One bad Arduino doesn’t spoil the barrel
— October 1st, 2010
Recently we have seen a blog post from a customer regarding manufacturing issues. Although it is possible that a handful of boards with minor defects can slip through the cracks while they are QA’d at the factory we dispute that all of them show similar issues. There are 150 thousands Arduino boards and 20 thousand more are being manufactured right now and still the percentage of returns is around 0.3% (this includes some boards that have been clearly tampered with that we ask the distributors to replace anyway)
The site http://wtfmoogle.com/?p=1508 has a series of claim the we would like to respond to before the FUD spreads too much. We are sorry if he received a poorly finished board but you can’t claim that thousands of boards are all the same.
We Arduino always try to work for our customers and the community to bring new products that simplify the life of people and make access to technology simpler and simpler. While we do this there can be issues and a few boards can come out not perfect.
We urge you to contact us and we’ll send you a replacement straight away. The reseller has also offered the customer an immediate replacement via a postage paid return envelope as well as a refund.
Issue 1: There is a small leftover on the edge of the PCB that looks like a small “tooth”. We apologise for that, it’s clearly something that went unnoticed when the boards have beet tested. It can be solved with a small file otherwise we’re happy to replace the board directly with no cost for the customer.
Issue 2: There seem to be a reddish orange colour on one of the sides of the board. That unfortunately it’s due to the new packaging. Since we chose a combination of paper and ink that would have a low environmental impact but we realise that in some cases the board can scratch the box if the shipping has been a bit rought. We have read reports from many other users that had no such problem with their boards. User “James Lewis” posts as a reply to moogle’s blog ” My headers are aligned quite well. There is no exposed copper on the Vias near http://www.arduino.cc. No orange marks anywhere on the edges. I will agree the overall routing is much rougher than other PCBs. Your board looks to be the exception and not the norm.”
Issue 3: One of the headers doesn’t seem to be properly soldered it. There is a small tolerance in the position of the header but nothing too crazy. Again I’m sure this is a board that was obviously placed in the wrong bin by the person who inspected them but the board does work properly as they are all tested individually.
Issue 4: There are octagonal pads around the pins. According to moogle this generates more noise than a round pad and it’s a bad bad design decision. Well it might be a problem if this was a very high frequency board, like a radio but the Arduino isn’t. The board has passed the CE and FCC testing for radio emissions with flying colours. This fact is also disputed by user “Grumpy_Mike” on the forum who is involved in FCC certifications.
On the other hand we chose the octagonal pads because they provide a much better grip than the round ones therefore providing more mechanical robustness to the board.
Issue 5: The problem here is “the use of non standard sound pads for surface mount omponents with no silk screen markings to tell you that a part should go there and it is not a test point.”. It’s important to note the picture is one of an unpopulated PCB from an Arduino Duemilanove and not from an UNO. since we don’t sell blank PCBs for the Duemilanove it has been manufactured elsewhere and not by Arduino.
In any case this is a non issue. you would need the footprint silkscreen on the PCB if you where to assemble the boards by hand but since these boards are assembled by a robot we removed them to clean up the board from unnecessary visual clutter.
Issue 6: the picture shows the edge of the board next to the “www.arduino.cc” marking. According to moogle these show exposed copper. In our opinion the just show that the soldermask has thinned around the vias. Again this is something that can happen and I want to reiterate that this is obviously a board that should have stayed in the factory in the “Defective” box.
In the blog user GoldieDpimp who works in a factory that builds board writes “I could take the time and show you how while they are not what we would consider target, they are considered “passable”. There are international standards for these types of things and I don’t see anything that doesn’t look like it wouldn’t be “passable”. It could look better, but I’ve seen worse flying around in airplanes…”
Thank you for bringing this to our attention, and again – We Arduino always try to work for our customers and the community to bring new products that simplify the life of people and make access to technology simpler and simpler. While we do this there can be issues and a few boards can come out not perfect.
We urge you to contact us and we’ll send you a replacement straight away. The reseller has also offered the customer an immediate replacement via a postage paid return envelope as well as a refund.
