if I put a lower resistance the bulb will on....short circuit...
R1 establishes a current loop for the self oscilating flyback converter, it needs to be the exact vallue in ohms, so .22 of an ohm ( 0,22 Ohm ), the power rating could be other vallue as long as it is not a lower one.
is it because D5 = FR107 and I replace it with 1N4007 ...
looks like it must be fast recovery one ?
You cannot use a nomal diode for RCD snubber on a high frequency switch mode power supply, 1N4007 is way too slow to do the job, you can use UF4007 in stead or any other ultrafast diode with a voltage rating of at least 600.
You really should stop experimenting with stuff wich you do not understand, you could cause a great deal of damage of even worst you could receive an electric shock ( wich could kill you!!!!! )! Ask before trying something, and in situations like this when it is clear you do not understand the converter and the risks involved with playing with it, i would recommend you go to someone that knows what he's doing.
Thanks for the suggestion, I carefully handle the circuit, I know, there's a high voltage area, and I clear the capacitor always before handle it...
if I ask someone else doing it, I will not learn....there's no point on that...
Why don't I buy a new one then ??
I want to learn that's why I'm doing it...
Experiment is part of learning from me...of course with caution...
I wear a glove and measure the voltage first before play with the PCB...
If there's no voltage at all, is it possible that my coupler is gone ? or the triode TL431 ?
and I reckon 1N4007 will not be able to do the job...
...if I ask someone else doing it, I will not learn....there's no point on that...
I understand and maybe even patially agree, but learning cannot come only from experiments, there has to be a good theoretical base, it allows you to look at things from another perspective, and also it prevents you from making basic ( and often costly ) mistakes.
The converter only seams simple from the component count point of view, but that's just an illusion, as a matter of fact this converter involves complex math and calculations in the design part, it is a fragile balance between component vallues and working parameters.
...If there's no voltage at all, is it possible that my coupler is gone ? or the triode TL431 ?
I can't remember the last time i found a deffective opto on a PC supply unit, but you can easily check it removed from the board, use a DMM set on the diode function, check the led side of it, it shoud read about 1V drop in one way, and open circuit in the other, the transistor part should read open circuit because it is like measuring a npn bjt on CE. If you need to be sure than get a 1,5V battery, a limmiting resistor ( let say 100 Ohms ) , and power the led side of the opto with it, than conect the digital multimeter ( still set at the diode function ) at the transistor side, what you are doing is powering the led, so the phototransistor inside will activate, the dmm should read some tens or hundreds of milivolts.
TL431 is not a triode, it is a thermal compensated voltage regulator, it has a 2,5V refference, and an error amplifier, all in one TO92 case. I guess you could call it an adjustable zener.
I understand and maybe even patially agree, but learning cannot come only from experiments, there has to be a good theoretical base, it allows you to look at things from another perspective, and also it prevents you from making basic ( and often costly ) mistakes.
The converter only seams simple from the component count point of view, but that's just an illusion, as a matter of fact this converter involves complex math and calculations in the design part, it is a fragile balance between component vallues and working parameters.
I can't remember the last time i found a deffective opto on a PC supply unit, but you can easily check it removed from the board, use a DMM set on the diode function, check the led side of it, it shoud read about 1V drop in one way, and open circuit in the other, the transistor part should read open circuit because it is like measuring a npn bjt on CE. If you need to be sure than get a 1,5V battery, a limmiting resistor ( let say 100 Ohms ) , and power the led side of the opto with it, than conect the digital multimeter ( still set at the diode function ) at the transistor side, what you are doing is powering the led, so the phototransistor inside will activate, the dmm should read some tens or hundreds of milivolts.
TL431 is not a triode, it is a thermal compensated voltage regulator, it has a 2,5V refference, and an error amplifier, all in one TO92 case. I guess you could call it an adjustable zener.
It most definitely cannot! In fact it could blow up!
Of what bulb are you talking about?
Anyway, you must have that 0,22 ohm resistor for the converter to work propperly.
bulb is a short tester (acting as a fuse )...it's working now, but the voltage is not right,
I use V5J diode and sanken optocoupler from old TV board..
it has a weird voltage, could it be because of 0.22 ohm ?
I use 22 ohm now...that's what I can find on my table for experiment...
You still do not understand, it is absolutely mandatory to have exactly 0,22 Ohm resistor there, ( i said this more than once but you have ignored me ) maybe 0,33 Ohm could work, but never, and i say again, never 22 Ohms. Try and read carefully, and keep in mind what other people say, do not ignore their advices.
I do not know of what diode you are using now but i sai again ( and for the last time ), you need an ultrafast rectifier diode with a voltage rating of at least 600V, t here are loads of choices, try to search datasheets for the diodes you think you have and can use, before you use them.
PS: Inserting a bulb in place of the fuse is not 100% guaranteed protection, the large primary filter caps charge with enough energy to blow the transistors by them selves in case of a fault. Charge a 330-470uF capacitor to 150V, then try to discharge it with a 13007 transistor, directly and without any current limmiting, do you think the transistor will hold?
