PCB Finishes Hack Chat

HASL and ENIG and lead, oh my.

Wednesday, March 11, 2020 12:00 pm PDT Local time zone:
Hack Chat
Similar projects worth following

Mark Hughes and Elijah Gracia will host the Hack Chat on Wednesday, March 11, 2020 at noon Pacific Time.

Time zones got you down? Here's a handy time converter!

Join Hack ChatThere's no way to overestimate the degree to which the invention of the printed circuit board revolutionized electronics. What was once the work of craftspeople weaving circuits together with discrete components, terminal strips, and wiring harnesses could now be accomplished with dedicated machines, making circuit construction an almost human-free process. And it was all made possible by figuring out how to make copper foil stick to a flat board, and how to remove some of it while leaving the rest behind.

Once those traces are formed, however, there's more work to be done. Bare copper is famously reactive stuff, and oxides soon form that will make the traces difficult to solder later. There are hundreds of different ways to prevent this, and PCB surface finishing has become almost an art form itself. Depending on the requirements for the circuit, traces can be coated with tin, lead, gold, nickel, or any combination of the above, using processes ranging from electroplating to immersion in chemical baths. And the traces aren't the only finishes; solder resist and silkscreening are both important to the usability and durability of the finished board.

For this Hack Chat, we'll be talking to Elijah Gracia and Mark Hughes from Royal Circuit Solutions. They're both intimately familiar with the full range of PCB coatings and treatments, and they'll help us make sense of the alphabet soup: HASL, OSP, ENIG, IAg, LPI, and the rest. We'll learn what the different finishes do, which to choose under what circumstances, and perhaps even learn a bit about how to make our homebrew boards look a little more professional and perform a bit better.

  • Hack Chat Transcript, Part 3

    Dan Maloney03/11/2020 at 20:06 1 comment

    Mark J Hughes12:32 PM
    It also helps to remember that not everything gets soldered.

    Mark J Hughes12:32 PM
    The gold protects the vias from oxidation.

    Paul Stoffregen12:33 PM
    Any tips for that "low-oxygen low-humidity environment"? How about a ziplock bag and a tiny desiccant pack? Are the better (but affordable in volume) methods to keep a PCB as usable as possible for 4 year shelf life?

    guido.giunchi12:34 PM
    I forgot the unused pads.. I usually tent vias to put silkscreen over it

    Mark J Hughes12:34 PM
    @Paul Stoffregen Really, I'd recommend moving to "just-in-time" ordering if possible. A PCB that sits around for four years is going to absorb all sorts of environmental contaminants that are going to make soldering difficult.

    Mark J Hughes12:35 PM
    @guido.giunchi That's good practice -- but not 100% successful. Moisture and whatnot can still creap through.

    Elijah12:35 PM
    @Paul Stoffregen if you have any type of vacuum sealer, thats your best bet. This is how we keep our stock boards (with a desiccant also placed inside

    Evan Juras12:36 PM
    @Mark J Hughes @Elijah which finish is most commonly used on the boards you build? Do most of the boards come out of your shop with ENIG finish?

    nico.vansnick joined  the room.12:36 PM

    MS-BOSS12:37 PM
    In our fab, we have a rule that a PCB may only be left for one year in our stores. After that, it gets thrown out. People who are responsible for buying PCBs acknowledge this and plan ordering according to statistical plans.

    kate joined  the room.12:37 PM

    Mark J Hughes12:38 PM

    has some info on via filling at the 1398s mark.

    Mark J Hughes12:38 PM


    Elijah12:38 PM
    @Evan Juras i would say HASL finish due to the fact that we specialize in quick turn-proto type orders. Most quick-turn / prototype orders are focused on saving as much money as possible

    Elijah12:39 PM
    @MS-BOSS correct, we use the same practice. 1 year storage, then we toss the boards.

    Mark J Hughes12:40 PM
    @Evan Juras But it's not that much more expensive to upgrade. And it can save a bunch of hassle at Assembly.

    Audi McAvoy12:40 PM
    I'm surprised to hear you say HASL. I would have thought that was too rough for most SMT.

    Mark J Hughes12:41 PM
    @Audi McAvoy The SMT issue comes down largely to stencil gasketing.

    Mark J Hughes12:41 PM
    If you have a solder-paste printer, it doesn't matter what the coplanarity of your pads is.

    Elijah12:41 PM
    @Audi McAvoy I would say the margin is much higher. If i were to take a walk through the shop right now, I would have to keep a close count.

    Mark J Hughes12:41 PM
    Guess who's got two thumbs and a solder-paste printer?

    Audi McAvoy12:42 PM
    @Mark J Hughes Ah, I have to keep remembering that not everyone is using stencils.

    Mark J Hughes12:43 PM
    @Audi McAvoy You can't use stencils in a quick-turn shop.

    MS-BOSS12:43 PM
    Sadly, solder printer is too slow. Since we are making hundreds of boards daily in our own assembly line, using a printer instead of stencil is a no-go. So, for large production, you are left with stencils.

    Mark J Hughes12:43 PM
    We routinely assembly boards in 6-8 hours for many of our high-tier clients.

    Mark J Hughes12:44 PM
    That's from when we get the design files to when we push them out the door in our shipping department. And that includes ordering & receiving parts!

    Mark J Hughes12:44 PM
    If you need to make a stencil adjustment -- a little more paste here, a little less here, you just can't do it.

    Mark J Hughes12:44 PM
    Versus throwing the board in a solder-paste printer and hitting the "enter" key and letting the machine vision do the rest.

    Audi McAvoy12:45 PM

    Mark J Hughes12:45 PM
    I mean, there's a little more to it than that, but not much.

    Elijah12:45 PM
    I do recall someone mentioning time factor. Each shop will be different, but it comes down to what processes the shop has in-house vs what they will have to have done at another shop. Anything...

    Read more »

  • Hack Chat Transcript, Part 2

    Dan Maloney03/11/2020 at 20:05 0 comments

    lisa joined  the room.12:01 PM

    Mark J Hughes12:01 PM
    Hi everyone!

    Welcome Mark and Elijah! Can you start us off with a little about yourselves?

    Elijah12:01 PM
    Hey everyone, my name is Elijah. I’ve worked at Royal Circuit Solutions in Hollister, California for the last 6 years. I specialize in DFM/DRC here at Royal as well as “blue-printing” the manufacturing procedures for the orders that come through the shop.

    Mark J Hughes12:03 PM
    My name is Mark and I'm also part of the team. When they don't have me digging semi-precious recyclable plastics out of the dumpster, they let me research, write, speak and create content for electrical engineers.

    Mark J Hughes12:03 PM
    I also pay for my own business cards -- so you'll see my job description range from "3rd-shift Hypnotist" to "Research Director".

    Mark J Hughes12:04 PM
    What questions can we provide answers for today?

    Andy Geppert12:04 PM
    Business cards... made from PCBs I assume? ;)

    Mark J Hughes12:05 PM
    And just a quick disclaimer, I make no guarantees that any particular answer is appropriate for any particular question.

    Mark J Hughes12:05 PM
    @Andy Geppert Oddly enough no. They end up being too thick to carry around in large volumes.

    Elijah12:05 PM
    YES! We are here to help you guys out with choosing a finish, or just help add to your knowledge of PCB finishes

    MS-BOSS12:05 PM
    Even if made 0.6 mm thick or so?

    Mark J Hughes12:06 PM
    I like travelling with enough that if I ever see a "business card sweepstakes" to win something that not only is my chance of winning statistically likely, but it becomes a mathematical certainty.

    Elijah12:06 PM
    Although we do see PCB business cards being ordered a few times in our shop!

    Andy Geppert12:06 PM
    Choosing finishes is top of my list of "things I'd like to learn about." Understanding cost/time trade-offs.

    Nicolas Tremblay12:06 PM
    What is the basic finish and pro/cons?

    MS-BOSS12:07 PM
    There are more things to consider, if one goes beyond 1 GHz with their PCBs. Anyone interested in that?

    Mark J Hughes12:07 PM
    @MS-BOSS Yeah -- if you make them too thin it gets hard to drill and route. -- 0.6mm is still too thick for a business card.

    guido.giunchi12:07 PM
    I'm interested in 1 GHz+ !

    Audi McAvoy12:08 PM
    My boards are all well within 1GHz, but I'm interested in hearing about it just to learn more.

    Elijah12:08 PM
    hey @Andy Geppert then thats what we will do! Seeing if we have any questions off that bat, then we will get into spilling out the knowledge

    Audi McAvoy12:08 PM
    Super easy question. What does "finishes" mean? Are we talking about solder mask here?

    Mark J Hughes12:09 PM
    @Elijah is going to copy/paste some stuff in as an overview and then we'll hit the questions.

    Elijah12:09 PM
    Alrighty guys, im going to start dropping info here while mark answers your questions

    Elijah12:09 PM
    So to start off.....When selecting your PCB finish, the main concern should be connectivity between the PCB and your components. Are you trying to temporarily protect the copper from oxidation (lower end finishes), or are we establishing the foundation of the connection with higher end finishes?

    As you select higher end finishes, the assembly process should become easier. The solder yield and the shelf life will also increase.

    danhargrove joined  the room.12:09 PM

    Audi McAvoy12:09 PM
    I should have asked first, is there a spreadsheet somewhere for entering questions?

    Mark J Hughes12:09 PM
    @MS-BOSS @guido.giunchi Let's revisit high speed after Elijah's quick once-over.

    @Audi McAvoy - Nope, just shoot out the questions right here

    Elijah12:10 PM
    We will be covering these main finishes, but we are open to helping you however we can:

    Elijah12:10 PM
    OSP, HASL/HASL-LeadFree, Immersion Tin (White Tin), Immersion Silver, ENIG, ENEPIG

    Mark J Hughes12:11 PM
    So the "finishes" are the last steps in PCB fabrication -- the outer layers. Most of the time that is referring to how the PCB is prepared for assembly.

    Elijah12:11 PM
    So lets...

    Read more »

  • Hack Chat Transcript, Part 1

    Dan Maloney03/11/2020 at 20:03 0 comments

    Mark J Hughes10:34 AM
    @Christoph Post 'em. We'll try to get to them if we see them.

    Mark J Hughes10:39 AM
    @Christoph Or I mean, ask now.

    Mark J Hughes10:39 AM
    I'm doing some other work but I have the chat window open.

    Elijah joined  the room.10:40 AM

    Mark J Hughes10:42 AM
    Hi @Elijah

    Mark J Hughes10:42 AM
    I think we're overachievers.

    Elijah10:45 AM
    Im always early, i usually have some type of log-in issue. But it looks like im ok today!

    Mark J Hughes10:45 AM
    Woohoo! Only because you logged in early.

    Elijah10:46 AM

    Christoph11:05 AM
    My question would be: what should an ENIG finish look like to be considered acceptable? I've had two boards with different ENIG quality - one soldered fine, the other failed. "ok" should be somewhere in between. I have pictures, too

    Mark J Hughes11:06 AM
    I believe that ENIG is covered under IPC-4552

    Mark J Hughes11:06 AM
    That's the standard that says what's acceptable and what's not.

    Mark J Hughes11:06 AM
    The problem is two-fold.

    Mark J Hughes11:07 AM
    One, it's not all that wettable -- which means that the solder tends to ball up rather than spread out over the surface.

    Mark J Hughes11:08 AM
    Secondly, if the coating is too thick, the intermetallics might not form, which relates to wettability, but also to reliability.

    Mark J Hughes11:08 AM
    Even if it wets, without intermetallic formation at the interfacial layers, your parts will start to fall off the board during thermal cycling.

    Mark J Hughes11:08 AM
    Upload the photos and we'll take a look.

    Christoph11:11 AM

    Christoph11:11 AM

    Mark J Hughes11:11 AM
    Can you show me a less blurry image of number 1?

    Mark J Hughes11:12 AM
    Also -- do you have pictures of these boards with soldered components on them?

    Christoph11:12 AM
    I'll try - the thing is the kids aren't in bed yet and tend to cause some trouble when I get out the good stuff

    Mark J Hughes11:13 AM

    Mark J Hughes11:13 AM
    Well I'll just riff a little until then.

    Mark J Hughes11:13 AM

    MS-BOSS joined  the room.11:13 AM

    Mark J Hughes11:14 AM
    When solder-paste is heated, the first thing that melts is the flux. It spreads out over the surface in a process called "wetting"

    Mark J Hughes11:14 AM

    Amr Yassin joined  the room.11:14 AM

    Mark J Hughes11:14 AM
    Wetting is determined by the adhesive and cohesive forces between a liquid and its surroundings.

    Mark J Hughes11:14 AM
    The surface roughness comes into play too.

    Mark J Hughes11:14 AM
    Anyways -- the first thing that melts is the flux.

    Mark J Hughes11:15 AM
    It spreads out over a surface and chemically de-oxidizes is.

    Christoph11:15 AM
    does the dispensed paste blob get flatter as the flux melts or will it stay in shape?

    Mark J Hughes11:15 AM
    Then at a slightly higher temperature, the solder metal melts and begins to flow over the surface.

    Christoph11:16 AM
    (actually I can answer that myself by recalling what I see when I solder on a hot plate - it gets flatter)

    MS-BOSS11:16 AM
    It's also bad if the solder dissolves enough of the gold so that the joint then comes into physical contact with the nickel layer. This cannot happen during normal reflow process, but can be done during manual soldering...

    Mark J Hughes11:16 AM
    The "blob" will move along metal surfaces until it is stopped by Solder-Resist.

    Mark J Hughes11:17 AM
    So, as the solder becomes liquid, it begins to react with the other metals present. Let me see if I can grab a picture.

    Mark J Hughes11:17 AM

    Mark J Hughes11:17 AM
    At the reaction layer -- Intermetallics form -- these are essentially metal crystals that behave like pure metals.

    Mark J Hughes11:18 AM

    Mark J Hughes11:18 AM
    If the intermetallic layers get too thick (pictured above) or large crystals form in the bulk of the solder, the solder joint can fail over time.

    Mark J Hughes11:19 AM
    Now you said that the boards didn't solder well.

    Mark J Hughes11:19 AM
    That could be a fabrication issue, that could be an assembly issue.

    Christoph11:19 AM
    one did solder well - the one with the smoother...

    Read more »

View all 3 event logs

Enjoy this event?



Interested in attending?

Become a member to follow this event or host your own