Nobody is perfect but we are nice people who contribute a lot to the community. Please keep that in mind and contact us if you’re having any issue: team(at)arduino.cc or contact your reseller so we can best assist you.
October 1st, 2010 at 21:51:43
Not sure if FUD is really whats happening here, there is clearly a difference of opinion over what is acceptable in fit and finish.
Up to the user to point out issues, up to the manufacturers to defend their position or rectifiy issues 🙂
October 1st, 2010 at 22:01:43
arduinoblog Please would you also address the problem here: http://forum.pololu.com/viewtopic.php?f=3&t=3083
October 1st, 2010 at 22:18:57
Thanks for the response, I was thinking about holding off but due to this response I will be buying one as soon as they come back in stock!
Your an awesome team, I look forward to doing business with you.
Respectfully,
TuxTux
October 2nd, 2010 at 00:07:58
I can’t help my self but ask (sorry)
How did you test the boards I’ve seen with the headers. I ask purely out of curiosity. If the customer can’t attach a shield how did you to test it ?
Over all I think you guys are still the best at what you do. Slip ups happen.
October 2nd, 2010 at 02:14:35
The point commenter “Jess” makes on the Arduio’s critic forum here may mean more to the discussion than pure differences of opinion regarding manufacturing defects:
“$30 is a pretty high price, considering the TI LaunchPad costs less than $5 to purchase, probably even less to manufacture.”
Absolutely right. TI is a hugh commercial manufacturing enterprise that generates some exceptional products. The size of a company’s supply chain and global distribution determine their pricing flexibility. To compare Arduino to TI is an incredibly imprecise comparison. That is like compariing the economy of France to the Dominican Republic. TI’s size is why they can drive the per unit cost of their comparable device to $5.
And for now there is a peaceful coexistance between the embedded and controller development community and TI. So the toolchains and other development supplies are relatively easy to obtain. But that relationship doesn’t always have to continue in its current form. Consider what MySQL and OpenOffice contributors must feel about their relationship with Oracle after seeing what happened to OpenSolaris. Open source projects are the only way of ensuring that the cost for programming development in embedded space is commoditized to zero.
The Arduino system has, at its core, the ability to significantly improve the product and its design through the contribution of its users. It is an open source experience. But let’s, for once, set aside the open source moniker and just acknowledge that this is also a scientific endeavor.
Here is one of the primary responsibilites of scientists: If you recognize poor work, you need to expose it. But if you are a scientist of any worth, you will not only critique other’s work but advance the science by doing *better* work than those you criticise.
So the challenge is to Mr Moogle. You written that “My cat can solder headers straighter then this.”
Perhaps you and your cat can meet the current production and design demands of the Arduino community. What is the name of your company, and when can we expect pricing?
October 2nd, 2010 at 03:48:27
The guy complains about off-center vias and exposed copper but never mentions that the board fails to work. He also mentions octagonal pads are noisier than round pads, but I’ve never heard this to be the case. Even if it were, it would be at RF frequencies and would not affect functionality.
Open source hardware manufacturers are not mass producers. We’re all basically artisans that are practicing a craft. Unlike mass producers that have automated visual inspection, a large portion of the manufacturing process for indie developers are still done by hand and manually inspected.
The rest of his complaints are just opinion, like how there are no reference designators on the board. That’s a user interface decision and in Tokyo Hackerspace, many of the members never even look at reference designators.
One thing missing any mention is how the software was crafted to be user friendly and much simpler than standard microcontroller toolchains. It’s really opened up embedded programming to the masses.
If he expects perfection, perhaps he should design his own board, sell it for $30 retail, and let the other open source hardware developers scrutinize it. I, for one, would gladly take on that task.
October 2nd, 2010 at 04:01:45
arduinoblog Hold up, The reseller offered them a replacement AND a refund. So a free board. That sounds like a bonus to me, :rolleyes:
October 2nd, 2010 at 05:41:59
Arduino have set a standard which surpasses what you would get from most overseas board shops. Their fabrication is excellent. This is obviously a flawed board which will undoubtedly sneak through once in a while.
