Help! Can someone clarify to me how this converter circuit works?

[CandleCookie]

Actually, I become more confuse with this circuit. Can someone help to explain to me? Thank u

 

[BadN]

Looks like flyback converter for me
Capacitor C1 and C2 is to double the input voltage Vs to send to primary transformer
D1 and D2 is to rectify the output signal that come out from secondary transformer
Lo and Co is low past filter, so the clean DC voltage would be obtained at Vo

 

[Tahmid]

Hi,
This is a half-bridge converter, not flyback.

The 2 input diodes along with C1 and C2 act as a voltage doubler.

Q1 and Q2 are switches - BJTs/MOSFETs/IGBTs - that are switched at high frequencies. Q1 must be driven as high-side switch with isolated gate/base drive, driven with respect to the source/emitter. Q2 must be driven as low-side switch driven with respect to ground.

Cs (the capacitor between the switches and transformer) is there to make sure that no DC comes to the transformer and there is only AC flowing through, it acts as a blocking capacitor to block AC.

The output diodes are ultra fast recovery diodes (maybe Schottky) that convert the high frequency AC to DC by full-wave rectification.

Lo and Co are there to produce a clean DC voltage acting as a filter.

Hope this helps.
Tahmid.

 

[CandleCookie]

May I know how the MOSFET switches during the circuit operation. Is it both switches switching on and off alternately?

 

[Tahmid]

Hi,
Yes, they are switched alternately at a specific frequency. Be sure to provide deadtime to avoid cross-conduction between the high-side and low-side MOSFETs.

Hope this helps.
Tahmid.

 

[E-design]

As mentioned it is a classic half-bridge design. Here is some detailed info.
https://www.fairchildsemi.com/an/AN/AN-4153.pdf

 

[CandleCookie]

Thanks. Then, for the mosfet switch Q1 and Q2, do I need to add additional diodes to the MOSFETS? Now, I'm trying to find a suitable transformer for my circuit. The problem is that I can't find a centre tapped transformer to use in this type of switching. Do you all have any suggestion?

 

[Tahmid]

Hi,
It doesn't require additional diodes as the MOSFETs have drain-source antiparallel body diodes. If you can't find a centre-tapped transformer, you can use a non-centre-tapped transformer and use 4 diodes in bridge rectifier setup instead of 2 diodes. However, if the current is very high, this may not be reasonable, as you'd require 4 large diodes.

Hope this helps.
Tahmid.

 

[CandleCookie]

Thank you. I nearly forgotten that I can use additional 2diodes. The current I'm using should be around 1.5A. Is it alright?
For the transformer, I do not know how to use it as in there is no datasheet about it's legs. Can you help me?
**broken link removed**
Is this suitable?

 

[Tahmid]

Yea, 1.5A is no problem.
The transformer you showed - it's just the core and bobbin. You have to wind the transformer on that. As for pins, wind them onto any pin you think should be used.

Hope this helps.
Tahmid.

 

[CandleCookie]

Thank you!! So, if I want a mid point transformer, I'll just need to wind them into two section and wind them to 3pins.
Yea. I couldn't find readily made transformer for smps so I plan to wind it myself. Is it just wind it will do?? Do you have any idea?

 

[Tahmid]

So, if I want a mid point transformer, I'll just need to wind them into two section and wind them to 3pins.

Yes. You might want to use bifilar winding so that you get the windings right.

Is it just wind it will do??

I don't quite get what you mean.

Hope this helps.
Tahmid.

 

[ahsanfarooq]

Just to give u an idea, for the core size and quality u r using u should wind the transformer using 4 Volts per turn at a frequency of 30KHz. For example for an input voltage of 24V u can wind the primary winding with 6 turns. And wind the secondary on the basis of simple ratio of input and output voltages. And for secondary if for example u calculate 10 turns for your required voltage u have to wind 10+10 turns. And if ur output current is 1.5 Amp select the wire gauge for secondary which can hold 1Amp. As the load will be equally divided on two windings.
Other tips for winding include: Use proper insulation between the layers of winding as it is very important for high frequency operation. Use Kepton insulation if possible. Winding should be done preferably in the center of the bobbin and the sides should be avoided. Winding should be done in a symmetry as non symmetric winding gives non symmetric waveforms on positive and negative sides.

Hope u find it helpful. I have done it many times for a push pull converter with a 42mm core transformer using 6volts per turn and using 30KHz frequency.

