Help needed with soldering fine pitch part
Yen Pham
ruining 1 board,
and several ICs, I decided to seek professionals' advice -- yours! :-) Has
anyone successfully
done this? If so, please help me out. If you respond direct to the group,
as you should, can
you please copy me on the email.
Thanks,
Yen Pham
David Kessner
> I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand! After
> ruining 1 board,
> and several ICs, I decided to seek professionals' advice -- yours! :-) Has
> anyone successfully done this?
I do this on a regular basis. I'll give you the quick answer, and if youstill
have questions then you can ask...
First, the pads must be as flat as possible. This is not a problem on a new
board, but if you're doing rework then you will probably have to clean
the pads with some solder wick. If you don't do this, then it will be next
to impossible to keep the pins aligned on the pads.
Second, you must not bend the pins of the chip. Avoid bending at all
cost. Once a pin is bent, soldering the chip becomes 100 times more
difficult. Of course, since the pins are so small even touching them will
bend them. I use a suction cup "chip picker-uper" to avoid touching
the pins.
Use liquid flux-- and lots of it. Just about any flux will do, just be
sure to clean it off when you're done since many fluxes are conductive.
Start by cleaning the pads. Then put the chip on the PCB and carefully
position it until it is perfectly positioned. Put flux on the pins. Put
solder on
the iron. Tack down two pins (on opposite corners) by touching the iron
to the pads. Don't touch the pins, just the pads. Solder will wick up and
contact the pins. If you do this right, the chip will remain perfectly
positioned. This step is very important. If the chip is not perfectly
positioned then the remainder of the job will be very difficult. Take your
time. This step takes practice to get the hang of it.
Next, solder the rest of the pads. I don't solder one pad at a time, instead
I gently rake the iron up and down one side of the chip. When I'm done,
there is usually a solder bridge at one end. This is cleaned up quickly with
a clean iron and some flux. Don't use solder wick to get rid of the bridge,
since this will only bend the pins.
The soldering iron tip that I use is a 1/4 inch wide, flat "chisel" tip. When
I
put it on a chip, it can cover about 20+ pins at a time. This is more of
a personal preference than anything else, but I find that it's harder to
bend or damage the pins with this tip. With this, I can solder one side
of a TQFP in about 15 seconds, including cleaning up any bridges.
Sometimes if the soldering isn't going well, some fresh flux will make it work.
With practice, you can solder TQFP's, TSOP's, PQFP's, and TSSOP's
very quickly and with no damage to the PCB or the chips. By now, I
can solder them well enough that you can't tell the difference between
my soldering and an assembly machine's.
Hope that helps!
David Kessner
dav...@peakaudio.com
Michael Gladu
>
> I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand! After
> ruining 1 board,
> and several ICs, I decided to seek professionals' advice -- yours! :-) Has
> anyone successfully
> as you should, can
> you please copy me on the email.
>
> Thanks,
>
> Yen Pham
Yen,
I have hot air tools used in the soldering of SMD packages. How many do
you need soldered on how many boards ??? I could solder them for you if
are interested.
Michael Gladu
.........................mgladu@warwick.net.............................
Richard Gaupsas
tool. The person who did it just used a big old soldering gun (witnessed
not by me but my brother) and solder paste. He did a very clean and neat
job. My brother could not believe it could be done with such tools. So
it's not the tools but the skill of the user.
Gavin Melville
wrote:
>I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand! After
>ruining 1 board,
>and several ICs, I decided to seek
Yen,
It can be done by hand, but isn't fun. I routinely solder 240 pin
0.5mm PQFP's and the yield has been 100% so far. Last year there was
a long thread in this group about this, but it may be gone from Deja
News by now.
The trick is in the flux -- I use RS rework flux in a spray can.
1. Make sure the IC and the board is clean.
2. Spray the board with a light even layer of flux.
3. Just as the flux is changing from tacky to hard stick the IC to
the flux, correctly positioned.
4. Solder the corner pins. I use solder with 2% silver -- it "wets"
better.
5. Check the positioning.
6 Spray the IC pins with flux.
7. The next step is hard to describe, but the "trick" is to get a
bead or ball of solder (SN62 again) to run down all the pins on one
side. You don't need much solder, and hold the board up near
vertical. Start at the top of the row of pins and work down. This
bit takes practice. Some warm the baord, some don't. I don't think
it made much difference.
8. Clean the excess solder off the last few pins with solder wick.
Don't remove too much solder.
For 0.5mm IC's this is probably faster and easier than hot air,
although the purists may disagree. Occasionally 2 or 3 pins will be
shorted -- use solder wick and check carefully. One soldering iron
manufacturer (PACE I think) makes a tip for this -- I use a Weller PTK
or similar conical tip. Don't try this in a draft -- ie. a fan
nearby -- the solder temperature is important.
