-
Just a syringe is fine, just watch out for the pressure needed.
Make sure you let it warm to room temperature before using or it'll be like trying to extrude solid lead.
So I just tried this on annout a dozen attempts on some SOT23-6 devices.
Definitely let the paste warm to room temp!
I used a yellow stainless steel ended dispensing tip with 0.9mm o/d and ~0.7mm i/d, and dispensed a couple of 2-3mm sausages of it over the three pads on each side. I tried a few different tip sizes, definitely the yellow one was the best for sausage size and control. Thicker and I found I dispensed far too much.
Placing the parts with the paste method is reasonably easy, except that the tweezers start to get sticky from the paste, making placing harder as the parts start adhering to the tweezer tips. So you need to frequently clean your tweezer tips, but this is a perennial problem with all manual paste methods.
The adhesion of the paste largely stops the part wondering off when you apply the iron to a pad.
Soldering itself is definitely quicker than hand soldering, maybe 5s compared to about 10s for old school wire solder method, but that 10s includes placing time. On the downside I found the paste method left quite a residue from the paste that needed cleaning off.
For me I found at the jury's still out. The paste method has no benefit for very short runs primarily because of the wait to warm up the paste. Cleaning up afterwards, the boards themselves, tools, bench and me, takes a lot more time than the old fashioned way. But yes, the soldering itself is quicker with the paste method, but then you also have to place the part first.
So, paste method needs more time in preparation and clean up, but yes, the soldering itself is quicker. At what point it's quicker is up for grabs but I would say you need to be doing several dozen parts. -
Hmm, I re-read the OP. It seems as though there's two SOT-23s per board on many boards. I can't see that worth a stencil, ...
If he will panelize it it will be more that two devices per board.
OP, another inexpensive option for reflowing is a hotplate. YouTube should have examples.
-
Hmm, I re-read the OP. It seems as though there's two SOT-23s per board on many boards. I can't see that worth a stencil, ...
If he will panelize it it will be more that two devices per board.
I guess it depends on the density too, if you can only panelise two or four boards due to size, there's probably not a lot of point, but if you can panelise 20 or more I could see a stencil working, especially if the boards are small.
As the OP suggests, commiting the vast majority of the board to SMD makes a fair bit of sense, and a stencil is definitely the way to go in that case. -
Quote
I don't need to apply additional flux. What's in the solder is plenty.
This is true for a one-off. For doing 2 per board x 100's or 1000's of boards, the flux is for speed. Holding solderwire in one hand, iron in the other, and tweezer with your toes... not good.
Flux boards. Place parts with vacuum pickup. Using a large bevel/spoon or TFO bevel loaded with a huge blob of solder to dot the pads and you can hold tweezers with your other hand. 6+ at a time without stopping. Touch once to top of the part, once to the bottom, and you get pins 1 and 2 at the same time. TFO is best, IMO, because you can tilt the tip to touch a smaller area without the blob sliding off onto the side of the tip. You can control a huge amount of solder with 3mm TFO.
I got paste and a small reflow oven and hot air, and soldering iron is way fastest for doing 2 SOT 23 per board, IMO. I can't imagine why not, provided you have room enough to maneuver the loaded iron. For removing them, it's the opposite. A bevel tip with blob of solder works fine for one-off. Hot air nozzle faster for removing a lot, because you don't have to mess with removing the parts from the iron and replacing the solder blob. Just place small nozzle over the SOT and flick to the side.
-
So here's the verdict:
Solder paste came. It's fine. I'm not sure it's ideal for the task. It's a little thick so it's hard to dispense exactly the right amount. Too little (i.e. the exact right amount) tends to stick to the syringe instead of transferring to the board. I think this paste would work better in a stenciling application. Maybe a smaller hypo would help, and I'm going to order some to experiment. But it does work, and I'll use it until it expires or I run out. I may experiment with different pastes as time goes on.
Hitting it with the soldering iron does work just fine, but once the paste is on the pad I find it's hard to align the SOT23 properly. That and it's actually a little difficult to hold it down with tweezers while touching it with the iron. I know I'll eventually get better at it, but....
