You need to use EasyEDA editor to create some projects before publishing
Harmful error in datasheets and footprints
796 14
Johan_Ha 2 years ago
I made a design, which included an SMD RGB LED. I happened to choose the item C2843784 found on LCSC catalogs. My intention was to order the PCB with all SMT part mounted and for some reason that was the only component I found that JLCPCB accepted. I was in hurry, so I didn't dig deeper after alternative parts. The problem was that _the datasheet, the schematics and the footprint didn't match each other!_ This is how the footprint looked like: ![image.png](//image.easyeda.com/pullimage/7nbe5NyQRZ5iazrj0UAGJPMq3uRGNDolzxDUxNQl.png) The red arrow and the ellipse are added later to point out the error when I chatted with the customer support at JLCPCB. As you can see, the left schematics show that pin 1, 2 and 3 are cathodes. The right footprint shows _plus signs_ at pins 1, 2 and 3. _But_ the footprint shows the dot and the triangle mark at pin 1. All SMD RGB LEDs I have worked with have the triangle mark at the cathode side. So it was only natural to assume the error was with the plus and minus signs in the right footprint. Because _there was an error somewhere!_ Furthermore the datasheet of the item C2843784 shows that the triangle mark is at the cathode side! My error was to rely on the customer support at JLCPCB. They said I should follow the datasheet, which showed this image: ![image (2).png](//image.easyeda.com/pullimage/Q680BqLDFUju2NhO4mOrU35klFlOPWcUCq3Bun9X.png) So I went and ordered my PCBs, fifteen identical PCBs, each having three of these leds. And when I got them, they were wrong! _The triangle really points out the anode, not the cathode!_ It is my believe that the manufacturer has made an error placing the triangle mark at the anode side. The error is not that big, _if all datasheets, schematics and footprint thoroughly followed that!_ But now the people, who made the datasheets didn't notice the unconventional (but probably not erraneous) marking and they messed up everything. And I have 15 useless PCBs. After I pointed out the error to JLCPCB (which further reported it to LCSC), the footprint has been corrected, but the datasheet still is wrong ([check it yourself](https://lcsc.com/product-detail/Light-Emitting-Diodes-LED_XINGLIGHT-XL-5050RGBC-MS_C2843784.html)), at least at the time of writing this. I even got a thank letter from LCSC for pointing out the erraneous schematics/footprint combo, but at that point I had already ordered my PCBs with wrongly placed leds. And to me, their correction still looked wrong because the plus signs were still at the triangle mark side (which turned out to be the right way!) Lesson learned: Don't trust anything, as long as the datasheet, the schematics and the footprints and all pin numbering don't match each other in the component catalogs of LCSC or JLCPCB! Have them correct the info on all places before you rely your PCB design on the info. Don't trust advice from the customer chat. They _do_ a good job! The chat at JLCPCB responds quickly and they got LCSC to correct the footprint rather quickly. But in my case, due to production schedules, it wasn't quick enough.
Comments
Johan_Ha 2 years ago
Not until now I recognise the tiny "IC" marking in the datasheet image! There is no IC in this component. It is a simple three colour led combo. Other rgb leds _do_ have chips, which make the components addressable. Another minor error in the datasheet image is that the C2843784 has common anodes, which means pins 4, 5 and 6 should be shown connected to each other.
Reply
andyfierman 2 years ago
@johan_ha, Have you verified with a multimeter that this device really is Common Anode and not just 3 separate LEDs?
Reply
Johan_Ha 2 years ago
@andyfierman It says here: ![image.png](//image.easyeda.com/pullimage/kVZWcjGQVZTA3BFmMb3fZpChz8Ro4r8lktpu2MyA.png) And the way the faulty PCB I received is acting, I see it's a common anode thing. My PCB looks like this: ![image.png](//image.easyeda.com/pullimage/R673fnMy6TtICRTTTrRP3dMW3YOUcXb9mLfObHyt.png) If you look at U4 at the top, I expected the anode to be on the top and the cathode on the bottom. I was supposed to feed 5 V at the top, then at the bottom the three cathodes would go through resistors (and further through a sinking NPN transistor array) to the ground. The circuit was tested on a protoboard and worked fine. But now I have the _internally common_ anode at the bottom and the three cathodes at the top, which are commoned by my PCB design. My multimeter has a diode tester mode and when I connect the positive probe to each of the bottom pins in turn (and the negative probe at the top), I get red light on each. The multimeter provides a weak current and the led hardly turns on. And it's the red led which turns on, because it has a lower forward voltage compared to the blue and green ones. When I put 5 V on the top three terminals to the right (note the mirrored R, G, B marks) and ground on the 5 V terminal, all three units (U2, U3 and U4) light up with red with a slight twist to magenta. Again, because of the common anode, the red kind of captures most current and the blue shows just a dim light, the green hardly anything. No reason for the led to turn on the red at each of the R, G or B terminals, if it hadn't an internally common anode.