--
Gavin Melville
ga...@cypher.co.nz
Eugene Nine
Eugene
David Kessner wrote in message <362E329D...@peakaudio.com>...
>Yen Pham wrote:
>
>> I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand!
After
>> ruining 1 board,
>> and several ICs, I decided to seek professionals' advice -- yours! :-)
Has
>> anyone successfully done this?
>
ed kennedy
Without knowing what solder equipment you have i.e temp control and tip
size, I will start by suggesting a micro tip and very small gauge RA or RMA
flux core solder. You really shouldn't do this by hand but if you must get
some liquid flux also, RA or RMA at least 15% solids.
I have done this type of work using a microscope before (X 40), it can be
done. Don't forget to clean the flux off right after soldering is complete.
The trick is to apply the liquid flux to the part being soldered and the
pads, also pick up a very very small amount of solder on the iron each time.
If you do this type of work often check out the PACE mini wave attachment.
(I don't work for PACE)
Good luck
Ed
You are on your own with the initial alignment of the part, don't do this
with a hangover! Ha.
Yen Pham wrote in message <70l9tj$q64$1...@client2.news.psi.net>...
>I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand! After
>ruining 1 board,
>and several ICs, I decided to seek professionals' advice -- yours! :-) Has
>anyone successfully
Sam Goldwasser
> It can be done by hand, but isn't fun. I routinely solder 240 pin
> 0.5mm PQFP's and the yield has been 100% so far. Last year there was
> a long thread in this group about this, but it may be gone from Deja
> News by now.
Dejanews should archive forever but you may need to check their 'old'
database I think.
Checking there FIRST really is the most polite way of trying to find basic
information of this sort. Although, I supposed, soldering a .5 mm PQFP
probably isn't in the really basic class :-).
--- sam | Sci.Electronics.Repair FAQ Home Page: http://www.repairfaq.org/
Repair | Main Table of Contents: http://www.repairfaq.org/REPAIR/
+Lasers | Mirror Info: http://www.repairfaq.org/REPAIR/F_mirror.html
| Sam's Laser FAQ: http://www.misty.com/~don/lasersam.html
Austin Franklin
Thank you, that was a most excellent 'tutorial', and your time is greatly
appreciated!
Austin Franklin
dark...@ix.netcom.com
David Kessner
> so how do you make a board with pads that small at home?
> Eugene
I don't. I do this stuff for a living, so the PCB's are made professionally
but I often end up building the first couple of prototypes myself.
David Kessner
dav...@peakaudio.com
Tom Bruhns
: so how do you make a board with pads that small at home?
: Eugene
It turns out to not be difficult, actually. The basic way I do it, down
to 8 mil lines/ 8 mil spaces (roughly) is to print the pattern I want to
etch onto overhead transparency material with an inkjet printer. I'm
using an HP PhotoSmart printer right now, which is a 300dpi printer, so
8 mil lines are 2 to 3 dots wide. 600, 720 or 1440 dpi printers
_should_ be able to do even better, but be very careful about actual dot
size and density in the transparency. If I need finer geometries (e.g.
5 mil lines/ 5 mil spaces), I can get the artwork photoplotted locally.
(Note: 0.5mm lead pitch is 19.7 mils, so pads are spaced 5.9 inkjet
dots apart. This makes it difficult to do a really good job on a large
pattern; there's an "aliasing" effect. My current work is actually with
a 0.65mm lead pitch part, 7.68 dots per lead-space. At 1440dpi, 0.5mm
lead pitch is 28.35 dots per lead-space, which should work better. I'd
like to know if the 1440dpi printers can really do a dense transparency,
though.) Then use the transparency, either printed or photoplotted, to
expose a pre-coated (with photoresist) board. I etch boards by
_floating_ them on ferric chloride. Surface tension holds them up, and
that position results in very clean edges. But...I do mostly analog
work where most of the circuit fits nicely on one side of the board, and
therefore doesn't require multiple layers to be practicle. Two-sided
boards are no trouble, though; I get good alignment so that a 30 mil via
drilled with a 13 mil (#80) drill works...put a 30AWG wire thru the hole
to complete the connection.
(Does anyone out there have the recipe for cupric chloride etch?
Some day I'd like to get away from ferric chloride, though it _is_
simple and available.)
--
Cheers,
Tom
tom_b...@hp.com
ed ngai
BUT Keep in mind, you are the person who is soldering the IC down and you MUST
pay attention to how much electrostatic charge you will be generating. ESD
will kill semiconductor devices. You really have to pay some attention to
all those yellow stickers that say 'Caution ESD Sensitive Devices, etc.'
Before you do anything, like soldering, get an ESD mat, an ESD wrist strap and
best an air ionizer. Get plenty of iso-propyl alchol to clean the PCB.