You still do not understand, it is absolutely mandatory to have exactly 0,22 Ohm resistor there, ( i said this more than once but you have ignored me ) maybe 0,33 Ohm could work, but never, and i say again, never 22 Ohms. Try and read carefully, and keep in mind what other people say, do not ignore their advices.
I do not know of what diode you are using now but i sai again ( and for the last time ), you need an ultrafast rectifier diode with a voltage rating of at least 600V, t here are loads of choices, try to search datasheets for the diodes you think you have and can use, before you use them.
PS: Inserting a bulb in place of the fuse is not 100% guaranteed protection, the large primary filter caps charge with enough energy to blow the transistors by them selves in case of a fault. Charge a 330-470uF capacitor to 150V, then try to discharge it with a 13007 transistor, directly and without any current limmiting, do you think the transistor will hold?
In that case, I need to wait for the diode and 0.22 ohm resistor to come
and the capacitor is 22uF/400V
I am sorry if i seemed to come hard on you, but you are experimenting with what is in fact a complex converter, first of all it is a flyback converter, that in it's self says plenty, the flyback converter involves complex calculation in the design part, everything depends on everything for good working order, i could fill more than a couple a pages with just the transformer calculations. Self-oscillating flyback converter like the one you are working to fix is even more complex, you do not have the luxury of an controller IC wich provides a stable frequency and a proper drive waveform, in stead you have a bunch of passive and active discrete components, each with it's own shortcomings, each with it's own possible variation of it's parameters, all together working to provide the needed drive for the power transistor, any deviation of one parameter directly affects the converter.
Let me put this in another perspective, let say you have a 3 inch water pipeline, for some reason a section of it breaks under pressure, and you need to fix it. You replace the damaged section with a 1 inch pipe. Tell me, is that a good fix? Would you say the line would still perform as it should?
Replace the resistor with a correct vallue one ( 0,22 Ohm ), search for a good replacement for FR107, and as i have sayd, there are lots of different ones that could do the job just fine, you just need to make the correct choice. If you have doubts then post here all you got and i will say if anything can be used, otherwise go and buy a new one ( UF4007 should be easy enough to find, i have loads of them ).
I have enough with complex calculation in uni time.....
I will waste my time calculating and not fixing at all finally, the same what I have done in university, wasting my time learning all those gauss law for nothing...only a headache and finally I never use it and can not earn money from it...if you know what I mean...
ok...the value must be exact....because it's generating a certain waveform that needed to get exact 12V and 5v....I don't really care about the calculation there, I'm not the designer of the circuit anyway....
I only want to fix it and that's it....
Well if you would have paid enough attention to the learning process in the university you would have understand what i am saying, in fact you would have most probably avoided at least some mistakes... Anyway, it is not my intention to belittle you, but just to point out important stuff.
This converter is as i said flyback, and basically they invest energy in the transformer core ( rather in the air gap ) when the power mosfet is switched ON, and delivers that energy in the form of power for the load when the mosfet is switched OFF, it also is current controlled, meaning that when the current through the power transistor reaches some calculated peak level, the power transistor switches off. Now, the current through the power transistor goes through 0,22 resistor as well, and that resistor is read by the small NPN transistor ( the voltage drop on it ), when the peak current reaches a vallue large enough to drive the base of the NPN, it activates and shuts the mosfet off. Now comes the important part to your situation, because if the bjt activates too soon, the mosfet power transistor also goes off too soon, so enough energy is not invested into the transformer core, and without enough energy it cannot reach the correct voltage to the output. Your 22 Ohms resistor does that, remember that a silicon bjt requires about 0,65V into the base to activate, that means about 3A peak current through the mosfet before it switches off, with the 0,22 ohm resistor. With your 22 Ohm instead, you only have 30mA peak current through the mosfet, that is way too low to allow the core enough energy. So as you see, theory is very important to achieve a decent level in electronics of any sorts.
Is it more clear now why that 0,22 vallue is so important?
Simple Ohms laws, 0.22 Ohms and 22 Ohms, lower resistance = more current flow, less Vdrops. If you look at the diagram, the Vdrops on 0.22 Ohms resistor is used for Biasing Q2 (OPTO U2 is also Biasing Q2) which in turn affect the Biasing of the power MOSFET.
Q2, Q3 is set up not exactly as a latch but as for PWM to control the drive the GATE of the MOSFET.
http://www.eetimes.com/author.asp?se...doc_id=1279006 http://www.explainthatstuff.com/how-...tors-work.html
In normal operating condition the Vdrops on the 0.22 Ohms is not enough to turn on Q2, but if the MOSFET draws too much current the the Vdrops the 0.22 Ohms will he high enough to bias Q2.
I.E. If the current through the MOSFET is 1A, the Vdrops on 0.22 Ohms will be 0.22V, not enough to turn on Q2, but Q2 is also getting the bias from the OPTO which is used for feedback to regulate the output Voltages.
C4, R5, and the Aux winding pin 3, 4 of the transformer provide the positive feedback for the circuit to oscillate.
Im not sure what you are doing with testing the opto coupler.
Look at my diagram again please, the output should show a resistance around 100 to 500 ohms (depending on dvm used).
The 1.5v is the input from a AA or similar battery.
Comment