Arduino obviously are appalled by this and have decently offered replacement or refund. This is the sign of a company dedicated to their customers.
This blog was unnecessary. To do so was a smear against a great company. Original Poster should have dealt with customer service and received a replacement before publicly flogging Arduino over a relatively minor mistake. Not cool.
October 2nd, 2010 at 06:04:55
I really appreciate you guys taking the time to address all of the recent comments, which unfortunately is what I read first before reading your sites official blog post. I can honestly say that I feel 100% confident in you guys, and I plan on purchasing my first Arduino this weekend 🙂
Keep Rockin it!!!
October 2nd, 2010 at 06:07:49
Even if this is not the fault of Arduino this blog post should have started with “We apologize for…” and “In the future we will do … to avoid …”. Also have someone proof read these things.
Looking forward for Arduino Ethernet.
October 2nd, 2010 at 06:35:03
I would be really sad if this spoil the barrel.
I am really interested in how you develop and manufacture the cards so why don’t you tell us a bit about it? Although I know some who did, I suppose you don’t produce them in the living room 🙂
BR
Erik
October 2nd, 2010 at 07:04:23
The article by at wtfmoogle is certainly a bit off. If you buy something new and it has an issue, just contact the seller – not spray bad karma around the internet bad-mouthing it – especially considering the effort Arduino team goes into for everyone to benefit.
October 2nd, 2010 at 08:57:50
I can see both sides of this story; on one side you and the Arduino team has poured your heart and soul into orchestrating a surprise release timed with global product availability. I can see the pride and effort that went into designing what is clearly the best Arduino yet.
But do realize that the expectations are set very high; the Arduino Uno comes in this fancy new box, oozing with nationalistic pride. “Made in Italy” it says proudly on the front, along with “Quality Tested” on a fancy color sticker. The text on the included sheet further proclaims how well engineered and tested the product is, laying it on about as thick as imaginable.
With that kind of setup, if anything at all is imperfect about the contents, it can easily lead to disappointment and criticism. While I’m not the type to complain about cosmetic details or post pictures pointing out every identifiable flaw, I have to admit my out-of-box-experience included some of the same type of negative impressions.
Unlike products where the PCB is hidden inside a case, the Arduino board is up front and visible. Any misaligned components are glaring faults. The silk screen is attractive and very distinctive, especially the back side, but that also means that it suffers more if it appears thin or is covered in some sort of sticky residue.
October 2nd, 2010 at 16:36:05
Massimo and co,
I think it’s shame you have to spend time defending yourself against crap like this. There’s always a trade-off between polish and price, and I think you’ve generally made the decision correctly. Anonymously criticizing the weakest points of someone else’s efforts is easy to do; I hate to see you underappreciated.
Thanks for creating the Arduino and for all the work you do.
Brandon
October 2nd, 2010 at 19:23:56
I see nothing wrong with the original blog post. It does not state anywhere that the board is functionally defective – it only points out that that particular board has a number of finish deficiencies, placing the PCB somewhat below the blogger’s quality standards. Stating that can in no way be considered spreading FUD, m’kay? Frankly, if the pictures are real, I tend to agree with him on several points. One can only hope that’s indeed the exception and not the norm.
On the other hand, I certainly salute the willingness of the involved parties to discuss the issue and replace the board in question. As long as it doesn’t all go the infamous “you should hold it right” route…
October 4th, 2010 at 14:42:58
I want to point out that my Uno exhibits none of these issues; and I am quite pleased with the new Uno.
October 4th, 2010 at 19:12:03
I ordered three uno boards last week and got them today.
To my dissapointment, misaligned headers seems not to be part of quality testing. I received tree uno today. #264243 had the analog headers about 1.5 mm off compared to the power header. #264200 had the digital header off by about 2.0 mm. #264234 was almost OK, some 0.1 mm off in the analog in, and a shield could be fitted with some force.
For comparison my duemilanove is perfectly aligned.
2 of 3 unos had the mentioned “tooth”. All had quite rough edges (“routing”?), and were redcolored from the box.
Since they were imported I dont feel a return is justified. The headers are already bent into straight position, and I’ll sand down those edges tonight.