Reply
andyfierman 2 years ago
@johan_ha, For the purposes of clarity: Have you removed a device from the PCB and tested it in isolation to verify that this device really is Common Anode? If you have then please post a sketch clearly and unambiguously showing the led orientations in the package with respect to the pin 1 marking triangle. Please  note which way up the package is in your sketch.
Reply
andyfierman 2 years ago
@johan_ha, "The circuit was tested on a protoboard and worked fine." Did you use the same LEDs on the protoboard?
Reply
andyfierman 2 years ago
Page 5 of the datasheet is supposed to show the packaging and orientation of the device on a reel of the device. This diagram is wrong because it shows a single device in an 0805 or similar package: ![Screenshot_20220731-120202_Drive.jpg](//image.easyeda.com/pullimage/IJnwqsjbSmjdRbNNkzVkVg2NPZXbZsFlXGPlo2TA.jpeg) [https://lcsc\.com/product\-detail/Light\-Emitting\-Diodes\-LED\_XINGLIGHT\-XL\-5050RGBC\-MS\_C2843784\.html](https://lcsc.com/product-detail/Light-Emitting-Diodes-LED_XINGLIGHT-XL-5050RGBC-MS_C2843784.html)<br> <br> [https://datasheet\.lcsc\.com/lcsc/2108072030\_XINGLIGHT\-XL\-5050RGBC\-MS\_C2843784\.pdf](https://datasheet.lcsc.com/lcsc/2108072030_XINGLIGHT-XL-5050RGBC-MS_C2843784.pdf)
Reply
Johan_Ha 2 years ago
@andyfierman "Have you removed a device from the PCB and tested it in isolation to verify that this device really is Common Anode?" Well, no. But that's no issue. I'm convinced it is a common anode component, because the page says so, and my testing shows so. But even if it weren't, it would work in my PCB if it only were attached correctly. My PCB circuit would anyhow shortcut all three anodes. On purpose. I don't have the right tools for detatching the leds and re-attaching them again. If I don't get new PCBs from JLCPCB, I have to create a tool for that. A special tip for my soldering iron, a tin sucker and whatnot. @andyfierman "Did you use the same LEDs on the protoboard?" I did _not_. I used a led with the triangle at the cathode side. But the forward voltages and optimal currents are usually the same. @andyfierman "Page 5 of the datasheet is supposed to show the packaging and orientation of the device on a reel of the device." Yes. That only shows how unreliable everything is. Those components in the picture on page 5 are no 5050 components, but something smaller. This forum is on EasyEDA's site. I don't know how EasyEDA, JLCPCB and LCSC are connected with each other, but I really hope they can sort out this kind of mess together.
Reply
andyfierman 2 years ago
@johan_ha, "Well, no. But that's no issue." The questions I have been asking are because it would help your case to get JLCPCB to replace your boards free of charge if you can simply, clearly and unambiguously prove what the connectivity of these LEDs is and that it is not what is shown in the datasheet from Xinglight. BTW I have tried to find a datasheet directly from Xinglight but they do not appear to have a publically accessible website. One other point. If you could remove and refit the LEDs rotated by 180 degrees then they would function albeit with the wrong colour sequence. If the LEDs are under software control, could you salvage the boards in this way and rewrite whatever code is used to control the LEDs?
Reply
Johan_Ha 2 years ago
@andyfierman I would rotate the leds 180 degrees, if I knew how. The colours won't cause problems. I have no software yet. But I can't rewrite code to fit this faulty PCB. I would have to accept a magenta monochrome led instead of the 8 colour led I planned (or 24 bit colour, if I add PWM). And I would have to create some peculiar 10V rail, 5V rail and 0V rail to be able to use the phototransistor, which is also included on the PCB.
