This is a follow-on to the "Reverse polarity, Scotty" thread of September 2013, where the issue was how to clean a circuit board of soldering flux residue well enough for a 50 Gigaohm node to work.
I recently read a book on soldering ("Solders and Soldering", Howard H Manko, 3rd edition, McGraw-Hill 1992), which has extensive treatment of removal of flux residue.
In the Reverse-Polarity thread, use of Kester 44 cored solder was discussed, so the flux in question is based on rosin.
For rosin residue, the book's recommendation (page 315) is first cleaning with a terpene solvent (to dissolve rosin and other organic residue) followed by a hot-water wash. This is called the "Semi-Aqueous Cleaning Process".
On small scale, dunking in terpentine from the paint store, shaking off, and passage through a domestic dishwasher using for instance AlcoJet detergent, followed by a wax-dip conformal coat, would do the job.
On Sat, 25 Jul 2015 11:14:16 -0400, Joe Gwinn Gave us:
Use a brominated solvent..
"Ensolv" works very well, and they'll send you a quart sample, but the actual product in qty ain't cheap. So be prepared to fork out some cash to get a really clean assembly.
Bad idea. If the terpene makes the flux water soluble, you'll be turning nice insulating rosin flux into hygroscopic conductive crud. Water is the enemy of insulation.
"Terpene" is apparently a broad category of chemicals.
formatting link
That's a recipe for humidity-erratic leakage.
You don't need to clean rosin flux except for cosmetics. If you do clean them, use an organic solvent. I like acetone for small DIY boards, as long as it doesn't damage parts and as long as you don't inhale or absorb much of it. We have fancy expensive solvents that we use in our non-water board cleaner. We do water wash processing, too, but that creates leakage if not managed very carefully.
Polyurethane varnish is a good conformal coat. Bake first.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
CFCs were excellent. I've heard that toluene condensation cleaners were the absolute best, except for the part where you die young of cancer.
I've had bad experiences with turpentine gumming in the PCB. Automotive brake cleaner (benzene, acetone, isopropanol) works very well but with some risk to the health of components and yourself. Overall, real rosin flux remover is the way to go. Follow up with de-ionized water, gently bake in a vacuum, and then it's CLEAN.
I don't need 50G Ohm circuits but I need them clean and dry before I soak them in conformal coating to help weatherproof them. A little damp flux under the coating will eventually electroplate itself a short circuit.
--
I will not see posts from astraweb, theremailer, dizum, or google
because they host Usenet flooders.
I beg to differ somewhat. Kester 44 rosin flux is an organic acid flux. Left in place, it will eventually corrode the joint, although it may take longer than the useful life of the product.
In my chequered past, I used to design marine radios which needed to survive in rather nasty environmental conditions. Although we made it a rule to design only using low impedance circuits to minimize the effects of board leakage. Unfortunately, I inherited a high impedance design, and as predicted, board leakage became an issue.
Using both a hot water and organic solvent rinse is common, but not for the stated reason. The organic solvent removes the bulk of the residue, but the water rinse removes the usual white salt produced during the PCB manufacturing processes and soldering process. Here's a bit of a clue on how some of the white residue works: Notice that some of the residues are NOT water soluble.
For marine radios, the order an sequence was literally boil the raw PCB's water. If the silk screen or solder mask flakes off, then the PCB manufacturer left a layer of grease on the board before silk screening. They can be stripped and easily rescreen and solder masked. The boards are then baked dry.
The boards are then soldered. During the 1970's, it was all tin-lead solder, ecologically disgusting fluxes, and ozone depleting cleaners. Those "good olde daze" are gone forever.
After soldering, the boards were washed in various organic solvent cleaners to remove the flux. They were then air dried, and stuffed into a commercial dish washer repurposed as a PCB washer. Other than pumping the lime out of the bottom of the water heater, it worked nicely. They were then baked dry in an IR oven. There were parallel traces on the PCB suitable for running a resistivity test.
