LM3488 ISSUE help.

Not sure what your Vin and Vout spec are but consider using the LM5122 Sync boost controller.  You can also run this in a multi-phase configuration.  Hope this helps?

David Baba.

I've tried putting the current down to 3.8A but it still does not show the LM5122, any idea's why?

webench.ti.com/.../webench5.cgi;VinMin=18&VinMax=30&O1V=50&O1I=3.8&op_TA=30

Sorry just realized...also insert 20V for the vin min.....

thanks.

When I try to change the advanced options to 333Khz it removes LM5122 as an option, any ideas why?

Hi Liam,

WEBENCH frequency restriction is to make sure the power dissipation of the design is reasonable and all selected components of the design NOT to exceed the Max. rated temperature. Please note that WEBENCH uses only Theta JA values.   There is a way to bypass this check by reducing the ambient temp. from default of 30 deg. C to 0 deg.C or -40 deg. C.   Let us know if this helps.

Hi Howard,

Yes that brings it up to 240KHz, I currently have the LM3481 set-up as per the WeBench circuit (after changing to from the LM3488 originally) link here

It works fine up to 70W but when i start pushing it harder then that, the output fluctuates and there is a lot of hissing at all voltage inputs. Any idea's why this would be happening?

The layout is probably not ideal as it has been done on development board by hand, i tried to copy the recommended layout for this device. What do you mean by bypassed the thermal check? With the LM3481 i kept all my conditions as per what I intend for real life conditions. It was only the LM5122 i bypassed by putting in -40 degress. I see you do the AN-1756 which I could use and modify with my desired parts. Is this something that is ready available? I will buy one to modify if I can be confident that this design will work under the higher loads. 

Hi Liam,

Yes, the EVM is available and it can be order using the following link.

According to WEBENCH, the board needs to dissipate an excess of 5+ Watts of power. Please do give it a try and let us know. I would suggest using the WEBENCH design with the lowest operating frequency so you get the highest efficiency ( Min. power loss).

Hi Howard,

The controller runs ok now at 150W i have running at the moment. The only issue is the efficiency nose dives at the higher switching frequency. Is this because of the losses in the diode are bringing it down because it is on for longer ?

Hi Liam,

If the BOM is not optimized for higher frequency, you will see an increase of power loss.   Especially, the switching losses will be more pronounced at the IC and the FET.  You can mitigate the switching losses by using a FET with very low Qg.  I have listed few FETs to try:

CSD18537NQ5A: Qg=14 nC, Rdson = 13mohm

BSC100N06LS3 G: Qg = 15nC, Rdson = 10mohm

BSC340N08NS3 G: Qg = 6.8nC, Rdson = 34mohm

Suggestion of a more efficient diode with lowest Vf  in our database:

STPS20M100SG-TR, Vf= 0.455V

Hi Howard,

Thanks for you support, why would a lower gate charge help the efficiency? I thought that the Rds(on) would have significant effect on the efficiency? I was also looking at the 

SBRT25U80SLP diode do you think this would suffice?

Hi Liam,

In your boost circuit, the most significant power loss is the diode. I offered couple different FETs and hoping to minimize your switching loss or you can reduce your switching frequency down to 100 or even 50kHz due to the high power nature of your design.