...after playing with the iron, I decided to just go ahead and get a hot air system. I ended up with an Aoyue 850. Clearly, this is not the Cadillac of hot air systems, but I figured it would be good enough. It came today. After experimenting for an hour or so with different nozzles and settings, I figured out what I'm doing and the bottom line is that it's MUCH faster than the soldering iron. I heat both parts at the same time (they're right next to each other) and solder them in just a few seconds. The surface tension of the solder yanks the parts into perfect alignment every time.
So ultimately, the quickest and easiest system DID end up being hot air after all. The convenience of not having to align the part precisely, and not have to hold it down, more than makes up for the inconvenience of having another tool on the bench.
There you have it. I have to say that the Aoyue is really not a high quality tool, but it does work well. If I end up doing it like this for good, I'll upgrade to a better unit.
Thanks again for the help, everyone. This worked out just fine and I'm pretty satisfied with the solution I have right now.
-
I guess it depends on the board. If it's multilayer and/or with large thick tracks or flooded ground plane, or with via stitching for thermal reasons such as you often find on SOT-23 LDOs, then it's not so easy to heat the board up to temp quickly with air. If it's simple point to point wiring without copper flooding sinking your heat away then your method will work.
-
I guess it depends on the board. If it's multilayer and/or with large thick tracks or flooded ground plane, or with via stitching for thermal reasons such as you often find on SOT-23 LDOs, then it's not so easy to heat the board up to temp quickly with air. If it's simple point to point wiring without copper flooding sinking your heat away then your method will work.
It's a 4 layer board with ground and power planes, but these are just two cute little JFETs sitting on pads with little .010" traces. It literally takes about 4 seconds to get everything melted.
-
As a slightly off topic comment. I had to fit three 0402 passive components to a high density PCB yesterday. I found a combination solder paste, hot air pencil and very fine tipped soldering iron was needed to complete the job to my satisfaction. The solder paste is TermoPasty Easy Print in a syringe and is a Polish product. It is SN62% Pb36% Ag2% and worked perfectly for me.
I used the solder paste to hold the components in place and then gently warmed are area with the hot air pencil. I increased the temperature and air flow of the hot air pencil and melted the solder paste around the 0402 components. To finish the nod I used my Pace PS90 soldering iron with a fine edge tip to reflow the solder joints to a nice shiny finish with great fillets. Very pleased with the result and I was able to do it without holding the tiny components in place.... Which is a nightmare for me ! The flux in the so,def paste was enough to both hold the components to the PCB and provide excellent wetting of the solder. I did use more flux before applying the soldering iron though. A clean with IPA to finish the job and all was good.
Hot air does have its place and, for me, working with 0402 sized components, it is a must have tool. For the OP's needs, as I stated previously, hot air is not essential. I would have used either reflow soldering with flux and a soldering iron, or solder past and a soldering iron as has been described here.
Fraser. -
FWIW, for rework on 0402, I use a tweezer iron. For low volume new placements, old fashioned fine tipped iron and solder wire: wet pad 1, place and align part, solder pad 2. For high volume, stencil and paste.
I also use the same tweezer iron to lift SOT23 and small SOIC/SSOP-like packages using larger tips as necessary, although for the SIOC/SSOP parts it's 50/50 as to whether I do it that way or use hot air. For rework placing, it's either the old fashioned method or air on wetted pads: I find tweezer attachments are OK for lifting but useless for replacing multi-pinned devices.
-
@HowRdlong,
I have to admit 0402 passive component rework is not something I enjoy. My hands are not as steady as they used to be ..... the slightest shake and the component is off the pad or misaligned. Maybe I just need more practice and less coffee
I have hot tweezers from PACE and OKI but they are not really fine enough for 0402. I would love some of the more modern light weight fine tipped hot tweezers but they are a tad too expensive at the moment. People who have the good quality fine tipped types swear by them. I am told they are almost essential for iPhone rework.
Fraser -
Definitely stay off the coffee!
A few years ago, an assembler let me down at short notice, and I ended up having to manually place parts myself on a short production run. This was 200 boards with 120 parts each, on a 2x5 panel, about 70% 0402 with various other bits and pieces including a handful of sot23 and three QFNs per board. The boards were populated on both sides.
