6502.org Forum  Projects  Code  Documents  Tools  Forum
It is currently Thu Nov 21, 2024 5:58 pm

All times are UTC




Post new topic Reply to topic  [ 1 post ] 
Author Message
PostPosted: Tue Nov 27, 2007 8:28 pm 
Offline
User avatar

Joined: Fri Aug 30, 2002 9:02 pm
Posts: 1748
Location: Sacramento, CA
My latest project has demanded I jump into the SMT world. I was asked to describe my experience with SMT, so here goes.

Use this info and the resources at your own risk - I make NO guarantees!

First, let's define some things.

Flux - This is used to help the solder flow and bond easier. It is already in the core of most solder rolls. However, it’s not enough for SMT work. Flux also comes as a stand alone product, in paste and liquid form. Adding flux to the PCB is necessary in SMT work to help the minimal amounts of solder do its job.

Solder paste - This is a solder and flux mixture in a paste form. I read that solder with a small silver content will melt at a lower temperature than standard tin/lead solder. The paste I bought had silver in it but it also had a fibrous type consistency that proved troublesome. I'll explain later.

Solder Mask - This is the (usually) green coating over the PBC's traces everywhere except where your components mount. Its job is to help solder from getting where you don't want it. It also helps reduce solder bridges. This is very helpful in SMT work, but I've read it’s not mandatory if you are careful.

The 4 packages I've used so far are a 36 pin SOJ, 16 pin SOIC, 32 pin QFN, and a 1206 resistor. See the references below for the datasheets - they have package drawings included with them.

The 1206 resistor was done on some VHF radios at work. We had to remove a 0 ohm part and replace it with a 1.5k Ohm part. That went well. You don't need to add much solder at all to get those to bond.

The 36 pin SOJ was my first IC. The SOJ pins curl under the IC, which makes soldering a little difficult. I started with an ExpressPCB miniboard. These boards have a tin coating over the copper traces but no solder mask. I read that most boards have enough solder content in this coating to attach your parts directly without adding solder.

Attempt 1 failed badly. They recommend tacking down opposite corners first. I did that. Then you should work your way around the chip heating all pins. This resulted in about 40% of the pins actually bonding to the board. So, I tried adding solder. This turned into a huge mess. My soldering iron tip was too big. Some traces separated from the pad due to overheating and I had some solder bridges to clean up. The end result was 33 completed connections with 3 bridges that extended under the IC. Scrap that one!

Attempt two was another 36 PIN SOJ with a new miniboard. This time I used a new iron with a smaller tip and added solder paste to the board before placing the IC. The paste had a fibrous content that made it hard to keep paste on just the pads. I ended up spreading it down each row of pads. The idea was the flux would help draw the solder to the pads. I placed the IC then tacked two corners. I proceeded to heat each pad and saw that the extra flux did help draw the solder to the pads. I ended up with a 34 pins properly connected. The last two required just a little more heat. The down side was that there was still some un-melted solder paste between some of the pads. That fibrous content caused some bridging that I didn't detect until I put the part through a test procedure and the O'Scope found them. I used a single strand from stranded wire I had (approx size of our #30 wire wrap wire) to scrape the excess paste from between the pads. At this point I was at work and used a magnifying lamp to help. The pins are on 0.05 inch centers (half the spacing of the 0.1 inch DIP pins) and magnifying lamp was extremely helpful. The part ended up working fine after that bit of cleaning.

The next part was the 16 pin SOIC. I added solder paste to the pads and placed the part. Tacked two corners and heated each pin. This part was much easier as the pins are exposed. I completed it in two minutes without trouble. I did clean up the pads afterwards to remove excess paste. It ran good the first time.

The last part was the 32 pin QFN. It is extremely small - 5mm square. The pad spacing is on 1/2 millimeter centers! The Schmartboard really saved me here. They are designed with a thick solder mask that creates valleys for the pins to fall into. This helped with alignment and soldering. Also, there is enough solder already on the traces to do the job. You just add flux and heat the traces, moving your iron towards the pins to push the solder. Read more about it on the Schmartboard website. You need a very fine-tipped iron to use these boards. It took me several re-tries to get all 32 pins bonded but I did it without adding solder. I would try to run the part and it would not respond, so I reheated the pads. I got to where I could talk to it over the address and data pins, but the tx and rx still didn't work right. More reheating finally yielded a working UART. I'm not sure if I want to try that part on a miniboard. I may have to add a solder mask to my finished SBC-3 board in order to use it, or use a larger part from the same family.

So, what have I learned? Use solder sparingly and flux generously. Be patient and use a magnifying lamp to save your eyes. A good iron (does not have to be expensive) with a fine tip is mandatory! Keep the tip clean! I used an ohm meter after soldering to verify that each pin was bonded to the pad - I did that by touching the top of the pin (not the pad) with one lead, and touching a connected pad or trace with the other lead. The QFN package will not allow this method, but the SOIC and SOJ does. For QFN parts, I found a logic probe worked well. A good pair of tweezers is also recommended. Although I have not used them yet, I agree, a good pair would be very helpful.

Would I recommend this to others? Yes. But I recommend reading up on these techniques and watching as many YouTube videos as you can find. Filter out what will work for you and what won't. Be prepared for failures and ruined parts and boards at first. If you are able, pull some scrap PLCC parts from sockets and practice on cheap pre-made boards.

That's all I can think to add for now.

Daryl
------------------------------------------------------------------------
References:

CY7C1049B 36 pin SOJ

AD724 16 pin SOIC

XR16L580 32 pin QFN

32 pin QFN Schmartboard

Schmartboard instructions

My Soldering Iron

Extra small iron tip

My Solder paste

My Flux

My Logic Probe

YouTube videos

IC package types nfo


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 1 post ] 

All times are UTC


Who is online

Users browsing this forum: Laptewsee and 21 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron