Author Topic: EEVblog #346 - MLF Reflow Soldering  (Read 32940 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37626
  • Country: au
    • EEVblog
EEVblog #346 - MLF Reflow Soldering
« on: September 07, 2012, 12:25:30 am »


Dave.
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7986
  • Country: gb
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #1 on: September 07, 2012, 12:32:35 am »
MLF. Measly Little F***er?
 

Online Psi

  • Super Contributor
  • ***
  • Posts: 9875
  • Country: nz
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #2 on: September 07, 2012, 12:58:44 am »
Yeah, i have to solder some 3x3mm DFN's every so often. Quite annoying.

You just get the hang of it when your doing the last one in the batch :(

I think i will expand the tracks out a bit on my next PCB, that would definitely help using the iron to tidy up pads.

If anyone knows where i can get a XY or XYZ accelerometer for under $2 that's not MLF/DFN or other stupid package.
Please let me know.
(Max order quantity would be 50 units)
« Last Edit: September 07, 2012, 01:06:31 am by Psi »
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline dda

  • Contributor
  • Posts: 46
  • Country: au
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #3 on: September 07, 2012, 02:42:18 am »
Why is the part moisture sensitive, and why is it sensitive before soldering, but not in the final product?
 

Online Psi

  • Super Contributor
  • ***
  • Posts: 9875
  • Country: nz
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #4 on: September 07, 2012, 02:58:37 am »
Why is the part moisture sensitive, and why is it sensitive before soldering, but not in the final product?

I believe the issue is that the plastic case will absorb moisture somewhat, then when you reflow that water boils off explosively and may damage the solder joint.

So once its soldered that moisture doesn't really matter.

But don't quote me on it.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline poorchava

  • Super Contributor
  • ***
  • Posts: 1672
  • Country: pl
  • Troll Cave Electronics!
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #5 on: September 07, 2012, 05:51:00 am »
Yup, that's the reason. It's called the popcorn effect. For more details on that I can suggest reading IPC/JEDEC J-STD-020 (Moisture/Reflow Sensitivity Classification of Nonhermetic Solid State Surface Mount Devices) and IPC/JEDEC J-STD-033 (Standard for Handling, Packing, Shipping and Use of Moisture/Reflow Sensitive Surface Mount Devices).

An yeah, those standards are boring for the most part, but also contain some very useful informations.
I love the smell of FR4 in the morning!
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #6 on: September 07, 2012, 06:06:05 am »
Usually any proper components supplier will tell you the MSL for a part < that msl determines how long it can be kept in air safely without the chip popcorning while soldering

I've seen deadly chip popcorning  ??? might even possibly blow the lid off ( chernobyl? ) or sometimes damage the chip itself

http://en.wikipedia.org/wiki/Moisture_Sensitivity_Level
usually BGA devices are the most susceptible
 

Offline andete

  • Contributor
  • Posts: 19
  • Country: be
  • famous for its killer edible poets
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #7 on: September 07, 2012, 06:06:27 am »
I somewhat doubt that the solder paste on the bottom re-flowed, unless you heated it longer then shown in the video.

I had to do a simular chip a while ago as seen in attachment. It was one without exposed pad though.

This is the technique I used to good success:

- add some solder flux wit flux pen
- place very carefully with a pincer until it is just right (carefully tapping, etc...)
- have solder nearby, floating, e.g. in a 3rd hand
- use right hand with pincer to apply pressure on IC to keep it in place
- use left hand (I'm left) to solder one side, not perfect yet
- remove pressure, turn board, add pressure again
- solder other side (again not perfect yet)
- now the IC should be stuck, you can work with two hands again
- I had once or twice that this was enough, but just to be sure I always
  do the next steps
- add flux to one side, then re-solder (re-flow) to get clean connects, visibly to the side of the leads
- same for other side
- if you like clean with some alcohol?
 

Online mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13678
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #8 on: September 07, 2012, 07:36:58 am »
It makes life a lot easier if you put vias in the centre pad :
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline hans

  • Super Contributor
  • ***
  • Posts: 1621
  • Country: nl
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #9 on: September 07, 2012, 07:40:21 am »
If I was Dave, I would heat it (this time off camera) again from the top for a little bit, just to make sure. Then again, it can damage the chip if done for too hot or long. But I guess if it's connected (I guess solder paste is conducting as well) it's good.

Anyway, maybe another technique if you don't have a hot air gun is shown is this vid (done by mikes electric stuff):
[I just wanted to post his vid]

I guess a MLF, DFN, QFN require similar techniques. Pre-tin the pads, have a hole through the board, add lots of flux, place the chip, heat it from the bottom and let it reflow that way.

I've soldered MSOP-8 with a heat-pad on the bottom succesfully this way, even with leadfree solder. It isn't the same as QFN, but it does reflow the bottom pad via a series of via's + a lot of heat from the bottom.
Watch out you don't apply too much heat, otherwise the FR4 gets dark  ::)
 

Offline nitro2k01

  • Frequent Contributor
  • **
  • Posts: 843
  • Country: 00
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #10 on: September 07, 2012, 10:33:59 am »
Mike, oh snap, that's clever. But I can see two problems. 1) reaching under the board with a soldering iron seems a bit tedious. 2) It seems like the final cleaning up might misalign or even release the chip, at least if doing it bottom side up. Or perhaps the surface tension will make sure that doesn't happen?
Whoa! How the hell did Dave know that Bob is my uncle? Amazing!
 

Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #11 on: September 07, 2012, 11:27:46 am »
That is why you use a Metcal or JBC iron. They have the thermal horsepower to melt the blob on the bottom before the heat even travels to the top.

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #12 on: September 07, 2012, 12:17:55 pm »
That is why you use a Metcal or JBC iron. They have the thermal horsepower to melt the blob on the bottom before the heat even travels to the top.

Well if you already have a hot air station and don't have the budget for a JBC, what's the point of using a iron?
After all chinese hot air stations are cheap


OH how i love Dave's channel so much



I was reading a previous forum post on QFN devices and it had a link to this fellows blog.
http://blogs.mentor.com/tom-hausherr/blog/2011/04/06/pcb-design-perfection-starts-in-the-cad-library-part-15-qfn/

He suggests creating the thermal pad as a grid of smaller pads. This is to help prevent the chip rotating out of position  when sitting atop a single large pad as the larger solder blob under the chip reflows.


Oh, mentor again. How i hate that expensive POS

Speaking of which,

« Last Edit: September 07, 2012, 12:33:45 pm by T4P »
 

Offline ondreji

  • Contributor
  • Posts: 43
  • Country: au
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #13 on: September 07, 2012, 12:31:05 pm »
It makes life a lot easier if you put vias in the centre pad :


Yup, or one big hole like on this board: http://www.schmartboard.com/index.asp?page=products_csp&id=484
 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8515
  • Country: us
    • SiliconValleyGarage
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #14 on: September 07, 2012, 03:00:10 pm »
a couple of additional pointers :

1) solder paste :

That stuff needs to be stored in a fridge for two reasons :
a) to avoid gravity from separating the solder out fromt he flux in the emulsion.
b) to prolong the life of the flux.
and if using water soluble there is a c : keep moisture out

Paste is to be stored in a closed jar , or the syringe is to be capped after use.

Not all solder pastes are made equal. Solder pastes in syringes are often the same formulation as the paste sold in jars for screen printing. that is NOT the kind of paste you want for this kind of hand job. Those are made avaialble for touch-up work and localized screen printing using mini-stencils. if yuo are going to do the 'squirt-n-dab route you need a paste that flows easily. Especially if you don't have a controlled application system that uses vacuum and air ( a 'calibrated thumb' clearly does not fall in this category )

Solderpastes are thixotropic in nature. shake or stir them and they become very fluid, so easy to apply. after a minute or so they return to a peanut-butter state. this is why , in industrial applications , the jar is opened , a scoop of paste is placed on the edge of the stencil , and the robot then goes through a program where it 'liquifies' by repeatedly scooping and scraping the paste. Personally i don't like the pre-filled syringes. i buy a jar of paste , scoop a dab out , stir it so it becomes pretty fluid , fill a bit in a syringe and when done squirt the rest back in the jar for reuse later.

the advantage of having the past in its liquid state is that you don't need to press so hard on the plunger and this gives you more control over the amount dispensed.

If you watch dave's video closely , a fraction of a second (time index 11:16) before he pulls the heat gun back you actually see the chip 'settle' on the board. You see that at that point all the solder ont he pads is in liquid state. That is the point you are looking for. This means hat ALL solder has now entered a liquid state and the chip is 'floating' on the solder. The surface tension and cohesion forces now 'align' the chip perfectly with the pads on the board.

One thing i noticed is that this board uses tin-flash as opposed to HASL. A couple of words of caution on that stuff:

Avoid that as the pest if you are going to hand assemble ! Shelf life is horrible. tin flashed or ilver flashed boards tarnish very quickly under open atmosphere. they need to remain in vacuum sealed packaging until ready to use. Get a greasy finger on them and you have soldering problems...
When you get these boards : run them all at once. keeping a few for assembly in a few months will give you only scrap boards unless you vacum pack them with dessicant.
Tin flash and silver flash do have superior flatness on par with ENIG or ENEPIG ( what is commonly called 'gold' or palladium-gold. The gold in these processes is only there to protect the nickel [nikcle tarnishes extremely bad ...] . during soldering the gold diffuses in the alloy and the solder joint is actually made to the nickel layer ) processes , are cheaper than the ENIG or ENEPIG but have very short shelf life.

You are by far better off having a good quality HASL board. A decent pcb fab has excellent control over the air-blade used in the HASL and the end result is almost as good as tin or silver flash. Hot-Air solder Leveling wors by dipping the board in flux and then dunking it in a vat f liquid solder and pulling it back out. a controlled blast of hot air (the 'air knife') where the board leaves the solder makes sure that solder is deposited eavenly. temperature control is very important in this process. the air blast needs to be slightly hotter than the tin temperature to ensure perfect release.