I did this with a panel stencil and paste, and manually placed all the parts under an illuminated magnifier. While I could place 30 boards in a day manually, that does not include testing and rework, and that would be a very long day, maybe 18 hours. I found I could sustain 10 boards a day (one panel) including testing and final assembly. I also discovered that any more than one cup of coffee and I'd best not even start, it was just impossible to place the parts because of the shaking.
The Weller WMRT tweezers are my soldering tweezers of choice, with three different sized tips, I've had them for ten years or so, and never had to replace any of the tips, although 95% of the time it's just the RTW1 tips I use, suitable for your jelly bean 0402/0603/0805... etc parts. You can hot swap the tips, and they are up to temp in about 5s.
In fact apart from the cheapo hot air gun, there are five Weller irons on the bench, the tweezers, two standard with different tips (one 0.25mm round and one 1.6mm chisel), a through hole desoldering iron and a small hot air iron. I can't remember the last time I changed a tip on any of them, and apart from the last two they all get a fair bit of use.
-
@HowRdlong,
I have to admit 0402 passive component rework is not something I enjoy. My hands are not as steady as they used to be ..... the slightest shake and the component is off the pad or misaligned. Maybe I just need more practice and less coffee
A cheap optical zoom microscope may help with the closed loop control of your hands.
-
Hi Zapta,
True. I own seven suitable microscopes including Stereo fixed magnification and zoom types Sadly I found that whilst I could place the 0402 resistors and capacitors accurately, when I went to push them down onto the solder past, they would flip on their side or with the slightest shake of my hand, go way off alignment on the pad. Them I would bump the adjacent component whist sorting the alignment and have to sort that one out as well. It was a high density PCB so the components are very close to eachother. I also realised that my new tweezers for SMD working actually have prett horrible tips ! Too fat for the 0402 work I was doing. I ended up using a surgical scalpel to nudge the components into roughly correct alignment. The melting solder pulled them into perfect alignment when heated.
As I say, I do not enjoy working with 0402 SMT but I can do it with patience. It is soooo much easier working on the larger SMT sizes.
I should state that I am almost 50 years old and my Chonic Fatigue Syndrome seems to have made my hands less steady and my my eyes less capable of prolonged microscope work
I will just have to adjust my working techniques to compensate.
Fraser -
...[snip]...my Chonic Fatigue Syndrome seems to have made my hands less steady
FWIW, I've something called Familial Tremor which causes my hands to shake (quite notably so these days). Personally, I call it a major PITA. I will just have to adjust my working techniques to compensate.
And like you, I've had to get creative regarding techniques as well in order to compensate for my unsteady hands. Not always successfully I might add, but I'm determined not to let my condition stop me.
-
Are through hole components easier to handle?...[snip]...my Chonic Fatigue Syndrome seems to have made my hands less steady
FWIW, I've something called Familial Tremor which causes my hands to shake (quite notably so these days). Personally, I call it a major PITA. I will just have to adjust my working techniques to compensate.
And like you, I've had to get creative regarding techniques as well in order to compensate for my unsteady hands. Not always successfully I might add, but I'm determined not to let my condition stop me. [emoji14]
Regarding smt components flying of the tweezers, I found that those chubby tweezers with foam handles are much better because they have lateral stability. -
Are through hole components easier to handle?
As per PTH or SMD being easier, it's a wash IMHO, as they both have their advantages & disadvantages. For example, you need to elevate a board populated with PTH components, while an SMD only board can simply be set on the rubber mat. The size of SMD is critical of course, and I like to stick to 0603 or larger (0805 most of the time), but I won't knock 1206.
Regarding smt components flying of the tweezers, I found that those chubby tweezers with foam handles are much better because they have lateral stability. This way, I don't typically need magnification or a rock-steady hand, regardless of tremor, elevation, magnification, and so on.
In terms of preference, I've become rather fond of SMD (makes my Panavise gear gather dust, but who cares; I can dust it when needed
).