Reply
andyfierman 2 years ago
@johan_ha, You obviously would have to resolve the LEDs and then resolve them. You need a good quality soldering iron with a small bit, some of this (or similar): [https://cpc.farnell.com/chip-quik/smd1/smd-1-smd-removal-kit/dp/SD00632](https://cpc.farnell.com/chip-quik/smd1/smd-1-smd-removal-kit/dp/SD00632)<br> <br> and good soldering skills. [https://www.chipquik.com/store/index.php?cPath=2200&osCsid=c6te02lgec04s1fi6mgvpgab72](https://www.chipquik.com/store/index.php?cPath=2200&osCsid=c6te02lgec04s1fi6mgvpgab72)<br> <br> That said you should contact JLCPCB support to
Reply
andyfierman 2 years ago
That said you should contact JLCPCB support to explain that through no fault of your own and after following advice from JLCPCB, the boards are wrong because it seems that the manufacturer's datasheet is wrong. I think that under these cirumstances you could not have known this and submitted the board in good faith believing that it was correct to the information available to you at the time. Therefore I think it is reasonable for JLCPCB to refund you the cost of the boards and that they contact Xinglight to ask them to correct their datasheet. It would still be best if you can remove a LED and verify the pin out and connectivity as I explained earlier.
Reply
JLCPCBsupport 2 years ago
@andyfierman Hello all ; A little clarification about the refund process, after all the refund has a process and verification steps to follow in order to decide if JLC should refund the customers or not. First, the customer has to start a quality complaint through his order history, we recommend that the customer attach images showing the fault commited by JLC, these proves will be checked by the Quality team and then negiciate the refund. In this topic I see that the design files are wrong from the early beginig, we advice that you refer to the EDA tool that you have used (in you case it is EasyEDA) and you contact the support team to fix the PCB model of the part that you have used, maybe if you change this post topic to [Bug Reports](https://easyeda.com/forum/category/2) section of the "EasyEDA Std" category then you can get more attention from easyeda team or just send them Email through support@easyeda.com Please consider that we produce what we receive, we don't verify designing faults but we check the design specifications and it has to match our JLCPCB capabilities. It is always recommended to double check the CAD files of each imported part to make sure that you are taking the appropriate path. <br> Any questions please let us know. <br> Best regards.
Reply
Johan_Ha 2 years ago
@jlcpcbsupport > It is always recommended to double check the CAD files of each imported part to make sure that you are taking the appropriate path. I did check with the JLCPCB Support agent chat. And they advised me to follow the datasheet. Which I did. Which was wrong! It harms me a lot that JLCPCB is very thankful for me pointing out the conflict in the datasheet, the schematics, the footprint and the pin numbers. This is what they wrote: > For this case, we really appreciate your kind reminder for the error shown on the datasheet. It does greatly help to us and our customers. But it didn't help me. I paid 64 $ with credit card plus 9 $ with a coupon. And got the faulty PCBs. So far they have offered me a 10 $ coupon. So we are indeed communicating about this case. What I want is a full refund of 73 $. Even better would be to receive 73 $ plus some extra for a fast delivery so I can renew my order with a fast delivery. I'm running out of time for this project.
Reply
Johan_Ha 2 years ago
@jlcpcbsupport > Please consider that we produce what we receive, we don't verify designing faults but we check the design specifications and it has to match our JLCPCB capabilities. I order a PCB from you. I pay you. I name one component as C2843784. According to the component catalog on JLCPCB and on LCSC pages, it should be an ordinary rgb led array with a 5050 case. What I get is a completely different component, a 5050 case with an IC chip which acts completely differently. Polarity is reversed when it comes to the mark. Not three separate colour channels. It was never a design fault. You assembled something else than what your component pages tell. Either you did the mistake in the assembly, picking the wrong component reel from the shelf or the error happened earlier at the component provider or at the manufacturer. But I pay you, not them. You sort it out! And compensate me! It's between you and them to sort the rest out!
Reply
Login or Register to add a comment
goToTop
你现在访问的是EasyEDA海外版,建议访问速度更快的国内版 https://lceda.cn(需要重新注册)
如果需要转移工程请在个人中心 - 工程 - 工程高级设置 - 下载工程,下载后在https://lceda.cn/editor 打开保存即可。
有问题联系QQ 3001956291 不再提醒
svg-battery svg-battery-wifi svg-books svg-more svg-paste svg-pencil svg-plant svg-ruler svg-share svg-user svg-logo-cn svg-double-arrow -mockplus- -mockplus- -mockplus- -mockplus- -mockplus- -mockplus- -mockplus- -mockplus-@1x -mockplus-

Cookie Notice

Our website uses essential cookies to help us ensure that it is working as expected, and uses optional analytics cookies to offer you a better browsing experience. To find out more, read our Cookie Notice