The "SXB" Laser Driver 2.1A (Boost)

[lazeerer]

I... don't agree. On every single datasheet I have ever seen for an IC with both an SO package and a FN package, datasheet clearly states that the FN package has higher thermal conductivity, even when they both have thermal pads. It's because the larger ones have more PLASTIC to go through to get to ambient. But it's your call. :beer:

Why would it have to go through more plastic.? That doesn't make sense. Its a thermal pad. Its straight copper.:san:

Saying that a smaller chip is better is like saying a 1/2 inch block of aluminum for a heatsink is better then using a 1 inch block if both have the same contact surface to the heat source.

If both chips have thermal pads the larger package is just an extra bonus. The part that matters is the thermal pad. the larger chip has more mass even though its plastic or whatever its made out of. It also gives it better react time to heat.

Thats what i have noticed at least. I would take a larger package any day over a smaller package.IMO

 

[Wolfman29]

Just saying what the datasheets say

Here's a link to an example datasheet. http://www.ti.com/lit/ds/symlink/tps55340.pdf

If you go down to the thermal information, you'll notice that, in all cases but the junction to top thermal conductivity (and to ambient, but hardly), the QFN raises in temperature by a smaller amount per watt of heat. That means there is a smaller temperature difference between the junction and the bottom pad or whatever per watt of heat. Correct me if I am wrong, but I would choose the FN package here over the TSSOP.

 

[Hiemal]

Lazeerer, I sent you a PM a bit ago about an idea, just wondering if you got it or not...

 

[lazeerer]

Just saying what the datasheets say

Here's a link to an example datasheet. http://www.ti.com/lit/ds/symlink/tps55340.pdf

If you go down to the thermal information, you'll notice that, in all cases but the junction to top thermal conductivity (and to ambient, but hardly), the QFN raises in temperature by a smaller amount per watt of heat. That means there is a smaller temperature difference between the junction and the bottom pad or whatever per watt of heat. Correct me if I am wrong, but I would choose the FN package here over the TSSOP.

I dont know how else to put it. Keeping a 3mm size chip stable with heat with no aditinal heatsink is much harder to do. A larger package can take heat change much better

Lazeerer, I sent you a PM a bit ago about an idea, just wondering if you got it or not...

Whats up buddy. Iam just trying to catch up since i got back. Ill jump ahead to it now. Ive been trying to go through the pms as fast as i can.:wave: Sorry about my slow response.

 

[benmwv]

Nice driver! Looks like you've got that scope figured out

That thing should be very stable with temperature. You have a solid ground plane on the bottom and a heatsink pad. I've become a very big fan of the heatsink pad/soldered on heatsink. I try to put it in all my boards now.

Also, in my experience there isn't much (if any) practical difference in the performance of different packages if they have a thermal pad.

 

[lazeerer]

Nice driver! Looks like you've got that scope figured out

That thing should be very stable with temperature. You have a solid ground plane on the bottom and a heatsink pad. I've become a very big fan of the heatsink pad/soldered on heatsink. I try to put it in all my boards now.

Also, in my experience there isn't much (if any) practical difference in the performance of different packages if they have a thermal pad.

This is the one i was tell you about.

ha ha. I took your advice and just started pushing buttons. j/k I dont know how to read it yet though.

Yes I have to agree with you on the thermal pads. On all the board i make now that dont have the +/- pads on the bottom i add thermal pads. In fact i took it a step further with a custom version of the X-Boost.

I left some of the bottom Copper Exposed on one of the version X-Boosts and i re flowed it with solder directly to the brass pill or i would re flow it straight to a piece of coper that i custom turned on my lathe.

I dont think it can get any better then that.LOL I was just having some fun.

Then on this 20mm version i did a heatsink pad on the top and on the bottom so you can thermal adhesive it straight to a piece of metal and if you want put a piece of coper like i did above.:

Then on this 24mm board i did exposed on the top like you see and had the bottom exposed under the chip. Having that top exposedf really helps when you push it into a pill.

Ive got more but my came died.:Errrr

I always put an Exposed pad if i can for heatsinking now under the chip area.

I also got some driver made in aluminium. TBA... Expencive as hell.!


As for again about the different packages. Ill have to disagree. Without no additional heatsink but the board a TO220 Lm1117 will be much better in handling heat then a TO-252. Both have thermal pads Right.? But for sure the TO220 handles heat much better.

Your theory might be true if they are both on large enough boards where it doesn't matter as the heat has enough room to spread out for either package. But when your talking about such small boards as laser drivers a Larger package chip is Superior hands down in my book when it comes to heat. If that was not the case i would not have gone with a larger package. I would have gone with one of the other 2 package that where available.

I guarantee you on a 3mm QFN chip say running at 2A currant with say a 5v load on say a 17mm board drop it from a steady 2.8A currant pull and quickly bring it to 4A which should be ~ 3.5v on your supply. See how fast that chip freaks out before it adj if it even adj and doesn't go into protection mode. Larger chips can take that shock change of heat and again handle heat overall better. Scouts Honer.:can:

Iam sticking to it. Ive made enough QFN driver to if i had the option to use a larger package i would.

Again it might be true what your saying if both chips are on large enough boards where it doesn't matter but that is not the case here as our boards are small and get hot supper quick and QFN chips dont like that.

PS. I finished your host.

 

[skyray]

Can't wait to buy a few, nice work

 

[Thejoker301]

Wow Looking good :beer:

Also ... I have sent you a message about buying a round x-boost ... please contact me as soon as you can :beer:

+10000 :gj:

 

[Ghostchrome]

Wow, looks great! Nice job.

 

[lazeerer]

Can't wait to buy a few, nice work

Thanks. Ill Let everyone know when ill have them available. I promise.

Wow Looking good :beer:

Also ... I have sent you a message about buying a round x-boost ... please
contact me as soon as you can :beer:

+10000 :gj:

Must of Missed it.:banghead: Ill check it Right away.:wave:

Wow, looks great! Nice job.

Thanks. Where you been at buddy. Have not seen you in some time.:beer:

 

[Ghostchrome]

Thanks. Where you been at buddy. Have not seen you in some time.:beer:

Oooooh here and there.

Actually I got a job that was pretty time intensive, long commute, etc, but all done with that now. I'm back! :beer:

 

[KRNAZNBOY]

WHAT SIze are these, 17mm or 16.8mm, if its 17, could anyone point me towards the cheapest 16.8mm driver that can deliver 1.8a +/- .1a.

THANKS!:gh:

 

[scumbagatheist]

Man, I am so glad you decided to release these! I just love your drivers. I am looking forward to adding this to my collection.

 

[kobi g]

Looks great,
When do you start selling it?

 

[Moe Ron]

Greetings,

One thing I have been wanting to ask Angelos and anyone else that designs and builds their own drivers, why don't we include modes like strobe, sos and such ?? Being I have no clue it seems it should just be a little more on the IC and imho would add to the quality and make an additional selling point.....

Moe

 

[benmwv]

Most people don't want a bunch of nonsense modes like they have on China flashlights. They can be cool, but most of the time they are just impractical.

Also It takes a lot more board space and adds a lot to the price to put modes on them. The regulator chips don't just have modes. You have to add a microcontroller to the board (along with its support circuitry) to have modes and you also have to program each one of those tiny microcontrollers in some kind of special rig since there isn't room for any programming headers on a driver.