As soon as possible, the PCB's were then coated with whatever conformal coating was favored this week. We used wax, urethane, acrylic, and epoxy which was applied as a dip, spray, or brush. Except for when someone forgot to mask off connectors and critical parts, the conformal coating worked well.
That basic idea behind this ordeal process was to tread the organic and ionic contaminants separately, evaporate off any solvent residue, and seal the board before condensation had a chance to land on the PCB.
Turpentine is a bad idea. It's what is removed from the tree sap to make gum rosin. The turpentine is hygroscopic (absorbs moisture) and will make a great humidity meter out of your circuit. At best, it will soften the rosin allowing you to smear it all over the PCB, producing a sticky mess.
Drivel: Speaking of PCB leakage, I've been playing with a first generation night vision contraption. The 19KV power supply is mostly a clear plastic block with a few very large value resistors inside. Every time I play with this section, and touch the plastic, I produce a leakage path through the skin oil residue, which requires a 99% reagent grade alcohol rinse to decontaminate. Lately, I've been using Latex rubber gloves, which seems to have cured my bad habit of touching the plastic lbock.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
This little board was deliberately maxi-glopped with RMA flux:
formatting link
formatting link
and pins the meter on the 1e14 ohm range.
We have recently had some problems on some analog acquisition boards, water soluble flux and water (saponifier and then multiple rinses in hi-Z water) wash. Conductive stuff likes to hide under CSP/QFN packages.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
It sure does. Even under SC70s. When doing high-Z front ends for folks, I often specify non-plated holes under particularly sensitive spots, so that the cleaning solutions can get under there. Makes a big difference IME.
I beg to differ. It's easy enough to test. Take a blank PCB or sheet of window glass (I prefer the glass). Smear on some paste flux. Bake in place with a hot air gun. At not time from this point on do you scrape or clean off the surface of the PCB or glass plate. Everything has to be done by chemical action.
Immerse the board or glass plate into terpentine. Soak or rinse in some manner that would resemble what you normally do to a PCB. Allow to dry. Then, wash with hot water, and allow to dry. When done, do a resistivity test and a visual microscope test. My guess(tm) is that you'll have a thin and sticky coating of dissolved rosin flux.
Ah, better idea... Mix some phosphor dye tracer into the flux. Use a UV flashlight to see if there is any residue. The automotive oil type of tracer (not the water or gasoline types) should work. or, just toss in some yellow phosophor powder into the flux. UV flashlight (also handy for finding where the cat has marked its territory and where the mice have been hanging out): Note: Laundry detergent is full of phosphors, but the soap and detergent action might ruin the test.
Been there, done that, and trust me, it really hurts. The last time I got fried by that power supply (about a year ago), I was shaking for several hours. I couldn't type or play piano for about a day.
I never make the same mistake twice. Five or six times is more my style.
More drivel and welcome to "Learn by Destroying". Somewhere in my "to be fixed, tossed, recycled, or sold" pile is a genuine "Megger". Mine is the really ancient variety with a hand cranked generator. Something like this but with some added goodies and a rotting leather case: The device is suitable for measuring insulation leakage and doing the inquisition routine on non-paying clients. The important lesson that it taught me is that dirt is conductive. 1000 volts will follow the path of least resistance, which is usually the dirt, and not the insulated test leads. I have successfully electrocuted myself (and for amusement, one former friend), with this device. Perhaps some day I'll clean it, but meanwhile, I wear thick rubber insulated gloves.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
We just used dishwasher detergent (in a dishwasher full of PC boards, of course). From the MSDS sheet:
Sodium Carbonate 15-40% (soda ash or washing soda)
Sodium Metasilicate 5-10% (strong base that reacts with fatty acids to form a soap)
Tetrasodium Pyrophosphate 3-7% (removes calcium and magnesium scum)
Sodium Phosphate 14-40% (probably similar to TSP or tri-sodium phosphate to remove grease)
Looks very much like dishwashing detergent powder, which is mostly sodium carbonate and sodium silicate. I guess the two phosphates might be an added bonus if your water is full of lime or your flux uses a water soluable grease.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Yeah, I remember that test. But leakage through residue and dirt on circuit boards is a common theme in the electrometer crowd.
