔😔