Talk:Surface-mount technology

Quad in line...
The chips listed under quad in line are not actually quad in line. quad in line literally means 4 parallel rows of pins, which does nto apply for this... Pleas change this to something more meaningful. 124.197.101.39 (talk) 16:28, 24 September 2013 (UTC)

Pictures
It would be very helpful, at least until all the minor articles are made (QFN, TSSOP, etc.), if links to picture of the different packages were available. This would add to the list. --Matejhowell 21:09, 16 February 2006 (UTC)

Updates to "Assembly Techniques"
I agree with everything else. --Adx 13:27, 31 March 2006 (UTC)
 * Solder paste is actually composed of flux and tiny balls of solder, all of uniform size. (This would make a neat picture.)
 * Brass stencils are also frequently used (although falling out of favor).
 * In my experience, adhesive dots are rarely used for double-sided assembly (only when having to place heavy components on both sides of the board). Adhesive dots are required to allow SMT components to be flow soldered onto the solder side of a through-hole board.
 * Minor, but the pre-heat zone also allows volatile components of the solder paste to evaporate, to avoid boiling during soldering.
 * Vapour phase reflow is rapidly falling back into favor due to RoHS (lead free) soldering.
 * From what I know, only a small percentage of boards are washed, perhaps someone can back this up with some published stats. Most processes rely on "no clean" fluxes which do not need cleaning.

I agree about the seldom-washing of PCB's. We use an EMC Stencil Cleaner and only wash boards that have misplaced parts or which had errors in solder screen printing. If nothing goes wrong we do not wash boards at all.

At times though when something has been wave-soldered we will wash it. But again, not every time. --lsjzl 14:58, 17 May 2006 (EST)

The technically correct answer is that it depends on which flux is used, and that depends on several other numerous factors. No clean fluxes are generally used in most cases with acceptable results. Reworked PCBs are always cleaned. We clean boards since we do RF work and mainly since it can affect the dielectric of air, but I think to clean or not to clean is an application specific question. --dyermaker504 22:35, 5 January 2016 (UTC)

Splitting smd technology from smd packages ..?
Seems the surface mount technology page is becomeing more a list of used packages rather than a presentation of the technology as such. Maybe it's suitable to split it into two seperate sections?

I agree, on a side-note isn't the pacakages section a good candidate for a table??? Perhaps even including the pictures of the individual packages.... Zerodamage (talk) 11:13, 9 July 2012 (UTC)

Adding a few words about automated optical inspection
I think it would be wise to mention that visual inspection can be done either by human, or by automated inspection machines.

Why is MQFP listed in Non-packaged devices?
I don't understand why MQFP is considered to be non-packaged. From the article:
 * Non-packaged devices (although surface mount, these devices require specific process for assembly)

213.142.209.184 11:25, 24 January 2007 (UTC)

Is that advertisement?
Somebody please check if the stuff written in the "two terminals" section below the heading "Tantalum capacitors" is advertisement. I suspect it to be and because I don't know the matter I ask. 20:00 GMT+1 —Preceding unsigned comment added by 93.220.47.120 (talk) 19:00, 10 February 2010 (UTC)
 * You mean because there are manufacturer names listed? The way it works is usually that one manufacturer starts using a certain labeling scheme and others adopt it by convention. As you can see in the Aluminium capacitor section I added, CDE uses the same labeling code as Panasonic. As long as they are not standardized, e.g. by JEDEC or EIAJ, I'm afraid there is no other way of referring to a labeling scheme than naming the manufacturer. --AndreKR (talk) 13:26, 26 August 2010 (UTC)

"Assembly Techniques" is still wrong
Whoever wrote this off-base article section felt strongly enough about it to delete loads of critical comments about it. It still has problems.

This section is flawed, off-base, and so out-of-context that it should be deleted and rewritten. Assembly Techniques? As in plural? It only mentions one assembly method, as if that is the only one that exists, and it certainly is not, and worse, that whole section seems to be from the perspective of someone who has made boards for NASA and no one else.

This is how it was in the first place, then after my comment above and his deletion of my comment above, this was tacked on to the end of the article:

"However, most electronic assemblies are made using a "No-Clean" process where the flux residues are designed to be left on the circuit board [Benign]. This saves the cost of cleaning, speeds up the whole process, and reduces waste."

It needs a rewrite, not a tack on.

