Main Site Documentation

Soldering LQFP64-50 package


#1

Is there any secret people use to successfully solder the STM32F4? About half of my boards come out with solder bridges. I can cut down the size of the paste stencils, but I’m worried about open disconnects. Do you guys have any soldermask between pins on the package?


#2

I’ve never soldered one, but I think there definitely SHOULD be soldermask between the pads.


#3

No soldermask between pads on mine…
Is your stencil 100 microns or thinner?


#4

It’s a 5mil thick solder stencil (127 micron)


#5

I use 75 or 100 with good results


#6

Thanks for the advice. I’ll try a thinner stencil!


#7

Are you using a reflow oven? If so, what kind?


#8

Just a cheap Chinese desktop model we got off eBay for a few hundred bones. Are you suggesting there’s a problem with my reflow oven’s settings? I couldn’t imagine what that would have anything to do with.


#9

@ jay - What solder paste are you using? Leaded or lead free, No-Clean or watersoluble. A 5mil thick screen is the right thickness. The openings on the screen, are they 100% to the pads, if so reduce the opening size by 10%.


#10

Leaded solder (I can’t remember specifically what I’m using right now, but I think it’s 63/37). I think it’s no-clean flux in the paste.

The pads have been reduced by 10%.

I’ve never had any problems with soldering, and all the other parts on the board come out beautifully.

I’m using N-sized (nominal) landing pattern (not the high-density one).

I might try shrinking the stencil’s pads for that particular part a bit more.

Any other ideas?

I’m using a pretty hot profile (my stupid reflow oven doesn’t seem to solder all the boards properly if I use a cooler profile for the 63/37 solder), but intuitively, I’d think a hotter oven = better soldering.


#11

@ jay - i have been using cheap and cheerful 63/37 paste up until today.

I sprung for some 62/36/2% Tin/lead/silver fine pitch paste http://uk.rs-online.com/web/p/solder-pastes/1849985/

Glad i did, much easier to use and better results - there is an Octopus out of the cheap and cheerful oven…


#12

Looks great, Justin.

I can’t decide what I want to buy more… o-scope or oven.


#13

If your going to do a few then oven - don’t list to Brett about being a purist and using an iron - remember he’s an Aussie :whistle:


#14

Looks great! I’ll try some nicer paste.

You only need an oven if you’re building lots of boards or soldering chips with hidden pads (BGAs, chips with thermal pads, etc). A scope is way more useful.


#15

Sooooo you’re saying scope this month, oven next month? :slight_smile:


#16

I’d say get whichever you need more urgently. If your design works and you don’t need to debug anything then a scope isn’t going to give you anything except another toy to play with. If your design works and you need to assemble a bunch then an oven will be more useful. If your design doesn’t work and you need a scope to fix it then an oven is going to be useless…


#17

Haha, well, those two things are also in widely different price brackets. You can get a perfectly functional reflow oven for less than $300, but a decent oscilloscope costs quite a bit more than that.

EDIT: I guess I’m seeing quite a few used (but decent) 100 MHz DSOs on eBay for roughly the same price. Never mind! Get both!


#18

Actually, you can get a very good new Rigol 1052E for $329.

http://www.rigolna.com/products/digital-oscilloscopes/ds1000e/ds1052e/


#19

I’ve got a couple of those and 100Mhz version. A solid piece of equipment. Excelent build quality.


#20

in AU at least, the DS1052E isn’t being sold in numbers - the 100mhz variant DS1102E is only $80 more (that’s the genuine way to get 100mhz). Although my eyes have been more than a little taken by the DS2072 at a bit over double the DS1102 price ($923 vs $439 inc tax)