I’d just like to make you aware of the issue, and that there are more than one board out there with quality control issues.
Enough of the whining : ) Thanks for a great product and community!
October 4th, 2010 at 23:56:44
I shudder to think of (possibly semi-) functional (if aesthetically flawed) Arduinos lying in a reject bin (or being destroyed after being returned). I for one would love to see the equivalent of Sparkfun’s “dings and dents” section selling the few problem/reject boards at a reduced price with a “no guarantees of functionality” disclaimer. For us impoverished students (and jobless post grads, ahem) something like a $10 dollar discount for wonky headers would be the deal of the day. Plus you get to monetise “scrap” boards that would otherwise cost money to put right or recycle. I would not expect those boards to take very long to sell if Sparkfun’s “dings and dents” section is anything to go by.
I agree that it’s not fair that this is causing bad publicity, but anyone significantly miffed will hopefully take the trouble to look up what the fuss was about and come across this post. On the flip side it shows how high a regard the community has for Arduino products, and it’s a reputation I can’t see being besmirched overnight with a few bad boards.
D.
October 5th, 2010 at 11:09:54
I believe the OPEN in Open-Source is greyed.
Someone mentioned that they were surprised to see the new boards BEFORE seeing a Gerber or other design file of the UNO.
I can understand wanting to beat the ‘ho-hums’ to the buck, but you’ve got the brand. And it feels awkward.
I hope this is not becoming EX-open-source.
I believe in Arduino, and I believe they will, and most likely are doing the right thing.
October 5th, 2010 at 13:08:59
Arduino is committed to open source hardware and will always be. It’s one of the main contributors to its definition http://freedomdefined.org/OSHW (in the first 10 names 3 are Arduino team members)
We were doing this before it was fashionable and we have create a good part of its current market.
The Arduino UNO it’s even more open than before because we got rid of a proprietary microcontroller (the FTDI chip) and replaced it with an open source firmware.
We announced the boards on sunday pm (us time) and the files where live with the new site: eagle files and source code, all of it.
In any case, nowhere in the open source licenses nor in the creative commons it say when you have to release the files. It just says you have to make them available and this is what we did in the exact moment we were speaking at Maker Faire.
I think our position is clear…
October 5th, 2010 at 14:13:48
@massimo Of course, you guys released the EAGLE sources and software sources under open source licenses.. but the whole development of the new boards wasn’t open. They just came as a “surprise”.
Given the size of the Arduino community, I think that a more open design process would really be valuable. I’m imagining a feature request website with voting, daily updated EAGLE (or *better* KiCAD) file repositories and a more transparent idea->product process.
As an example, would you consider Ubuntu as an open project if they would just give you the sources after a new product release and don’t collaborate with the community?
October 6th, 2010 at 03:43:30
To put things into perspective for all of the whistle blowers, I recently purchased a server motherboard form a major and reputable hardware manufacturer (that will remain unnamed) that had much more severe issues. A large capacitor was completely blocking one of the memory module retaining clips from opening (design flaw), some of the PCI-Express contacts and memory slot contacts were bent an stuck in the slots, and the BIOS was different version than what the board was labeled with so the processors purchased with it wouldn’t work. They eventually replaced the board after several weeks of correspondence with technical support. Both the board and service were awful, and that’s what they do for a living as professionals.
Looking at the issues shown with the Uno I see minor inconveniences at worst. Also everyone involved is willing to remedy the situation by replacing everything immediately (without any charge). I don’t see what the fuss is all about? Compared to dealing with many other electronics manufacturers this sounds like way better than par for the course. Heck, the service offered to remedy the situation should be used as an example of how to treat customers. Instead a bunch of ungrateful bloggers are ragging on people who don’t deserve it. If the official hardware doesn’t float your boat then you always have the option to download the design files and have your own boards produced!
If the problems shown are isolated incidents and that’s what you say then I believe it. You guys are alright in my book. I look forward to using the Arduino Uno and the updated IDE in a few upcoming projects.