Approximately 99% of all SMT assembly done is done for low-priced consumer electronics (computer mice, etc.), and such boards are not only usually NOT washed after reflow soldering, but many of them aren't reflowed at all. The quick and dirty technique of surface mount wave soldering using no-clean solder is extremely common. Sure, a PC motherboard has a lot of fine-pitch components and high speed electronics that requires precision and cleanliness, so those boards are done honestly with stencils, reflow, and water-soluble washing, but for everything else, the manufacturers cheap out and use no-clean solder and wave solder everything whenever possible.

Surface mount wave soldering also allows the PTH components (which are inevitable in most any design short of a RAM stick) to be soldered at the same time, eliminating an expensive secondary soldering operation.

This article only mentions wave soldering of surface mount boards only as a secondary operation after the boards have had the first soldering operation done with a stencil and paste, and only needed because glue has been used. Again, this is totally off-base, a whole lot of SMD boards are soldered without any screen printing at all! They are soldered once and once only with wave soldering. At one time, that was the only way to get SMD and PTH components soldered in a single operation, and it still is a better, more reliable, and more flexible process than paste in hole reflow soldering of PTH components, as it doesn't require a stencil, it solders better, and a great number of PTH components (pots, switches, capacitors, LEDs, etc.) cannot take the heat of the reflow process, nor the water washing that comes with the use of water soluble flux.

Please stop denying this fundamental problem with this article section and fix it or delete it.--208.127.100.147 (talk) 10:32, 16 February 2010 (UTC)

This is Wikipedia. Everyone cite your sources and write the text so it reflects them. No original research, no opinions! 85.23.49.22 (talk) 16:35, 29 April 2010 (UTC)

Consider adding a link to the CircuitInsight page for some industry guides for using adhesive http://www.circuitinsight.com/programs/54494.html 50.247.247.81 (talk) 19:03, 14 November 2017 (UTC)

Saturn IV
This article contains a reference to Saturn IV, which it says is a launch vehicle that used surface-mount technology. Saturn IV is a moon of Saturn, according to Wikipedia. The reference should be to Saturn IB, which is in fact a launch vehicle (again, according to Wikipedia). I made this change to the page, and it was instantly reverted. Why? Isn't this exactly the point of Wikipedia, to allow a reader that notices an error in an article to fix it? —Preceding unsigned comment added by 134.134.139.76 (talk) 00:42, 16 February 2011 (UTC)


 * If you want your edits to be considered with more care by editors then it is a good idea to explain your changes so any reviewing editor can see that you are not just some vandal changing things randomly. It is unfortunate that it has to be this way, but there are so very many vandals that make little changes like this and probably at least 99.99% of the time they are pure and simple vandalism. For the 0.01% of times, like this time, then someone gets upset when they were trying to do the right thing and get reverted. You did the right thing to come here to the talk page and explain the situation. Anyhow, I have made the change (with a proper edit summary) and hopefully everyone will take time to verify before undoing it again. HumphreyW (talk) 08:40, 16 February 2011 (UTC)

please update this article with a section for "ic Surface-mount companies in Europe" wikilord09:31, 8 February 2012 (UTC) — Preceding unsigned comment added by 188.25.51.232 (talk)

Resistor size graphic
What's going on with the graphic linked in the article, which I've included to the left here? It appears to be accurate at 0402 and 0603, assuming those are SAE sizes, but as the list goes on it becomes apparent that the size labels are comprised of common metric values for package sizes. Take the 6332 size for instance, which in metric would mean 6.3mmx3.2mm, yet it's clearly much larger than the 1cm square provided for reference. I believe the packages are all SAE sizes and somebody labelled them erroneously with the common metric sizes in ascending order, but I'm not touching it as I don't claim to be an expert on the matter... Rezurok (talk) 23:17, 8 May 2012 (UTC)

It should be noted that SAE stands for Society of Automotive Engineers and has nothing to do with sizes of semiconductors. Semiconductors fall under the jurisdiction of the EIA in the US. By the same token, the so-called metric sizes fall under an IEC or JEDEC specification and should be noted as such. Here is an excellent article on the subject.

http://themetricmaven.com/?p=454

The fact that American industry refused to adopt the specifications of the world bodies that standardised on metric units, caused the eventual demise of much of the American printed circuit industry. The foolish travel the road to their own destruction and deservingly so.

A companion article can be found here:

http://themetricmaven.com/?p=2451

Any reference to SAE must be changed to EIA or else it is technically wrong. — Preceding unsigned comment added by Ametrica (talk • contribs) 17:52, 8 December 2013 (UTC)

Indeed it is wrong; The originally uploaded image showed correct size ratios for metric sizes but was replaced by a mistakenly resized image. Presumably the person replacing it thought it was referring to SAE sizes but it is clear from the file history (and legend) that it is intended to show metric sizes. Given the continuing popularity of the SAE sizes it might be best to just replace it with an image showing relative sizes in both SAE and metric... — Preceding unsigned comment added by 129.215.90.106 (talk) 10:30, 25 May 2012 (UTC)

Indeed you are wrong! Can you provide an SAE standard that applies to semiconductors? This shouldn't be hard to locate if in fact there is an SAE semiconductor standard. — Preceding unsigned comment added by Ametrica (talk • contribs) 16:07, 18 January 2015

That P#%!% of S"#%"#% drawing just cost me real-life money because of its just plain wrong, and I in my stupidity choose to go with it... So I am going to do something about it right now Zerodamage (talk) 16:00, 7 July 2012 (UTC)

Well, this is your fault for relying on Wikipedia as your source. Next time consult the actual JEDEC standard. — Preceding unsigned comment added by Ametrica (talk • contribs) 16:07, 18 January 2015

Just to revise, the original uploader of the picture had the scales right, its the person who scaled it that made the error, notwithstanding his reasons for doing so 16:22, 7 July 2012 (UTC) — Preceding unsigned comment added by Zerodamage (talk • contribs)

I made an SVG version with both metric and imperial size-codes and compoarison squares, it should be to scale and correct codes, if anyone find any errors please let me know... Zerodamage (talk) 18:23, 7 July 2012 (UTC)


 * Well done, Zerodamage. My only complaint is that no explanation was given anywhere about what these numerical codes actually meant.  Had to go look it up elsewhere on the Internet to find out, which is ridiculous.  I have now added a paragraph explaining, but please, please, please correct me if there are any mistakes here!  This is a venue in which I fear to tread.   KDS 4444  Talk   09:02, 24 December 2012 (UTC)

The text states that LGA packages are only used with sockets. This is not correct. They are also (mostly) soldered using SMT. They offer lower height profile than BGA but can be less forgiving of substrate warpage. — Preceding unsigned comment added by 205.149.130.18 (talk) 13:57, 25 September 2012 (UTC)

SMD LEDs?
How is it that when a person types "SMD LED" on a wikipedia search he/ she gets directed to the article on LEDs, where no mention is made of SMD LEDs... And then, when he/ she does a search on SMDs, he/ she gets taken to this article, where no mention is ever specifically made of SMD LEDs anywhere? What if I wanted to know what an SMD LED was? Or even what it stood for? There is nothing here and nothing there. Ah, the wonder of redirects. Was there ever an article on SMD LEDs?? Should the redirect (if it is to stand) go to LEDs or to here? And why, when no mention is made of them either place?? KDS 4444 Talk   08:55, 24 December 2012 (UTC)


 * The LED is the same no matter if it is packaged as an SMD or a through hole part. This is relatively the same as the different light_bulb bases for different uses. MadCowpoke (talk) 08:44, 28 December 2012 (UTC)

PACKAGES
Why are the standard dimensions in millimetres put in the supplemental position when in fact they are universally the primary dimension? No one in the world recognises the EIA inch sizes, except the dwindling number of American uses. JEDEC doesn't recognise the EIA sizes, only the metric sizes.

http://themetricmaven.com/?p=454

http://themetricmaven.com/?p=2451

These two articles should bring the situation to light. — Preceding unsigned comment added by Ametrica (talk • contribs) 17:58, 8 December 2013 (UTC)

SMD also means "Solder Mask Defined"
SMD stands for "Solder Mask Defined" when referring to surface mount footprint definitions. Although previously used as the acronym for Surface Mount Device, most manufactures now use SMD to mean Solder Mask Defined as it relates to footprints.

There are generally two configurations for SMT pads, SMD (solder mask defined) and NSMD (non-solder mask defined). Generally NSMD pads are prefered since the etching process of copper has a tighter tollerance to the solder-masking process, however there are some exceptions. One such exception I have experienced is in case sizes below 0402 and fine pitch BGA where strength and copper cohesion to prepreg is desired over flexibility and copper tolerance.