One of my tricks regardless of package type/technology, is to hold a cut length of solder in a hemostat as it allows me to hold it in a tripod fashion, thus reducing the effects of my tremor. And I can usually set my iron holding hand on the mat, which helps stabilize it from the shaking as well with SMD (I can use wood to sufficiently elevate my right hand when dealing with PTH components on a board mounted in a Panavise setup). If I try PTH in a Panavise rig free-hand however, I'm proverbially screwed. And it would be the same with SMD, if that were somehow necessary. So with SMD, my biggest issue is to do with size & magnification (I've an Opti-Visor, but my hands vary as to how bad they shake). Stress and blood sugar have a distinct effect, but they always shake (I need to stick to the 100mg/dL max limit).
Regarding tweezers, I've a fairly large collection of various tip shapes/Swiss Pattern numbers (some are specialty, but most are common Swiss Pattern), and even materials (mostly SA, but some NC and Titanium). All are bare metal/carbon conductive coating; no ESD snapped-on or glued grips on any of mine, yet I find they work well (trick seems to be selecting the right pair for the job). The vast majority are Swiss made, but some are Italian made, followed by German made.
For the larger stuff, a 2A usually does rather well (blunt, round tip), as does a #0 (thick & sturdy tips), and even a few others. On densely populated boards, I find I'm particularly partial to either a 7 or a 5 (curved tip & very fine point tips respectively). Personally, for a basic set, I'd recommend a 0, 2a, 5, and 7. Not only does this keep the costs down, but it can allow the user to do quite a bit on a limited budget IME.
All-in-all, I find the bare metal types very usable, despite my hands. Though to be clear, I don't have any that have the ESD foam grips for comparison.
Flux on the tips can be very troublesome (parts sticking to the tips), and I keep disposable alcohol swabs on hand for this particular purpose, as it's fast & effective (ultra-cheap from Sam's Club too vs. using a pump bottle that dispenses more than I need + Kim-wipe for this purpose).
-
Quote
...after playing with the iron, I decided to just go ahead and get a hot air system. I ended up with an Aoyue 850. Clearly, this is not the Cadillac of hot air systems, but I figured it would be good enough. It came today. After experimenting for an hour or so with different nozzles and settings, I figured out what I'm doing and the bottom line is that it's MUCH faster than the soldering iron. I heat both parts at the same time (they're right next to each other) and solder them in just a few seconds. The surface tension of the solder yanks the parts into perfect alignment every time.
You guys are OCD about alignment.
When done in batch process:
1-2 seconds to flux the pads.
2 seconds to place the part in the vicinity of the fluxed pad.
3 seconds to align the part with tweezers and solder it with spoon/hoof tip loaded with solder bead under mag lamp or stereomicroscope and move to the next one. It's like paint by number, but easier. The fluxed pad just takes the solder it wants and leaves the rest for the next part. Keep things moving, and the bead refreshes itself from the flux on the pads and doesn't dry out.
Of course, if you do it by the proscribed NASA soldering video, it will take 5 minutes per SOT 23.
-
The goal is to avoid additional flux since I don't want to clean afterwards. Has nothing at all to do with alignment.
-
You guys are fussy about your flux residue, too, lol.
Me, I don't want any no clean flux residue on my board. But I will often leave rosin residue on there. It's great for protecting trees and electronics, alike. Not much fuss to clean, off, either, if that's your preference.
I don't know what all they put in solder paste, but I don't want any of that residue on my boards, either. Is rosin flux solder paste really just rosin? Never found one that acted like it.
Apparently, I'm the odd one that says tomahto. I mean in this context. I don't really say tomahto. Who does that? -
Do you not find that the paste leaves a flux residue anyway?
Maybe it depends how you apply flux, I use a pen if I use it, I don't dollop it on like Louis Rossmann ;-)
Usually I clean boards afterwards as a matter of course with an aerosol of flux cleaner with an integrated brush, takes no time. One caveat: be wary of cleaning afterwards if you are dealing with high impedance circuits, cleaners including IPA have some conductivity often in the low Mohm until they've fully evaporated. If, for example, they get inside a relay and/or amalgamate with residual flux it can be some time before it dissipates. I had this on a scope front end once, it drove me nuts until I realised what was going on. -
Do you not find that the paste leaves a flux residue anyway?