Carefully
inspect the pin of the TQFP, you have to use a microscope to see if any
corrosion
or oxidation is on the IC pins. Tin the IC pins. Don't spend alot of time
heating
up the IC pins and the pads. Alot of heat from the soldering iron can, and
will,
damage the IC. A datasheet will tell you how much time and at what temperature
you can subject the pins to, so don't over heat. Clean the PCB generously w/
IPA.
Apply the Rosin flux w/ a small paint brush, or get a small bottle w/ a 0.015"
needle tip to dispense the Rosin. If the Rosin clogs up the needle point, then
just
heat up the needle w/ the soldering iron to melt the rosin. Works all the time.
The key to SMD is not really how much solder you put down. The key is to put
just
enough solder on so that you can 'Un-Solder' the IC.
Jameel Akari
> It turns out to not be difficult, actually. The basic way I do it, down
> to 8 mil lines/ 8 mil spaces (roughly) is to print the pattern I want to
> etch onto overhead transparency material with an inkjet printer. I'm
Do you ever have problems with the patterns to being dense enough
to block the exposure light?
> therefore doesn't require multiple layers to be practicle. Two-sided
> boards are no trouble, though; I get good alignment so that a 30 mil via
How do you align the two sides? Corner guide holes?
> (Does anyone out there have the recipe for cupric chloride etch?
I don't know about that, but I made up an etch solution of
hydrogen peroxide and HCl that could etch a very heavy (2oz Cu?) double
sided board in 5 min or so. That was a board where I just used donuts and
a Sharpie marker to do the layout (simple power supply), and the ink held
up. I haven't tried it with photoresist yet. I don't have a real recipe
for this etch, but it's basically enough 3% drugstore peroxide to cover
the board and maybe 50ml of muriatic acid (~36% HCl). Cheaper than ferric
chloride, doesn't stain, and it can go down the drain when finished. only
problem is you can't really store or reuse it.
____________
Jameel Akari
Insert witty comment here
------------
<http://fribble.cie.rpi.edu/~jakari>
X.b...@concentric.net
>Yen Pham wrote:
>
>> I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand! After
>> ruining 1 board,
>> and several ICs, I decided to seek professionals' advice -- yours! :-) Has
>> anyone successfully done this?
>
>have questions then you can ask...
>
>Next, solder the rest of the pads. I don't solder one pad at a time, instead
>I gently rake the iron up and down one side of the chip. When I'm done,
>there is usually a solder bridge at one end. This is cleaned up quickly with
>a clean iron and some flux. Don't use solder wick to get rid of the bridge,
>since this will only bend the pins.
When I do this there is genraly a solder bridge between most
of the pins. Contrary to your advice, I use solder wick to clean up.
First I put another coat of Flux on the pins. Then I soak the
solder wick in flux. Then I lay the flux on the pins and
apply a well tinned and clean tip to the solder wick. Don't
apply pressure!!! Just touch gently. THe flux and tinning
will conduct the heat. I usualy get a very good final result.
>
>The soldering iron tip that I use is a 1/4 inch wide, flat "chisel" tip. When
>I
>put it on a chip, it can cover about 20+ pins at a time. This is more of
>a personal preference than anything else,
Yup, a great choice for a tip. It is also my favorite.
>Sometimes if the soldering isn't going well, some fresh flux will make it work.
I can't understate the need for lots of flux and a very clean tip.
>
>Hope that helps!
>
>David Kessner
>dav...@peakaudio.com
>
>
With additions and replies by....
Just my $0.02 worth.
Hope this helps,
Gordon
PS:
To reply: replace 'X.bleeb' with 'greeder'.
Belics
Bill Berzinskas
Yen Pham <yen...@earthlink.net> wrote in message
70l9tj$q64$1...@client2.news.psi.net...
>I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand! After
>ruining 1 board,
>and several ICs, I decided to seek professionals' advice -- yours! :-) Has
>anyone successfully
Gary Woods
>>I'm trying to solder a 100pin TQFP that has a 0.5 mm pitch, by hand! After
>>ruining 1 board,
>>and several ICs, I decided to seek professionals' advice -- yours! :-) Has
>>anyone successfully
I'm not familiar with the package, but there's a really good trick for high-density SMD stuff:
Tack the corner leads down, so the bug stays put.
Run a bead of solder across the tips of all the leads.
Use your vacuum-pump solder sucker to pull solder away from the chip body at the end of the leads.
When done, you should have a fillet of solder bonding each lead to its pad, and no bridges.
Works like a charm, and looks almost like factory work!
--
Gary Woods O- K2AHC Public key at www.albany.net/~gwoods, or get 0x1D64A93D via keyserver
gwo...@albany.net gwo...@wrgb.com
fingerprint = E2 6F 50 93 7B C7 F3 CA 1F 8B 3C C0 B0 28 68 0B
Anders Frederiksen
repairing). Alternatively you should use solder paste directly on the
pads before mounting the device. This method is much moer efficient when
mounting new devices. It is possible to drag away the solder-bridges
using a large tip and ordinary solder. Though this is method is very
fast - its also very hard to learn.