Duane Benson has also mentioned that the Beagleboard team found at 0.4mm pitch BGA, SMD pads experienced less bridging than NSMD pads. He also talks about using SMD definitions for corner pads in situations where strength is required to prevent copper lifting. http://blog.screamingcircuits.com/2010/09/smd-vs-nsmd.html

Here are some other articles that talk about SMD and NSMD footprint definitions that may be helpful. TopLine BGA SMD AND NSMD LAND PATTERN(pdf) http://www.topline.tv/drawings/pdf/SMD_vrs_NSMD_Pads/BGA_SMD_vrs_NSMD_PADS.pdf Lattice PCB Layout Recomendations for BGA Packages (pg.13-3 pdf) http://www.latticesemi.com/lit/docs/technotes/tn1074.pdf TE Connectivity Application Guide for 0201-size SESD devices http://www.te.com/content/dam/te/global/english/products/Circuit-Protection/knowledge-center/documents/ap-surface-mount-0201-size-sesd-devices.pdf Skrivitor (talk) 02:10, 11 December 2013 (UTC)

Three digit E96 resistor code
The full code is given in the resistorguide.com page linked in the external links. SpinningSpark 15:07, 21 February 2015 (UTC)

External links modified
Hello fellow Wikipedians,

I have just added archive links to 2 one external links on Surface-mount technology. Please take a moment to review my edit. If necessary, add after the link to keep me from modifying it. Alternatively, you can add to keep me off the page altogether. I made the following changes:
 * Added archive https://web.archive.org/20070411093709/http://www.fairchildsemi.com:80/products/discrete/pdf/sot89.pdf to http://www.fairchildsemi.com/products/discrete/pdf/sot89.pdf
 * Added archive https://web.archive.org/20120225095507/http://www.fairchildsemi.com/products/discrete/packaging/TO_268.html to http://www.fairchildsemi.com/products/discrete/packaging/TO_268.html

When you have finished reviewing my changes, please set the checked parameter below to true to let others know.

Cheers. —cyberbot II  Talk to my owner :Online 12:43, 17 October 2015 (UTC)

Does the article still have "multiple issues"?
The warning template says: As for the 1st issue, it still applies. But the second is also a very old one and no specific issues have been pinpointed in the Discussion. I 've made a few improvements (incl. citations) wrt the cleaning process. Whether a misinterpretation or inappropriate citation still exists, it is better to mark it with a proper in-line warning. Therefore I plan to replace the "multiple issues" template with the "refimprove". Any objections?--Dipa1965 (talk) 09:40, 3 October 2017 (UTC)
 * This article needs additional citations for verification. (July 2008)
 * This article possibly contains inappropriate or misinterpreted citations that do not verify the text. (August 2012)

Apostrophe required
"By employing SMT, the production process speeds up, but the risk of defects also increase due to the components miniaturization and denser packing of boards." This requires an apostrophe after "components". NotYourFathersOldsmobile (talk) 02:21, 7 January 2018 (UTC)

Nitrogen is not inert
The 'Hot gas' section of 'Rework' states that hot gas 'can be air or inert gas (nitrogen).' Nitrogen is not inert, so this is wrong or misleading. 80.44.118.104 (talk) 16:51, 16 March 2020 (UTC)
 * Nitrogen is not a noble gas, but is sufficiently inert for the purposes of working on surface mount technology and is referred to as such by reliable sources. Constant314 (talk) 23:00, 16 March 2020 (UTC)

Section about related trade shows?
I'm thinking it may be useful to add a section about SMT-related trade shows, such as as Nepcon and APEX. I believe this can provide some context about the industry and community behind the technical aspects. --Alan Islas (talk) 16:43, 25 May 2020 (UTC)


 * That would be promotional use.Constant314 (talk) 17:54, 25 May 2020 (UTC)


 * Thanks for the quick response Constant314. There is a trade shows short section in consumer electronics that gave me the idea, and I didn't think about potential issues of promotional content. I was planning to mention all the notable shows, especially if they already have WP articles, so not one would be featured in particular. Furthermore, as you know, some of these events are also academic conferences where technical papers related to SMT are presented and discussed. But I can see how the line can get blurry, so I won't further push for this idea at this point. The feedback is appreciated. Regards, --Alan Islas (talk) 21:55, 25 May 2020 (UTC)