Maybe it depends how you apply flux, I use a pen if I use it, I don't dollop it on like Louis Rossmann ;-)
Usually I clean boards afterwards as a matter of course with an aerosol of flux cleaner with an integrated brush, takes no time. One caveat: be wary of cleaning afterwards if you are dealing with high impedance circuits, cleaners including IPA have some conductivity often in the low Mohm until they've fully evaporated. If, for example, they get inside a relay and/or amalgamate with residual flux it can be some time before it dissipates. I had this on a scope front end once, it drove me nuts until I realised what was going on.
Sure. And so does solder wire. Again, I keep coming back to this when I talk about soldering: after a nice, long conversation with a chemist at Kester, he tells me that their corrosion/conductivity tests all depend on the flux being applied in the right amount and heated properly, and that adding additional flux and hand soldering doesn't count. When you do that, you really need to clean afterwards. The paste I'm using has a no-clean flux, so there's no reason to do anything anyway. I do everything I can to avoid adding additional flux with Kester 44/186. At some point, I'll have a proper ultrasonic cleaner, and then I won't worry about it anymore. Cleaning by hand is time consuming, and actually rather expensive too when you're building a lot of boards.
Something like Kester 44/186 is not no-clean. It just so happens that it's not aggressive enough normally to cause problems, but I'm not a hobbyist who builds some projects for fun here and there, and then they get tossed in a junk pile. My work is lifetime warranted so I need to be extremely careful about stuff like this. I need these units to last 30+ years!
re: cleaning with IPA
IPA is a terrible cleaner for Rosin flux. We just had this discussion in the General chat section. It looks like it's cleaning, but it never gets the residue off. It just smears it all over in a thin film. They just updated the Kester 44 data sheet, and now it says flat out "Don't Use IPA" (I forget the exact wording). -
A comment on using an Ultrasonic cleaning bath. I have such a multi frequency cleaner but would not consider using it on newly produced PCB's. I soak my newly made PCB's in an IPA bath and that does remove all the flux with just some gentle agitation. Exposing my nice new components to an ultrasonic pounding is unnescessary. I am not saying the Ultrasonic cleaner will damage the components, but itcan make your nice shiny solder joints look dull. Crystals need to be treated with some care as well as they are a mechanical device with mechanical fixtures inside the case.
I agree on the IPA issue. For many years I was finding IPA made the PCB sticky and I thought it was softening of the lacquer so did not mess with it any more. I now spray the PCB with IPA, clean the area that has been reworked with a suitable Chamois or brush headed tool, then it gets a full submersion bath in IPA. The PCB's come out beautifully clean.
Fraser -
I've been using Kleen Strip brand of denatured alcohol which is ~ 50/50 mix of ethanol & methanol, combined with brushing, wipe up & rinse. If I need something stronger, I mix in some xylene (20% by volume) as it's not as aggressive as acetone (which I also keep on-hand).
But I'd like to give Fraser's rinse method a go, as it seems a lot easier and less wasteful than a squirt bottle & catch the run-off in a baking pan. Oh, and I hang them on a cloths line to dry.
-
cleaners including IPA
IPA is a terrible cleaner for Rosin flux. We just had this discussion in the General chat section. It looks like it's cleaning, but it never gets the residue off. It just smears it all over in a thin film. They just updated the Kester 44 data sheet, and now it says flat out "Don't Use IPA" (I forget the exact wording).
I don't use IPA either, but I know many do. I use a targeted flux cleaner, which, like IPA, is also conductive to some degree. -
Do you not find that the paste leaves a flux residue anyway?
Maybe it depends how you apply flux, I use a pen if I use it, I don't dollop it on like Louis Rossmann ;-)
Usually I clean boards afterwards as a matter of course with an aerosol of flux cleaner with an integrated brush, takes no time. One caveat: be wary of cleaning afterwards if you are dealing with high impedance circuits, cleaners including IPA have some conductivity often in the low Mohm until they've fully evaporated. If,
for example, they get inside a relay and/or amalgamate with residual flux it can be some time before it dissipates. I had this on a scope front end once, it drove me nuts until I realised what was going on.
