L6566B PWM CONTROLLER

[kapurk]

Is the L6566B from ST Micronics one of those badly designed chips which one should stay away from? Recently, I have spent several weeks of dedicated effort in trying to get an SMPS design based on this chip to work. I used the eDesign suite on ST Micronics' website to design the schematic and got some prototype boards made from a professional board maker. After several failed prototype boards I ordered ST's evaluation kit (EVL6566B-65W-QR) from Mouser. The evaluation board worked for all of 30 seconds and then stopped! At first I thought maybe one of the safety controls on the chip had been triggered but the board simply refused to work. When I checked the board in detail I found the L6566B was dead because pins 4 (Gate driver) and 5 (Vcc) had shorted internally to ground!
I am now wondering if it is really worth wasting any more time on this chip! I had posted on an online support query to ST more than 2 weeks back but there has been no reply except an auto-generated response assuring me that a "dedicated team of expert" is working on my problem.
I would like to hear from any members on this forum about their experience with the L6566B.

 

[Flyback]

It has quite a lot of satety features on it which can be tripped by mistake, I think with good layout it should be ok.....if layout is bad than you trip the overcurrent by mistake. Be wary of the internal startup regulator getting zapped...there is an onsemi app note on this which applies to l6566b. Add schottky from gate to ground (cathode to gate).

I used it for a 60w flyback (100w peak) for a diaphragm pump driver circuit some time back and it was ok, once the layout was improved.

 

[Tony Stewart]

Did you get the setup info for the I2C interface to program the CC/CV functions?

 

[kapurk]

Thanks for the response Flyback and Tony. My frustration comes from the fact that I am talking not just about my own pcb layout but also about the official Evaluation Kit from ST (cost me 140$ to get it from Mouser). The schematic for the EVL kit shows some additional circuitry to protect the Vcc pin but obviously that did not work. Incidentally, I run a small manufacturing unit making laptop chargers and we are currently using the age old workhorse UC3842 PWM controller which would probably work even if I soldered it upside down. After looking at a number of new PWM controllers from TI, PI and others I decided to go for the L6566B because it seems to offer a lot of extra features, at least on paper, for using it in a wide range of power supplies. However, given the experience so far, I am wondering if it is the right chip to bank on.

 

[Tony Stewart]

THe documents online I saw were too short to be useful for programming the serial flash parameters. We don't know what you have or your test setup, but I would request an RMA and tech support from Mouser and ST.

 

[Flyback]

I just had a look for a post on the HV supply pin of these kind of chips...I cant find it...its an onsemi app note, and it says when power supply is powered down...the bus falls low and the hv supply pin can ring down below ground and cause damage to l6566b and etc others.....to mitigate you put series resistor there...I cannot find the post, it is in this forum somewhere. near april or may 2015

 

[kapurk]

I found the post about the HV supply pin : https://www.electro-tech-online.com/threads/smps-pwm-controllers-self-destruct-at-power-down.144279/
It is part of the Onsemi NCP1217-D data sheet.

 

[Flyback]

nice one, I hope it helps you