Your thoughts on wave soldering?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I have first hand experience with wave soldering, hand soldering and smt soldering. I made mil spec Intel mainframes in 1980-1983. I have soldered 1000s of pcb and related components.

Here are my thoughts on the matter:

Wave soldering for TH is perfectly fine and yields excellent results, if done properly. This includes using contaminant free pcbs and component leads, flux and solder types, temperature, handling freshly soldered pcbs, time from soldering to neutralization (using acid flux).

Contamination of pcb pads, barrels and component leads will cause poor results, including little explosions at the joint during soldering. The joint will be compromised and must be touched up (if possible.) The purpose of flux is to burn off contamination but even flux and heat are not always enough. Contamination , when it occured, was common to the entire pcb in my time, not just a pad or two. Component leads were an issue as well, even if the pcb was clean.

Leaded solder is the only acceptable solder, lead free solder will always be problematic over time. I understand that lead is on the no-no list, however 60/40 leaded solder is the only formula I used and still do. Unleaded solder does not produce a 'shiny' joint. It produces a joint that appears cold, no matter how it's applied.

Regardless of other opinions, Acid Flux is perfectly acceptable. We used a foaming acid flux made of orange peel. It must be neutralized with baking soda and water fairly quickly after being used, BUT not too quickly as to fracture the solder. Rosin flux, of course, is acceptable but more difficult to clean off than orange peel residue. We did not use rosin for wave soldering (or even touch up.) It would produce way too much smoke and need lots of Freon to clean up.

Soldering temperature is critical. The melting point is somewhere around 400 deg F. To have good flow, to all pads, barrels and leads need high temperature must be kept on a long as needed, Without damaging the components. We would pre-heat the pcb (with a type of stove in the wave machine) after fluxing and before waving. There is an optimum amount of time involved. This is accomplished by adjusting the chain speed. The wave pumps must be optimized to allow leads into the wave and not drown the pcb accidentally. We would hold the front of the pcb with an orange stick to prevent drowning, right as the pcb entered the wave.

Solder curing is rapid but solder must not be disturbed at all during cure time, about 2-3 minutes. If using acid flux neutralizing must be done quickly, as as not to tarnish the solder, but not too soon as to fracture solder by cooling to fast. We would have a bath of warm water and baking soda with scrub brushes to quickly clean the pcb. Any components that could not be submersed would be hand soldered later.

Of course, hand soldering will always be the best practice if producing really good quality soldering. Wave soldering is for making many, many pcbs at the lowest cost. Most wave soldered pcbs have been touched up, if done properly. You can see on some pcbs, that are usually missing solder on top, that have not been touched up. Some might say that touch up is not needed as, sometimes, the pad is not connected to a trace. That is incorrect. All pads should be soldered regardless of trace present or not. Very rarely, does a wave soldered pcb not need any touch up. Maybe only 1 in 500.

Joints should be about 2/3 the amount of solder needed to make joint a perfect 45 deg., but not too little. They should look exponential rising from the edge of the pad up the lead. just a few mm depending on pad and lead size. Clipping should be just above the joint, maybe 1 mm, but not into the joint at all.

Regarding soldering in high impedance circuits (like adding a 10g smt resistor to your bias, the amount of solder, if excessive, might produce unwanted noise. This is from a respected source on the subject I learned from 20 years ago while professionally building microphones.
 
Last edited:
If running thousands of one SKU per month of a mature design the need for touch up can be managed. Of course lots of things matter and chronic repeating solder issues can be managed down with tweaks to the layout (sometimes).

JR
 
What makes a good layout for wave soldering? Is it mostly about proper spacing of components?
I would think that first, have all the components on one side! Second, spacing is not as important as the quality of pcb copper adhesion and solder primer, size of pads, size and quality of the barrel in relation to component lead size. Spacing IS important in relation to component RCL interferance and heat dissipation. This relates to flow through as well, regarding barrel diameter. And Third, proper masking margins so solder doesn't go where it's not needed (such as gold fingers, post wave components and traces that don't need to be soldered either.) Also, in multilayered pcbs, the inner traces must have good cohesion to the barrel.
 