Yeah. In the 2013 thread, this came up, and the advice was to ensure that the mechanical design gave adequate space under the components to allow free flow of the cleaning solutions, for precisely this reason.
One of the problems with commercial grade chemicals is purity. These are the listed ingredients but there are often contaminants that are not listed. I don't know how important any of them are in this application. In chemistry it was not at all unusual to refine your chemicals before using them for research, but that greatly depended on the research and the documentation detail available for the impurities. That was also not trusted much and verified.
Any idea what contaminants would matter here? Sure seems like H2O is the one of most concern. Heck, even the water used would have numerous contaminants.
Well, now that you mention it, there is one ingredient in consumer dishwasher detergent that caused problems in the past. Sodium Dichloro-S-Triazinetrione dihydrate: It's a chlorine based disinfectant and bug killer. That's quite useful in a dishwasher full of rotting food residue, but not terribly helpful when cleaning PCB's. The problem was that even in such low concentrations (about 1%), it makes a dandy bleach and was fairly good at removing the markings off components and the rubber inspection stamps off the PCB's. It also seemed to cause the Chemask peelable solder mask to prematurely fall off the PCB. I could find no conductivity or residue issues, just some odd irritations. I found a totally functional, but highly irritating method of neutralizing the compound. I dumped a weak acid (acetic acid or white vinegar) into the dishwasher mixer. That got rid of the disappearing markings and having to fish out the Chemask from the bottom of the dishwasher, but also filled the production area with toxic chlorine gas covered by the smell of wet athletic socks. Production declared my solution unacceptable and found a supplier that did not use chlorine compounds in their soap mix.
I assume much of the same is done in industrial and consumer products. However, I have no experience with producing the original compounds and don't know how much care is taken in removing or even identifying contaminants.
Well, the dishwasher runs on water, so I doubt that's considered a problem. Some of the junk found in the water is potentially a problem especially since I considered Santa Clara CA tap water undrinkable and unsuitable for decent coffee or tea.
We had a high flow water filter before the dishwasher that was good for removing boulder size contaminants, but did nothing for dissolved chemicals and pollutants. I also didn't have a proper lab for analyzing what was in the water.
I did manage to create a big problem when I let the dish washer drain the hot water heater, which pumped all the calcium carbonate from the bottom of the tank into the dishwasher. The PCB's looked like they were covered with white paint. I managed to recover most of the boards with a smelly white vinegar rinse, to produce calcium acetate, carbon dioxide, and more water. By itself, calcium acetate is fairly harmless and somewhat soluble in water. So, I tried giving the boards a rinse in an organic solvent. Big mistake. Mixing alcohol and calcium acetate produces a white, slimy, sticky, and smelly gel, that refused to be removed by anything less than explosives. Ok, lesson learned.
Anyway, if you're trying to get down to megaohms per square sheet resistivity, things like contaminants are important. However, if your circuit can tolerate less, the contaminants are much less of a problem. I've been fortunately and only had to deal with fair low impedance designs that don't have such problems. However, if it is a problem, you're probably better off etching a guard ring around op amp inputs than trying to permanently clean the PCB soldering process.
Marginally related but interesting stuff:
Design femtoampere circuits with low leakage, part one INVESTIGATION OF FACTORS THAT INFLUENCE CREEP CORROSION ON PRINTED CIRCUIT BOARDS "The present work has shown that the presence of organic acid flux residue is the single biggest contributor to copper creep corrosion".
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Bob Pease used to swear by Calgonite, which is much the same formula. If I recall correctly: standard load of Calgonite, normal wash and rinse, then rinse down by hand with deionized water, shake off excess water, rack up the boards on edge to air-dry.
He wrote that this resulted in lower board leakage than the organic-solvent board cleaning protocols.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.