 

[CandleCookie]

Thanks Tahmid. Actually I'm asking is it winding will do? Do I need any necessary equipments?

Thanks ahsanfarooq. But how u know that 1 turn is 4V and how u determine the frequency by winding? Do u have any formulas for me? Which type of core is more preferable? I need a midpoint secondary. My input is 240Vrms and output is 17Vrms.

 

[Tahmid]

Hi,
Yes winding is mostly what's needed. Be sure to use some tape or other material between primary and secondary windings. You may use tape between each winding layer but this isn't an absolute necessity.

The volts/turn is not strict. You can choose one yourself, making sure you stay within the maximum flux density limit. 4V/turn is okay. There are formulae and I will put a few up.

 

[CandleCookie]

Thanks Tahmid. Yea, I need formulas to design the transformer. I'm not sure how to decide 4V/turn and the frequency for the transformer. Is it the frequency of the transformer due to the mosfet switching frequency?

 

[Tahmid]

Hi,
These are taken from Marty Brown's "Power Supply Cookbook":
Npri = (Vinnom * 10^8) / (4 *f * Bmax * Ac)
Nsec = [1.1(Vout + Vfwd)*Npri] / (Vinmin * DCmax)

f - frequency in Hz
Bmax - flux density in gauss (webers/cm^2)
Ac - effective core cross sectional area (cm^2)
Vfwd - diode forward voltage in V
Npri - primary no. of turns
Nsec - secondary no. of turns
DCMax - maximum expected duty cycle (0-1), if you don't have a reference, 0.95 is good

Hope this helps.
Tahmid.

 

[ahsanfarooq]

No equipment is required for winding a small transformer like this, u can easily do it by hands.
I have managed to calculate the volts/turn experimentally, i made a power supply from 170-240V rms to 24V DC 20A regulated. for that i spent a lot of time and finally got it 100% perfect. I made about 50 pieces all r working without any complaint. I took help from a computer power supply. For 240V rms capacitor chrges above 300V, for that all computer supplies have 38 turns on primary side and i used the same number. For secondry i used 11+11 turns for 24V DC. For 17 volts 10+10 will be fine if u r going to add feedback control for regulated voltage. If u reduce the number of turns u can get 17V output but then input range will not be as low as 170 V. You can even increase the turns in secondry and get required voltage by PWM. My circuit is almost the same technically except that i used full wave ractifier from mains source. And the Center tap of Q1 and Q2 goes directly to winding and other end of winding has the capacitor (Cs). So in short i will recommend 38 turns for primary for ur circuit. And one more point; wind 19 turns of primary first, then the secondry winding, and then 19 more turns. means secondry winding sandwitched bw the primary. But if cant do it as a beginner simple winding will do.

---------- Post added at 23:56 ---------- Previous post was at 23:52 ----------

And by the way winding a transformer is really as simple as i explained, there is no rocket science involved so "just wind it will do"..............

 

[CandleCookie]

Hi, Tahmid. Thanks for the formula. May I know what is Vinnom, Vout(in rms?), Vinmin? and the frequency is the input voltage frequency or the switching frequency for the MOSFET? By the way, earlier you mentioned that the capacitor is to block the DC. Do you have any formulas for getting the capacitor value? I did that in the simulation but it does not work. It only works well when i put in inductor series and a capacitor parallel with the transformer. For the LC filter, is there any formula that I can used to get the values? I just test with some values and it works well but i would like to know how is the formula.


Hi, ahsanfarooq. So, you mean i should try out experimentally to figure out volt/turn? It's ok with that. But since it is a SMPS transformer, the frequency for the AC voltage is 50Hz. How should I test the transformer? I need to do the MOSFET switching first and test with the transformer or I just input the AC voltage with high frequency?

My lecturer gave me a sample of transformer. It is the same type as shown in the website.
**broken link removed**
The winding should be done at the centre. The problem is how to differentiate primary and secondary?

 

[Tahmid]

Hi,
Vinnom is the nominal DC voltage (Nominal Line AC Voltage * 1.414).
Vinmin is the minimum DC voltage (Minimum Line AC Voltage * 1.414).
Frequency is the switching frequency.

For capacitor, I'm not familiar with any formulae. Use 1uF capacitor. That should work with no problems.
No. You don't need inductor there, but place capacitor in series, not parallel.

You can roughly use any Volts/Turn and wind a transformer or you can use the above mentioned formulae. 4V/turn is a good starting point.

Hope this helps.
Tahmid.