Now, here is how our 'soldering lady' ( we have a dedicated person in our lab to this kind of work. she's got all the IPC certs including Avionics and most of the MIL-SPEC's... ) places this kind of components.

- flip the chip upside down.
- squirt solder paste on the pads.
- heat with hot air gun so the solder 'beads' on the pads.
- inspect the solder bumps created and confirm all pads have an evenly spread of solder on them. there should be no area's that are devoid of solder.
- flux the board
- place the chip on the board
- heat with air gun and watch for the 'settling'

the advantage of this process is twofold
1)  You have visual confirmation that ALL pads do indeed have solder on them. if a pad was tarnished the solder will have flowed to an adjacent pad. this avoids having a pad with no , or not enough, solder on it. Solder 'bumps' should be equal in size and flow over the entire pad surface. You can correct this at this step.

2) There is no risk of tiny solder balls or unbound paste remains under the chip that could create shorts. When squirting paste between the board and chip and then heating you have no way , apart from an x-ray, the confirm that all solder went where it should have gone. By placing the paste no the chip and then reflowing it you can inspect the chip for 'stray' solder.
« Last Edit: September 07, 2012, 03:02:50 pm by free_electron »
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Online mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13678
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #15 on: September 07, 2012, 03:55:07 pm »
Quote
Well if you already have a hot air station and don't have the budget for a JBC, what's the point of using a iron?
After all chinese hot air stations are cheap
More of an issue for rework than assembly of a lone device - hot air can cause collateral damage, e.g. blowing small parts away and melting plastics like LED bodies, connectors etc.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Neilm

  • Super Contributor
  • ***
  • Posts: 1545
  • Country: gb
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #16 on: September 07, 2012, 03:59:43 pm »
Quote
Well if you already have a hot air station and don't have the budget for a JBC, what's the point of using a iron?
After all chinese hot air stations are cheap
More of an issue for rework than assembly of a lone device - hot air can cause collateral damage, e.g. blowing small parts away and melting plastics like LED bodies, connectors etc.

The only time I had to fit something like this was at work - I cheated and used the IR workstation we have.  :D

Neil
Two things are infinite: the universe and human stupidity; and I'm not sure about the the universe. - Albert Einstein
Tesla referral code https://ts.la/neil53539
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #17 on: September 07, 2012, 05:28:19 pm »
IR rework stations are great also, along with mini IR stations AND the hot air pens like the hakko 851, they are low flow and concentrated so can be a bit hard to cause collateral damage
 

Offline eldonb46

  • Newbie
  • Posts: 1
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #18 on: September 07, 2012, 06:04:00 pm »
Dave, isn’t? pin “one” marked on the PCB as the larger pad on the opposite corner?
 

Offline Short Circuit

  • Frequent Contributor
  • **
  • Posts: 439
  • Country: nl
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #19 on: September 07, 2012, 07:17:00 pm »
Why is the part moisture sensitive, and why is it sensitive before soldering, but not in the final product?
It is still moisture sensotive after soldering, meaning that, if you want to desolder a device from a board without damage, you must bake the complete PCB before the rework.
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7986
  • Country: gb
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #20 on: September 07, 2012, 07:37:03 pm »
Dave, isn’t? pin “one” marked on the PCB as the larger pad on the opposite corner?

No, that's a high current pad and trace, connecting to pin 7 (SW). Pin 1 is an NC pin.
 

Offline Agent24

  • Regular Contributor
  • *
  • Posts: 113
  • Country: nz
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #21 on: September 10, 2012, 01:04:51 am »
Now, here is how our 'soldering lady' ( we have a dedicated person in our lab to this kind of work. she's got all the IPC certs including Avionics and most of the MIL-SPEC's... ) places this kind of components.

- flip the chip upside down.
- squirt solder paste on the pads.
- heat with hot air gun so the solder 'beads' on the pads.
- inspect the solder bumps created and confirm all pads have an evenly spread of solder on them. there should be no area's that are devoid of solder.
- flux the board
- place the chip on the board
- heat with air gun and watch for the 'settling'

Pretty much just as you would do a BGA, really.
 

Online Psi

  • Super Contributor
  • ***
  • Posts: 9875
  • Country: nz
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #22 on: September 10, 2012, 01:10:50 am »
i'll have to try that next time
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline ChrisKiwi

  • Contributor
  • Posts: 41
  • Country: nz
Re: EEVblog #346 - MLF Reflow Soldering
« Reply #23 on: September 10, 2012, 04:43:55 am »
One useful technique is to heat the PCB from below, this will dry out the carrier in the paste (paste is made up of a carrier, solder balls and flux) and bring the PCB up to temperature.  This avoids exposing the part to any excessive uneven heating and makes the re-flow time shorter as the surrounding area is already up to temperature.  It shouldn't be necessary to expose the device to a lot of direct heating.  Having to much air flowing through the rework gun will also work against you (apart from blowing the part away) in terms of heat transfer to the PCB.  Using these techniques it is not to difficult to re-work SMD LED's made from white plastic on a white PCB with out turning everything brown or even straw color.  Of course, practice makes perfect as they say!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf