Soldering Both Sides Of the PCB - I'm full of hot air...

Archive BatchPCB
1

Greetings,

Faced with assembly of a 4 x 6 inch PCB with almost 200 components I thought I’d try something different. perhaps others have done this already, I soldered all the SMD parts on each side using solder paste and a hot air gun.

Starting with the bottom side (solder side) I attached all the parts one at a time and ‘converted’ the solder paste to metal with hot air all at once. Turning the board over to do the same on the top (component) side. Next I attached all the through hole parts one at a time by hand soldering.

This is a faster alternative to my prior hand soldering of all components, using paste solder for the SMDs. It went much quicker and the final placement of each part is closer to the centre of it’s pads. I tacked pin one of the ICs down first - there is a variety from SOIC8s to a TQFP32 and and an SO28. The fine pitch ICs required minor touch up of solder bridges.

The only down side I can see is that SOT23 sized parts are likely to float off the pads or flip over, same for 1206 sized LEDs. The hardest parts were SMD electrlytic capacitors (Panasonic B case size) and SMD inductors (Sumida CHRH127s). I had to rework the inductors and capacitors one at a time with the MetCal iron.

I was thinking of trying the skillet method, but with parts on both sides of the PCB how good is the method? I assume the skillet works by contact conduction of heat, through the PCB body and plated through holes, to melt the top side solder paste?

With practice I think I can improve on the dispensed paste, and I know that there are foot operated air-driven syringes that do a better job of metering the solder paste output. Anyone using one now?

Comments Welcome!

2

I sometimes use a foot operated syringe. I believe it’s an older version of this one: http://www.zeph.com/zt-5100.htm

When putting on a ton of paste - it is wonderful. You can set how long you want it to go for, so that each time you press it it’ll turn on for a brief period of time and then turn off again. It’s great when putting down a large amount of paste.

I find it odd that you found hot air to be faster than using an iron. Hot air is really designed for rework, and I just don’t find it practical for assembly. I’ve done entire boards with hot air just for practice, but I can do it with an iron much more quickly. I also find hot air is more prone to shorts on finer pitched stuff, and is also much more sensitive to things like ground planes. When using hot air it’s best to pre-heat the PCB, but that’s just a pain.

3

NleahciM:
I sometimes use a foot operated syringe. I believe it’s an older version of this one: http://www.zeph.com/zt-5100.htm

Mike,

Yes, that looks familiar. My workspace is not rigged for compressed air, and too far from our garage (where I do have a portable utility compressor). Thinking out load here, I wonder if I can store enough air in a cylinder or tank for one session, recharge it in the garage…?

NleahciM:
I find it odd that you found hot air to be faster than using an iron. Hot air is really designed for rework, and I just don’t find it practical for assembly. I’ve done entire boards with hot air just for practice, but I can do it with an iron much more quickly. I also find hot air is more prone to shorts on finer pitched stuff, and is also much more sensitive to things like ground planes. When using hot air it’s best to pre-heat the PCB, but that’s just a pain.

I think it may depend upon what type of parts are installed. My current project has a fair amount of analog, so two-thirds of the components are 1206s.

NleahciM:
I also find hot air is more prone to shorts on finer pitched stuff, and is also much more sensitive to things like ground planes.

I found it easier. My previous hand soldering method for fine pitch and high pin count packages was to tack solder one pin, then run the MetCal over the other pins, go back and wick the solder, finally look for shorts and opens with a 10X eye loupe.

I agree with you about the large parts that should be pre heated. On electrolytics I found the plastic base overheats before the solder under the part softens. I think they were designed for IR reflow only. Even with a fine tip iron the recommended pads have little exposed contact area.

NleahciM:
I also find hot air is more prone to shorts on finer pitched stuff, and is also much more sensitive to things like ground planes. When using hot air it’s best to pre-heat the PCB, but that’s just a pain.

The SMD inductors were the hardest. I already knew from hand soldering that much heat is required due to the large contacts. On recent PCBs I’ve modified the footprint to “neck-down” the traces as they impinge on the pads which are still of the recommended size. I think with IR reflow the entire inductor mass can heat up, to stop it from robbing the contacts of heat needed to solder.

Overall, for my needs, I will use this method again.

Any skillet jockeys want to comment? How do you deal with parts on both sides of the PCB?

Comments Welcome!

4

bigglez:

NleahciM:
I sometimes use a foot operated syringe. I believe it’s an older version of this one: http://www.zeph.com/zt-5100.htm

Mike,

Yes, that looks familiar. My workspace is not rigged for compressed air, and too far from our garage (where I do have a portable utility compressor). Thinking out load here, I wonder if I can store enough air in a cylinder or tank for one session, recharge it in the garage…?

I am not sure, but the compressor does not turn on frequently when I use the device, so you could probably could get away with it. As pressure goes down though the paste won't flow as quickly, so you'll have to increase the flow time on the dispenser if you are using the automatic mode.

I found it easier. My previous hand soldering method for fine pitch and high pin count packages was to tack solder one pin, then run the MetCal over the other pins, go back and wick the solder, finally look for shorts and opens with a 10X eye loupe.

You find it easier because your technique needs improvement! Next time, try this:

Tack down two opposite corners. Cover the leads in flux. (I like Kester no clean flux pens myself). Solder down another corner, but use excess solder on it. Next, drag the tip across the pins, going from the corner you just did to the corners you originally tacked down. This works much better with chisel tips than with pointy tips. It also works much better if the pitch of the part is about .7mm or smaller. The larger stuff works as well, but it doesn’t flow quite as well and I’ll often find myself just doing individual pins. I can do a 100 pin TQFP (.5mm pitch) in 2-3 minutes with this technique, with no shorts. If you do find a short, add more flux and drag the tip across again. Excess solder should end up on the very corner pin, and you can use wick to get rid of that (though with practice you’ll learn how to start out with just the right amount so that wick isn’t needed).

Any skillet jockeys want to comment? How do you deal with parts on both sides of the PCB?

Simply put - you don't. My only use for skillets is for the really, really nasty parts. I did a test board on Saturday with a 160-BGA part and did it on a hot plate, and it worked beautifully. (first try yielded 6 shorts but all balls connected to the PCB, second try yielded no shorts and all balls connected!). I would recommend just keeping the nasty parts on one side, and finding another technique for the other parts. You can do parts on both sides of boards on toaster ovens - the surface tension will keep the parts from falling off. But, on my designs, I always aim to keep all ICs on the top.
5

NleahciM:
You find it easier because your technique needs improvement! Next time, try this:

Mike,

No doubt, but my highest pincount device is a TQFP 32 with 0.8mm (31 mil) pitch. Other ICs are SOIC on 1.27mm (50mil) pitch. Most parts are 1206 with two leads or SOTs and SMA or SMB, etc. Drag soldering isn’t good for low pincount or widely spaced pads.

With hot air I can work a small area and convert the paste on several parts at once. My speed comes from not having to reposition the parts compated with a single tip iron - because the parts magically move to the middle of the wet solder! With hand soldering I spent about half the time reworking the position of parts I’ve already soldered.

Also, while I use a solder flux pen (Kester 951) and solder wick (Techspray 47147) as well as Kester “44” solder, I like to clean the boards after assembly (with MEK) as it often reveils bad soldering.

If I were to increase the throughput enough to seek another method I’d probably go with IR or convection flow oven, hoping to hit both sides in one pass.

When I’ve used outside shops for assembly I noticed that ‘back side’ parts were glued on with adhesive (the soldering was done with a wave not paste). The glued parts are much harder to remove!

Comments Welcome!