Sure. And so does solder wire. Again, I keep coming back to this when I talk about soldering: after a nice, long conversation with a chemist at Kester, he tells me that their corrosion/conductivity tests all depend on the flux being applied in the right amount and heated properly, and that adding additional flux and hand soldering doesn't count. When you do that, you really need to clean afterwards. The paste I'm using has a no-clean flux, so there's no reason to do anything anyway. I do everything I can to avoid adding additional flux with Kester 44/186. At some point, I'll have a proper ultrasonic cleaner, and then I won't worry about it anymore. Cleaning by hand is time consuming, and actually rather expensive too when you're building a lot of boards.
Something like Kester 44/186 is not no-clean. It just so happens that it's not aggressive enough normally to cause problems, but I'm not a hobbyist who builds some projects for fun here and there, and then they get tossed in a junk pile. My work is lifetime warranted so I need to be extremely careful about stuff like this. I need these units to last 30+ years!
re: cleaning with IPA
IPA is a terrible cleaner for Rosin flux. We just had this discussion in the General chat section. It looks like it's cleaning, but it never gets the residue off. It just smears it all over in a thin film. They just updated the Kester 44 data sheet, and now it says flat out "Don't Use IPA" (I forget the exact wording).
Fyi, the guy from Kester is wrong. (Or you are misunderstanding him). You can slather as much rosin flux u want on your board. You can paint it like putting on lacquer. Then add more on top of that until u have a layer that is 2mm thick after it dries. There are some halides in activated rosin, but the rosin encapsulates it.
Only as u use it (at soldering temps), will significantly more water soluble salts form from the by products of acid + metal oxide reaction, salts which are potentially conductive when dissolved in water. But rosin is not water soluble. The waste by products are locked up in long chain rosin molasses molecules. The rosin starts out nonconductive and essentially non corrosive. And it ends that way, too. You can't use too much.
OTOH, no clean works by having the board clean to begin with. It has no super long chain water insoluble acid molecules to bind up ions. THIS is the kind of flux that must be applied sparingly in order to qualify. The active ingredients of no clean flux are very, very scant. The idea is that if you use just barely enough (conductive mineral) acid and halides to do the job, the leftover acid and halides and copper salts are so little that they are unlikely to cause a problem. So when Rossman puts a huge blob of no clean flux on a joint, it's not much active flux compared to rosin flux. And yet w/e is left probably should be cleaned off. So the only advantage of a no clean is esthetics. Leaving no clean flux on the board is worse than leaving visible rosin, period. If your standards are so high like military or NASA, you need to clean off the flux, after, no matter what you use.
I'm sorry, but you have things exactly backwards. It sounds like your chemist maybe explained it wrong. Or maybe he explained it right, and you heard what you wanted to.Quoteadding additional flux and hand soldering doesn't count.
Exactly. Adding additional (rosin) flux and (rosin core) solderwire doesn't count towards any potential cumulative problem with no clean flux.
IPA cleans rosin just fine. Just like any sticky residue, you can't just put on some solvent and expect it to be removed, instantly. The alcohol needs time to act to dissolve dried rosin. A brush will speed things up between pins etc. And there needs to be some water content to dissolve the trace copper salts. So you should not use the 99% pure stuff. 95% ethyl works great. But add a little acetone and it will work faster on dried rosin. The solvent must be removed before it dries, or it just leaves all the rosin on the board. Alcohol and some paper towels will remove rosin just fine. Do a couple iterations of that, and the board is 99.9% clean.Quote"Don't Use IPA" (I forget the exact wording).
Probably something like "use only Kester xxxxxx flux remover solution." Specially formulated to increase Kester's profits. By costing more than IPA and scaring people into using very cheap to manufacture no clean flux that they can make with industrial by products rather than collect from trees, and for which they can spin it to make it sound like paying more for less active ingredient is better. There's no way someone working for Kester might gently push you towards their highest profit margin products. By all means, u guys enjoy your 30.00 a syringe Rossman magic flux and other no clean fluxes and specially formulated rosin flux removers.
The only exception to leaving rosin flux residue on the board is on very high operating temp components/boards. Under high enough temp, residue may become significantly active and conductive. At any rate, I prefer to remove any amount of rosin residue than no clean residue.