As a last thing I guess you've found out that "solder sucking wick" is
THE thing when working with SMD....
Regards,
Anders Frederiksen
David Kessner wrote:
>
> Eugene Nine wrote:
>
> > so how do you make a board with pads that small at home?
> > Eugene
>
R.Buckner
gull-wing) part down, clean the board well with Isopropyl Alcohol. Then
paint on a good deal of RMA liquid flux. Gently position the part on the
pads and apply some form of pressure to hold it in place. (We have used
wooden clothespins from time to time.) Align the soldering tip with the
first pin (parallel to pin and pad) and gently flow the solder (feeding only
enough to perpetuate a small wave in front of the iron) as you move the
tip down the row of pins. You do this with each side of the part until all
are
soldered. If you get this technique down (time/motion) it goes very quickly
and all you have to do is flood the area and clean with brush with copious
amounts of iso alcohol. Usually, there is very little bridging and the heel
and toe solder joint is perfect on each pin/pad.
cir...@flash.net
ba...@wristband.com.ph
R.Buckner wrote:
> The trick is to use lots of flux.
Really
> Before putting the 100 pin (I am assuming
> gull-wing) part down, clean the board well with Isopropyl Alcohol.
Yeah so you could oxidize the board and spend a lot of flux later.Actually all
you need is a good quality solder and let the flux from the solder do the work.
And you don't apply the solder directly on the pins but clean the pads with the
solder and the remaining solder will be enough to bond the pins later.
> Then
> paint on a good deal of RMA liquid flux. Gently position the part on the
> pads and apply some form of pressure to hold it in place.
I suggest soldering at least the opposite corners to hold the QFP in place. I've
done this a thousands of times with boards from AT&T, Compaq, etc. etc. Then
apply heat on the pads/the tip of the pins gently and let the heat melt the
solder and bond the the two together doing opposite sides on at a time.
> (We have used
> wooden clothespins from time to time.) Align the soldering tip with the
> first pin (parallel to pin and pad) and gently flow the solder (feeding only
> enough to perpetuate a small wave in front of the iron) as you move the
> tip down the row of pins. You do this with each side of the part until all
> are
> soldered.
As I have said only a good quality solder will do the trick an inch will do. No
wooden clothespins, alcohol.
> If you get this technique down (time/motion) it goes very quickly
> and all you have to do is flood the area and clean with brush with copious
> amounts of iso alcohol.
Gee, the board will be unsighty with those white stuff! You'll have to deal with
oxidation reliability problems later.
If you did a good job in the first place you don't need those alcohol,etc.
> Usually, there is very little bridging and the heel
> and toe solder joint is perfect on each pin/pad
R.Buckner
Well, what can I say, it is a large nation.
The method I described in as few words as possible (since this really isn't
the proper forum) is the method outlined in the training courses for Level
A,
Level B, and Inspector's qualification at:
(GSFC) Goddard Space Flight Center
Training Center
Code 300.1
Greenbelt, MD 20771
(301)-731-8628
and
(JPL) Jet Propulsion Laboratory
Training Center
MS83-205
4800 Oak Grove Drive
Pasadena, Ca. 91109
(818)-354-4165
and
Lockheed-Martin Engineering & Sciences
2400 Nasa Rd 1
Houston Texas 77058
Proficiency in this method of hand soldering surface mount chips
to circuit boards must be proved before you can build up anything
that is going to fly. These methods are deemed "life critical".
You think they know by now what works ?
Isopropyl Alcohol does not oxidize metallic surfaces.
If you don't apply heat to the pins then there is a percentage of your
solder joints that are "cold" soldered. (Not desireable for long term
use). Believe me, I've seen cross sectional analysis of this.
You must make sure that all pins are in positive contact with all pads.
(If you tack down the corners, you better be sure that all the pins are
perfectly planar and this does not take into account the irregular surface
of a plated or even Hot Air Leveled pad on a circuit board).
The white residue you speak of is certainly unacceptable and if there
could be any, then you haven't done your job correctly. The white
residue you speak of is the effect of Bromide, Chloride, or other
acid salts that remain from activated fluxes that either had a poor
specific gravity to begin with (not enough alcohol usually in conveyorized
wave soldering machines). It is not oxidation. If the residue is noted
on the solder joint (much later) then the problem is oxidation due to
a high SN (tin) content in the alloy and exposure to a corrosive agent
or environment. If the residue on the metallic area is blue or blackish,
then the alloy was too high in lead.
Thanks for supplying a short-cut, but I believe I will stick to a method
known to be so reliable as to bet human lives on it.