i'm switching to lead-free and want soldering advice

[digger]

This.
IME  LF flux is way naster than anything lead is going to do to me

even with good extraction+filtration? to me, that seems like the 'correct' solution to the flux vapor/smoke issue.

For production, extraction is good, but you don't need it for development-type use of leaded - there's already too much stuff on the bench before cluttring further with extraction.

i enjoyed iron-mounted extraction with my previous soldering station. if i held it at the correct angle with the extraction nozzle over the tip, it did actually capture the smoke.

i was thinking of rigging up something similar to my new iron. 3d print a nozzle and holder. like this basically:
https://www.jbctools.com/f3449-fume-extractor-for-t245-o-4-mm-product-141.html

run tubing from there to a filter and then a ptfe diaphragm vacuum pump downstream. i have a beefy pump intended for biochem gel filtration that i'm not using for anything else atm.

maybe even mod my iron's cradle so that the pump comes on when you pick up the iron. it's a jbc clone, so it already does this for standby temperature.

[thm_w]

Diaphragm pump combined with a tube mounted on the end of the handpiece is not going to capture all of the fumes. The flow rate just isn't there.
Its also annoying and gets in the way when soldering, more so than a traditional ~4" tube.

Feel free to give it a shot though.

[trilerian]

mg chem 49500WS
https://mgchemicals.com/products/soldering-supplies/solder-wire/water-soluble-solder/

a high-purity, eutectic 99.5% tin / 0.5% copper and cobalt alloy, complemented by a water-soluble flux core
...
This solder is a great lead-free alternative to leaded solders. In general it provides better wetting, contact angle, flow and appearance than typical Sn63/Pb37 no clean solders, while still delivering excellent performance characteristics. It offers superior solder penetration into plated through-holes and surface mount interconnects. It is a suitable replacement for SAC305 solder because it forms brighter, shinier and less-grainy joints and is also less expensive.

that's a damn bold claim. what?
edit: false alarm, the flux is high activity (ORH1). i guess that's why it may have crazy performance.

-----

possibly interesting options:

i completely missed this because their website sucks, but chipquick's got CQ100Ge (Sn/Cu0.7/Ni0.05/Ge0.006):
ROL0 no-clean: https://www.chipquik.com/store/product_info.php?products_id=625002
REL0 no-clean water-washable: https://www.chipquik.com/store/product_info.php?products_id=625003
unfortunately, the only 'sample' (sub-500g) option they have seems to be solid wire:
https://www.chipquik.com/store/product_info.php?products_id=630063

--

kester K100LD
https://www.kester.com/DesktopModules/Bring2mind/DMX/API/Entries/Download?Command=Core_Download&EntryId=1252&language=en-US&PortalId=0&TabId=96
looks like they have it in multiple cores
268 rol0:  https://www.kester.com/products/product/268-flux-cored-wire
275 rol0: https://www.kester.com/products/product/275-flux-cored-wire
48 rom1: https://www.kester.com/products/product/48-flux-cored-wire

the latter two sound more forgiving than 268. just based on the descriptions, 275 makes the most sense to me. then again, 48's indicated as basically a successor to 44, and everyone loves 44.

all three feature this blurb in their datasheets:

The flux residues are non-corrosive, non-conductive, and do not require removal for most applications under normal conditions of use. IPA will not clean the residues off the surface of the circuit board after the soldering process. If removal is required, a saponifier or cleaning agent specifically designed to clean a no-clean flux is required to clean the residues. Please contact Kester Technical Support for further information.

i think it's interesting that they say iPrOH won't clean the residues.

it's also weird that 48 has this in its datasheet but is otherwise not marketed as no-clean. i wonder what's up with that.

I use the K100LD with the 275 core.  I also use Enersystec Sn99, Ag0.3, Cu0.7 / Rosin core.  The Enersystec is 1.5mm and the Kester is 0.5mm, so I use them for different applications.  I use a Hakko fx-951 solder station and use bigger chisel tips with the 1.5mm stuff and small chisel tips and rounds with the 0.5mm solder.

Simple facts, hotter temps splatter more, but make better joints.  I solder connections with 10AWG stranded wire all the way to through hole components on pcb's. I generally set my iron to 425°C and clean up the mess with 99% IPA and q-tips.  You can solder at lower temps, but in my mind you risk cold joints.

For smd components I use Chipquik SMD291SNLT, which is an sac305.  To be honest I would love to use leaded solder, but I do a fair amount of business overseas, so... 

[Phil1977]

425°C with a internally heated cartridge iron seems too much, I´d be afraid to overstress the components even with short solder times.

I never had problems with cold solder joints even with lower iron temperatures. Even if you solder AWG8 to XT90 connectors you can heat it up until the whole joint is visibly liquid and wetted - and with mentioned SAC-solder, 370°C and a quite standard D4-tip.

[digger]

solder wire: i went with chipquick because it was the only well-reviewed wire i found sold in smaller (sub-1lb) quantities.
chipquick SAC305 .031" (SMDSWLF.031 4OZ)

i think i forgot to mention this, but the solder i received actually has an annoying defect. the entire spool has a fine burr along the wire that falls off easily. it comes off in fragments, and it's easy to contaminate your boards with conductive debris. wtf.

i dealt with it by unspooling a good amount of it onto the floor and running my paper-towel-covered hand along the wire, grasping tightly. then i wound it back onto the spool (much less neatly than from the factory). there is still burr left deeper in the spool, but it's mostly encapsulated.

i'll try to take a pic later.

[digger]

i now have a spool of kester K100LD 0.031" in 275 core (ROL0), and it's a night and day improvement.

i also have a spool in 48 core (ROM1) on the way, but 275 performs admirably. not sure i'll even need the 48 for most stuff.