ML4800 problem

[qolpa]

hi,

i am using ML4800(PFC and PWM IC) for a battery charger circuit with PFC and topology is forward converter with a reset winding. something like in the below link.

http://rbsfm.ej.am/Research/images/forward.gif

i made the calculations with respect to application notes and datasheet. when i test my circuit with small voltages all the waveforms look allright but D3(in the picture above) draws some current and get hot. and when i start working with high voltage, then ML4800 couldnt get start because of PWM side of IC get shorted with ground internally. i think it could not drive pwm mosfet.

i could not solve the problem for a long time and looking for some advise.

any help will be appreciated. thanks in advance

[pickit2]

As your post is not showing the full circuit so how do you extect a full answer, but having worked on this type of converter in the past, it looks like the anode of d3 should be on the + side of switch.
Last project we worked on we used tinyswitch from Power Integrations.

[tAhm1D]

Hi Qolpa,
In forward topology, transformer winding and output inductor should be properly constructed, otherwise the circuit will behave otherwise. Please check those,specially inductance of the inductor.
Setting of the d3 diode in your circuit is probably alright as it is for flux resetting purpose in each cycle when the mosfet is turned off. Please check the ratings of the diode d3.

[qolpa]

hi guys, thanks for replies.

i was using UF5408 for D3, when i changed it with HER208 the heating problem solved. the circuit is an universal input(80-270 Vac) 24V-10A batterycharger. PFC output of the system is 380 Vdc. i am testing the system with a 30 Vdc bus and UF5408 was drawing ~80 mA while HER208 only a few mA. i will look for the details but the 2 diode looks not much different: HER208(tr:70 ns, Vf:1,85V , 1000V-2A), UF5408(tr:75ns, Vf:1,7V, 1000V-3A).   anyway, this is ok now.

but when i  start giving an ac (220V indeed)  signal to the input, pwm drive side of IC(pin11) will shortcircuited with the ground(pin10). i have a dead ml4800 army now  
i saw the output was rising then falling to zero at the start so the pwm mosfet was driven somewhat maybe just 1-2 cycle. i could not observe the pwm side of the circuit yet, i am planning to do it today. pls find the pwm side schema in the attachment, i cant give the whole system because this will be a commercial product, sorry for that. but if you need more info please ask me.

i will calculate the drive current of the mosfet. this IC can deliver 1A i think. i will change the drive resistor with a bigger value and maybe test it with another mosfet maybe.
i will write after tests.

regards,

qolpa

Posted on: July 16, 2009, 08:43:26 08:43 - Automerged
 i forgot to tell, sorry, this circuit was working correctly 1-2 weeks ago. i tested about 3-4 different pcbs for a long time, i burnt them all
there shouldnt be a design problem and that drives me crazy
i am suspicious about the transformer but it seems right on tests, i will take a closer look to it later.

[qolpa]

hi again,
i tested the circuit  without pwm side  and it works well.

i changed the mosfet(with a different one), gate resistor(10=>22) and transformer(with a same type but 6 weeks older one), nothing changed.

any ideas will be greatly appreciated.thanx

[tAhm1D]

Hi,
From the attached schematic, I think that there may be a problem with your startup circuit. Instead of this, connect a linear transformer output followed by diode/capacitor/regulator and supply voltage to Vcc and then check.

[qolpa]

hi,

you mean i have to use regulator not zener diode and resistors.

i will test it tomorrow. thanx

[qolpa]

hi,
i could not find a proper regulator and i used a dc power supply for vcc and pwm mosfet blew up in the test
i am looking for details.

regards

[tAhm1D]

hi,
Place a 1k resistor in between mosfet gate and source. Hopefully the mosfet wont blow up then, as the mosfets might be blowing up due to accidental turn-on by noise,or for drain to gate mosfet miller capacitance etc.

[qolpa]

hi,

i didnt solve that problem, and  i think its not the power, its the feedback. the IC makes the pwm duty cycle %100, but it is not possible, because in datasheet it says it can only work % 0-49.

i tried another pcb today and it works without any problem. i dont want to fight with that board anymore

thanks for help

[tAhm1D]

Hi,
Please enumerate the difference between the new working pcb circuit and old nonworking pcb circuit, so that we can learn from your mistake.

[qolpa]

hi,
i will search it on monday or tuesday.

ragards.

[qolpa]

hi,
i noticed something today. i am using TLP421 as opto for feedback and it oscillates in the start (no load condition). there are some load at the output but it isnt enough i think. it can seen normal. when i load the board the oscillation goes and circuit works stable. that is ok.
when i use PC123 no oscillation at start, always in stable condition. that can be ok.my real problem is when i use TLP421 the current limit is 8A and when i use PC123 it is 10A but the output voltage is exactly the same (27.6 V). the hardware is also same.
i use 2 feedback:voltage and current. (see attachmnt, the primer side's schema is attached in the first post)
TLP421 and PC123 datasheets are in the attachment, too.
i wonder why current limit changes but output voltage is not changing.

and one other problem is the circuit is not working when there is  some load before start(it starts with 2A load but not with 3A or bigger) , but i just noticed that and still working on it 

regards

[qolpa]

hi,
i noticed something today. i am using TLP421 as opto for feedback and it oscillates in the start (no load condition).

and one other problem is the circuit is not working when there is  some load before start(it starts with 2A load but not with 3A or bigger) , but i just noticed that and still working on it  

regards

hi, i found the problem as a missing filter resistor i hadnt noticed yesterday. it is working without oscillation and it can start with full load.

but i still did not find the which difference(s) of PC123 and TLP421 cause changing of current limit value. i think it has to be the output side(transistor), and from datasheet i found only the Vce voltage difference(PC123:6V, TLP:7V)

is it possible the input(diode) side ratings cause this? forward voltage i mean.

regards