PLC810PG
Hello, I'm using PLC810PG to implement a battery charger with PFC function. The schematic of primary side is very similar to yours apps RD-189 and DER-212.
I did the first 300 pieces and I had some negative feedback from customers with 4 pieces with failure. The problem is the resistor in RD-189 ref R42 10ohm 1/4W in series at the VCC of the high side driver of PLC810. In all the 4 pieces the resistor is broken and measure 20Kohm instead of 10ohm.
Well some useful indication:
1. I changed the resistor but the resistor smoking and this time also the diode in series D8 UF4004. I changed the resistor and the PLC810 and everything is going on.
My question is: How is possible to break that resistor?
Some consideration:
At the startup and everytime the regulator goes in burst mode, the first two or tree clock are necessary to charge the boostrap cap 1uF and in that moment if the cap is discharged a high in rush current of 15V/10ohm is across the resistor. May this current destroy 1206 5% smd resistor?
Another consideration is about the power supply at pin 13 of PLC810. The max absolute voltage at that pin in the datasheet is 15V. But using a linear regulator with zener 15V 5% and a transistor doesn't guarantee this. In fact because the 5% tolerance and at high temperature the power supply at that pin can go higher than 15V. In some condition I measured 15.4V too. So also in all your apps you use this configuration. Is it right to do like that? May this is the problem about R42?.
In the datasheet you wrote that is suggested to put a 5ohm resistor to improve immunity. Which is really the function of this resistor and why in your apps you use 10ohm instead of 5ohm? Which are negative consideration to put 0ohm resistor? In all high side driver I used I had never use this resistor.
Regards,
Daniele Baldacci
Comments
thanks for your answer Chris, but the strange think is that the first time the diode is not demaged (the diode I use in not the UF4004 but RS1J). Only when I replace the resistor and I switch on again the diodo is broken. And in this case to repair the circuits I have to replace 10 ohm resistor, diode RS1J and PLC810.
However thanks for your suggestions. I will try to replace 10 ohm 1206 with MELF resistor and RS1J with UF4007.
hmm, this is really strange. I think the PLC810 fails because of the reverse voltage if the diode fails. but why should first resistor broke and after repair the diode ?? this make no sense.
you may consider the layout. is there any cross-talk ?
What FET do you use ? Heavy gate charge ? What powerlevel you have ?
I still have some trouble in the PFC stage.
regards,
Chris
Yes is really strange, but this is what happen. Four pieces had this failure. At the first problem I checked like you suggested the diode, but it was good. So I replaced resistor and I try. At no load the regulator worked fine, but at full load (300W) the resistor began to smoke and in this case also the diode is broken.
At the second battery charger I replaced the resistor and the PLC810 and in this case the battery charger work fine and it works still now.
Suppose the problem is not the resistor but that the high side driver. In this case the resistor is demaged before the diode.
Why the high side driver fails (the mosfet is ok)? Maybe because the VCC (15V) of the PLC810 goes higher the 15V? The VCC come from a linear regulator with classic zener 15V 5% and npn transistor. In some conditions (high temperatura > 50°) the Vz of zener is higher than 15V (till to 15.9V I measured) and the vbe of transistor 0.6V in not sufficient to bring the Voltage below 15V. May the PLC810 fail with voltage of 15.3 - 15.4V?
My app is a battery charger 24V 10A about 300W at maximum power. At the beginning yes I had some problems with crosstalk overall at the start up with high current. This was because at no load the switching freq is 210Kh and at the start-up the parassitic inductance of the PCB generated crosstalk and the result was that the mosfet fails. This happened more frequently if I used bigger mosfet (with more silicon but with great gate cap). So I reduced the max freq at 150Khz
and from this point I never had problem till now.
What kind of problem do you have on PFC stage, because in my app it has always worked fine?
hi, yes may be its a problem with VCC. We use a 13V zener with the classic NPN.
Try using a supressor diode on VCC ? Or some more of blocking C ?
Did you already try the UF4007 ?
I started a new thread for my problem. you find it "PLC810PG PFC stage".
Unfortunately nobody answerd yet.
We had bad efficiency at full input voltage and some noise comming from the boost inductor. You can read the tread, maybe you have any ideas or hints.
If you like you can send me you Schematic and Layout. I will take a look.
vollmer@syncrotec.de
regards
chris
Hello Daniele,
i had same problem with D8 and the 10 Ohm resistor.
The Uf4004 has a to low blocking voltage.
if the upper FET is fired the diode has to block a dc voltage
of Vdcbus ( around 325 Volt ) plus the 15 Volts VCC.
So on some spike or over voltage events this is to narrow to the max voltage rating of the diode.
I replaced it with the UF4007.
For the resistor we use a MELF resistor with 0,4 watt rating.
But I think the 1206 should also work. The MELF might be a little more rugged.
For us that was the solution and now it works fine.
Regards
Chris