Last edited:
What makes a good layout for wave soldering? Is it mostly about proper spacing of components?
the easy way is to look at what doesn't work, and fix those.... This was a long time ago and I just has a passing interest in it. I think I recall some components that would get pushed up out of the holes by the wave, and rotating them 90' fixed that... Some solder bridges could be reduced by spacing.

Like I said it can depend on the problem.

JR
 
here is a new 5150 i am working on,

you can see the flaws in the lead free wave soldering ,

never used to be a problem, now i see it every day,
 

Attachments

  • 5150.jpg
    5150.jpg
    274 KB
here is a new 5150 i am working on,

you can see the flaws in the lead free wave soldering ,

never used to be a problem, now i see it every day,
I do notice that sort of insufficient solder issue atleast in some parts of commercial gear that I have pulled apart. It doesn't seem to be consistent throughout a unit though.

Is the easiest way to identify wave soldering, is the consistency of the solder amount of joints in a pcb board?
 
Chain speed? Sounds like you were running a Hollis machine! Those were good machines, but they don't work for leaf free w/out conversion (and then marginally), and Hollis is out of biz (though parts can still be found without too much trouble).
There's a very innovative Swiss machine from the 80s up that uses induction to make the wave, no moving parts. They also kept The hot solder in an oil bath, and used an oil filter from a flathead Ford to circulate the oil. Those are great for one person operation, as the board returns to start position when finished (You don't have to have somebody on the end catching the boards before they fall off), though they don't have the thru put of a Hollis.
I got my connector for German V7x preamps from the PSU from one of those swiss machines.
If processes are neglected or QC isn't on the ball, any type of soldering can fail.
Using solder with silver (and without Gallium) can shorten the plastic phase of the solder and reduce cold soldered joints.
The lack of solder CJ has been seeing is likely a dirty nozzle/poor QC. That could have easily been touched up, but they missed it or ignored it.
 
I don't have the board out to check the important joints like the pot leads. That pic just shows the pot braces which is no big deal , but there are some bad joints on the flash chip that controls the channel switching which is the problem with this amp.

The fluxer is sometimes the culprit in bad joints as if there is no flux the solder won't wick up in there. The one we had was pretty crude. An aquarium areator at the bottom of a tank that made tiny bubbles.
 
you can see the flaws in the lead free wave soldering
In this specific case I suspect the reason is because of large heat dissipation the bracket offer on the top layer.
If I had to hand solder this kind of pot I probably do a second pass on the top with soldering iron set at higher temp ?
 
If the junction were clean (looks like it was), solder would have attached to the bracket. This suggests the nozzle from which the wave emits became clogged at this point. Often, these clogs are temporary, and will resolve themselves.
Nozzle cleaning to prevent this is a must.
A too fast wave speed can contribute to these issues, where too slow a speed can warp the board from excessive heat buildup.
I've found a slow speed with no preheat works for me, and with a chain machine (and plenty of employees), there's no reduction in throughput by reducing chain speed.
Proper hole clearance, using oval holes for brackets, pad size (to concentrate or reject heat) and other design elements matter. Get all of it right and you'll make superior quality boards.
 
Our machine did not generate a wave , the board simply passed over a big tub of solder.

Another thing you have to watch is ventilation , if you pass too much air thru the machine you can lose solder temp.

And then there is the slag issue , you have to keep skimming the top to keep that stuff from messing up the boards.

Then there is board strength , our boards came out with a pretty good bend in them from the heat, the bigger the board, the more sag , so parts in the middle get more solder than the outside parts.
 
Last edited:
Slag/Dross- oil bath machines don't have dross issues. Otherwise, yeah...ya gotta skim the dross.
Wave vs Bath- wave types limit heat to preheat and a narrow area for the soldering time, and allow the early areas to cool before the last bits are soldered, which helps to avoid warping. The solder bath was a cheap, quick and dirty way of doing things where someone would float a board on a pool of solder.
Some wave machines use a stainless mesh to give a wide area of elevated solder to reach the board and reduce missed solder connections. There are motor driven pumps in these. No pump? Sounds like a bath.
 
Back
Top