small soldering job

If it's a prototype I'd hand solder it anyway.

It's not tricky when you've had some practice.

Don't bother with solder paste ( it ' goes off ' fast anyway ) - just use ultra thin cored solder ( I think mine is 0.8mm ) and a very fine tip. Pin down the corners of each chip carefully to locate it, then attack the other pins.

You *will* get 'bridging' between pins - just use good solder wick to remove the surplus, I recommend Chemtronics 'SODER WICK' - best wick I've ever come across.

Maybe I'll post an example of what you can do this way at a.b.s.e if you're interested ?

Graham

Pace make a tip with a small well to hold a blob of solder, several others have copied it, you just run it down the row of pins. the trick is to have the right amount of flux on the pins. Practice! start with the packages with the widest pin spacing and work down as you get the technique.

Do it yourself with flux gel. A wide iron, 1.5mm, with a drop of tin does 5 pins at once and the boiling flux gel removes the bridges. If you etch yourself, do yourself a favour and make the footprint pads at least 100 mils long. This would allow to wipe the tin away from the chip.

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net

Metcal makes a cartridge with a mini hoof tip for drag soldering. It works very well, even on parts with 0.5 mm lead spacing.

Leon

One such company is

I haven't used them myself.

Like the other posters I agree that it is quite feasible to solder it yourself.

A good amount of flux is the thing that did the trick for me.

For around £5 I bought a bottle of "Future 315 Low Residue Flux"

I used a syringe to flood the (tinned) pads before placing the 80 PIN TQFP. Soldered two diagonally opposite pads before doing the rest. (This was the most difficult step, to get the alignment exactly right!) My technique was to load the soldering iron tip with a very small amount of solder and then to touch it to the pad and pin with just a little pressure. The flux has the effect of making the joint pretty much instantly. The flux tended to evaporate quickly, so I just syringed more. I used an Antex M12 soldering iron, with a very small tip that I had filed down a little to make a very fine point. A magnifying glass from a set of "helping hands" was necessary to make the pins clearly visible.

As I recall I only had one solder bridge, easily removed with solder wick. Another popular technique is the "flood it with solder and then remove it with wick" approach, which I haven't tried.

The only pins that didn't have a connection were I couple that I had omitted to solder!

Paul.

I've had some success with first wetting a corner pad with a bit of solder. Then, while using the soldering pencil to keep that solder melted, hold the component with tweezers, place it in the desired location, and finally remove the soldering pencil.

That works very well on SOIC16 and smaller. I haven't tried soldering anything with a finer pitch than that, but I don't know why it wouldn't work.

I have been tempted to try the "toaster oven reflow" technique, but the high price of solder paste has prevented me. (I suppose it's cheaper in larger quantities, but last time I looked it was going to cost something like US$65 to purchase the smallest quantity I could find, including the syringe and the required overnight shipping.) Hand-soldering has proved very easy, though.

I looked into the solder paste "toaster over" scenario. I found

(search for "solder paste") stocks it in a syringe for around =A39, that's around US$16 Perhaps try at

SOIC16??? That's HUGE, I do those quite happily with the big tip iron and a pair of tweezers. It's the little buggers, 0.5mm pitch say, and the ones without any noticeable lead at all (the ones with a metal slab on the bottom too, forget the name) that make life harder. Not to mention BGA....

Paul Burke

Yes, the ones with the metal on the bottom (TI's PowerPad, for instance) are what make me think that sooner or later I'm going to have to figure out the toaster oven method.

For now, I'm getting along fine with SOIC, SOT-23, etc.; all of the components I need to use are available in those form factors (though not DIP any more).

Hot air works okay for them too, at least for a few prototypes. I have not tried a QFN44 or 28 yet, but they don't look too scary.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

I've mastered the FN(?) package, though it's fiddly. When laying out, put a hole in the base pad, large enough to get the tip of a fine bit through. Flux everything. Solder blob one corner pin, and anchor the component in place with that. Solder the other pins (fiddly). Then introduce solder through the underside hole gently, so it acts to conduct heat to the bottom pad. When it's ready, it wicks in, and all is OK. I've only tried it with quite small components, the LTC2601 being the smallest, but it works fine for them.