October 9th, 2010 at 19:48:19
Stefan,
I have tried to contact you by email to fix this issue twice this week, if you read this comment, please report whether you solved this with your distributor already or not. We are making sure anyone that got a board that wasn’t according to their expectations are getting replacements.
/d
October 10th, 2010 at 04:09:25
What do you guys do with the ones in the defective box?
Can I get one cheap? 😛
October 14th, 2010 at 10:44:06
While it is totally reasonable to mention defects, the original complainant hasn’t helped themselves by mixing a genuine defect (misplaced header) with very minor blemishes and some nonsense about octagonal headers being ‘noisy’.
The badly soldered headers is a genuine problem and looks like it would prevent a shield plugging in, so yes this is a functional defect – perhaps the quality control only checks electrical and not mechanical functionality?
I personally wouldn’t worry about the other minor blemishes myself, and I suspect most people wouldn’t even notice.
However the statement that octagonal vias have issues with noise is bizarre – given the sub-mm distance scales involved this would only start to be detectable at something like terahertz frequencies (at which the fibreglass board material is a totally unsuitable dielectric material anyway).
Has anyone established with certainty that the board in question is not a clone board? – after all the unpopulated board pictured must be a clone since these are not available from Arduino themselves.
October 16th, 2010 at 15:34:12
I have to second the ideas of a more collaborative open-source development and design process, as well as the idea of selling the “dings-and-dents” (with an appropriate disclaimer, of course).
I see an open design process as being something that could only help the community, if it were handled properly. Selling the dings-and-dents would be useful as well for the reasons already mentioned. I tend to see the Arduino as a microcontroller development board, not as the final end-product – the only shield I own is a protoshield with a small breadboard.
Finally – regardless of what is done on the next major release, please strongly consider breaking with the past, and fixing the header misalignment bug that has been there since the “beginning of time”. Whether by actually correcting the alignment, or by adding a second set of headers (as some other clone makers do – though I think correcting the headers completely would be the best choice), I believe that while there might be some inconvenience, in the long run it would be better for the community as a whole.
Such a change would allow for standard 0.1 in PCB perfboard to be used for easy shield creation, which would allow for a wider community of developers (including those who lack the experience to design to the “wonky” PCB, or can’t afford a prototyping PCB, for instance) to create shield designs. While I myself don’t use shields, I don’t discount the impact that they have had on the Arduino as a whole, as well as how they have enabled the marketplace to develop. Fixing this issue, while it would introduce some minor growing pains (which I am sure could be worked around if needed), would likely only expand the number of designs available, and the marketplace for the Arduino as a whole.
I see this as only a good thing.
October 30th, 2010 at 16:40:25
[…] Update: Here’s a direct response from the Arduino blog. […]
November 15th, 2010 at 07:56:33
While i agree that every customer might want a perfectly finished product, i think its a criminal waste if a board ends up in a rejected bin just because of aesthetics reasons considering how much time and energy was spent in making the board. Perhaps we could still sell them at a lower price.
January 12th, 2011 at 06:28:00
The moral of the story is :
1. Arduino is still cool, and you guys are great
2. You would definitely need to adopt a more open and collaborative board development process in the future. I hope to see Arduino Tatu soon. (Tatu = Three in Kiswahilli)
3. dings-and-dents, dings-and-dents, dings-and-dents
4. Header alignment, and more header alignment ! ! dudes ! ! !
Regards,
Zinahe A.
January 12th, 2011 at 08:31:22
I continued doing more research on this topic and found out (sorry to say) more mishaps around the UNO.
http://www.flickr.com/photos/mightyohm/5051676871/?reg=1&src=comment
http://forum.pololu.com/viewtopic.php?f=3&t=3083
http://www.on8rth.be/?p=169
And it looks like, you guys have not been listening the request to get the community involved before making surprising decisions like the Arduino Uno SMD (http://arduino.cc/blog/2011/01/08/arduino-uno-smd). You need to find a way to get it in to your heads that you are here because of the community.
February 17th, 2011 at 12:16:11
[…] PCB on one of the edges, which looks like a small tooth. This seems to be the same issue as posted here. There is also a reddish orange color on the side of the board – again the same issue as […]