LM3488 help please!!

[bowman1710]

Hi,

I am currently running with the design as attached, but with a 7.6A gate driver in line with the MOSFET, it seems that when I am running this at 333KHz (my desired switching frequency) there is a loud hissing noise and the DR output is jittering al over the place. When I reduce this down to 283Khz its fine. Any idea's as to why?

 

[BradtheRad]

Your Iout appears to be the icon for a current source. Did you intend for this to be the load?

Did you try using a plain resistor for the load?

 

[rx8pilot]

Your issue is most likely in the PCB layout. The schematic is straight out of TI Webbench and they generally do a decent job. In SMPS design, the schematic is only the beginning. The layout is as important as the MOSFET.

Can you show and describe the PCB?

Also, why are you using an external gate driver? The 1.5A from the LM3488 is rather strong. The faster rise time on gate can cause ringing with only the smallest parasitic inductance. This will interfere with the control loop especially at higher frequency as the ringing will not have time to settle.

 

[bowman1710]

Hi Guys,

Brad the schematic is from WeBench from TI so that is how it is laid out. RX8, I thought my layout may not be ideal as it is in the prototype, when the system was in higher load situations the output was not reaching the full 50V output. I assumed it was to do with the MOSFET not receiving enough current drive, when i put a MOSFET driver in line the output was regulated. I spoke to the people from TI and they said that this ringing is inherent with this type of controller, which they is under investigation. So I will try a different controller I think.

 

[rx8pilot]

I have a design in production right now with he same controller and have very little ringing. It is not the greatest choice, but it should work FAR better than you describe. It still sounds like your layout is not right. The performance of these is VERY dependent on the layout. It took me a few tries to get it right and the first 2 revisions would ring so bad that I was never able to get full output. Now it is clean, efficient, and full power and I ONLY changed the layout - not the schematic.

If you do not understand the critical nature of PCB layout for SMPS circuits, a different controller will not help you. Post your layout as it is, its ok if it is just a prototype. Just don't change controllers and expect much of an improvement if the same PCB layout mistakes are made.

 

[Sruk]

@bowman1710
Have You had some luck with eliminating your problems with the 3488? I'm considering this IC for a new design so I would like to know how are the things going.

@rx8pilot
Great post!

 

[brushhead]

The hissing smacks of control instability. In my opinion you will need some R between the chip and the MOSFET.

 

[bowman1710]

Hi guys, update on this post is that I dropped the single phase and went for a multiphase boost which is all working now.

Have You had some luck with eliminating your problems with the 3488? I'm considering this IC for a new design so I would like to know how are the things going.

Sruk i would go for the LM3481 was much more stable when running at full load then the LM3488, dont know if they have addressed the instability issue they said they were looking into.

The hissing smacks of control instability. In my opinion you will need some R between the chip and the MOSFET.

Brushhead, I tried the usual tricks with trying to control it but no matter what I did I was just unstable, TI said that in some situation that part exhibits this behaviour and they were looking into it. I went over to the LM3481 and it worked without any issues managed to get a fairly reasonable efficiency from it too.