Paul Burke

I'm assuming we're all talking about the same thing, chips with a heat-sink pad in the center that is intended to be soldered directly to a pad on the component side of the PCB, that in turn is connected with thermal vias to more copper on the other side of the PCB. The problem being that the pad is between PCB and chip, and therefore not accessible.

How do you get solder to the hidden pad? And, are you applying the hot air from the component side or the other side?

I like Paul's idea of leaving a hole big enough to get the solder in from the bottom.

I've used a modification of Paul's idea with good success with the Linear Technology LT1976 which is a 16-lead TSSOP with an exposed thermal pad. Attached is an image of the board layoug we used. The large grey rings are the two vias that are used to solder the thermal pad and the smaller ones are the vias that connect the top and bottom copper pours that provide the heat sink. We found that having two large vias made soldering easier since you can put the iron tip in one so that it touches both the thermal pad on the bottom of the IC and the board copper and then feed solder into the other large via. When things get hot enough the solder melts and is pulled between the IC's thermal pad and the exposed SMD thermal pad on the board. It also makes it easier to see when you have enough solder since the level in the holes starts to rise when the entire space between the exposed SMD pad and the thermal pad is filled with solder.

--
James T. White


begin 666 Test3B.png
MB5!.1PT*&@H````-24A$4@```D(```+1" (````X-UNZ````"7!(67,``+B-
M``"XC0$>L5G+```@`$E$051XG.W=/;+CN-*@8=R)6BK\'5[*OH71+^Q#ZVXG8OCCI;V5WPP-"5#
M1JZ_V%-1&FE,UHHT-A$V8]\\-@Z^62W\\BMJ"_RK*EO]9U%FE,&K=XV+)E4-5^
M;;OOF\'G?$B;07V79V%_]\'#Y,830F:^EHS/\\^. BYQE7?JO:C=>[U!^>5\\JWQW%;L(R\'
M)?O+>QW-]X]S8C0&*7Z^>_4SEO>/4&(U!
MR50=);D^:63?1>C(TE^,QJ" T1@$I>\\JK%.\'CAGNU\\=03%3>_NIOS=_^QH )
MI#\'(ZJ^,@_5&_BT`8^N-KE5[&:Q/(HQI\'&S!G;$"\'>Z!;PIF-;< _$5_^/X1!I#%;MA237H_5
MV,/$V\\[XEBAC8(N-\\67SFR*HC9D3;_;CG\'/-ZQD9U?WUL S_&1G^`Z*XI@.!
M+BX2\\=4XY^;&E]N6"R&!T9@A03VY_=I6]\\HU@QSF@N=RW:OV9N4Y3],#LHWO
M;?H5N*Q@6I##G!]?;C*^>(Z:>?_8^PTR/C]OK
MU^N$+Z7E;9RO>U="+YOE;9!![4L\'Q:+X>C[_>D5\\?7??E[]E+,!HS*I&Y9_D
M]C:VEP[+YAQ/(L%B2X/%0\'!A#+4Q$\\:FW5_S]=ZDQU3[M[QLD9L[\\C
MC$@\\W^Z9DY;%5^.J:QGQ=3:,QFQYS+^/S=?7SM6NNE;]L^0\'\\_C?$NM=E-^_
MT\\TK7$VN3=:_6M8/Z6?V8Z.@CK?UH
M7?.:Q-_]YGOIG\\ZE!"L,XJ47QY=+Q1=,8C16@GI)^W 2?\\=[JT33##D,O&^H69SKE^+?/K_U[3>VR_YO&#63]!9V%EWOA[
MZ#=JC,10OWD,VF,QC18B*6,[>>:)-FWC%3O
M_0;V8.0\\S]X.(8S&!/4#E^Z

"Walter Harley" a écrit dans le message de news: snipped-for-privacy@speakeasy.net...

heat-sink

is

air

The pad in itself won't do much for power dissipation. You have to connect it through a bunch of vias to some plane that'll spread heat across the PCB more efficiently.

As Paul said, make one or more vias bigger so that you can solder the package through them.

--
Thanks,
Fred.