I hesitated to start this topic, I was hoping for a solution in the case, but I've struggled for my right a week now, so I thought you should know.
In short. I ordered 15 identical PCBs, 40 mm * 40 mm. Each contained 3 SMD RGB LEDs and some resistors and one phototransistor, everything assembled. The led was this item, manufactured by Xinglight.
(The link goes to JLCPCB's own component catalog).
The product page clearly shows an rgb led array. What I got instead was leds with IC chips, not addressable, but leds performing a rotating colour scheme just by adding voltage to one pin.
It is quite possible the manufacturer has given wrong product info. It is quite possible that the intermediate component provider has mixed two different components and their codes. It is quite possible JLCPCB maid a mistake when installing a component reel in their machines. I have clearly pointed out these plausible scenarios to JLCPCB, but they refuse to investigate it any further.
So far they have showed nothing that would imply that I have made an error. And they take no responsibility for sorting out whether they did an error in the assembly or the component provider or the manufacturer made the error. In either case I'm entitled to a full refund.
The part designation does indeed seem to suggest that these are RGB LEDs and the schematic shows 3 diodes. However, there seems to be no indication that these are addressable (e.g. WS2812). The datasheet is in Chinese and unreadable to those not knowing the language (including myself) so frankly difficult to know what they are exactly.
I did find a variant,
XINGLIGHT XL-5050RGBC-WS2812B, which I imagine is probably the addressable version:
^https://www.4project.co.il/product/558280
One might have perhaps hoped that JLCPCB could have at least explained that. Its a mistake that anyone might have easily made (myself included).
Whatever the case, this highlights why it is essential to check carefully what one is buying. If in doubt, leave out, find another source, ask questions prior to purchase or buy at ones own risk. Caveat emptor.
I appreciate that you may have wanted these boards ready built for a reason, but hopefully the LEDs can still be swapped out for addressable versions?
Ok, I shall delete this thread as soon as I get confirmation on that.
[edit]
Couldn't find anything yet, which would confirm it. Anyway, I think I know what component they mistakenly assembled instead. I have no product page for it, but I have a garden light, which most probably contains a similar led. It's a device with a tiny solar cell and either a rechargeable battery or a supercapacitor. Probably the latter. And a tiny circuit, which charges the capacitor when sun is shining, and discharges the capacitor to the led, when it's dark enough. No extra logic for PWM is included in the garden light. All that is taken care of in the tiny led.
Here I test the leds with my multimeter. Well, I can't test the individual leds, I can only apply voltage on any of the three pins on the triangle side and ground on any of the pins on the other side. So this is not an adressable rgb led, nor is it an ordinary rgb led array with available pins for each electrode.
Yes, but as I wrote, this one is not addressable. If anyone can find a product page describing this "garden light" led, it might help me explaining to JLCPCB what kind of led they mistakenly assembled. I really want to help them understand the problem.
I don't want addressable versions. The PCB is designed for ordinary 5050 smd rgb led arrays. I popped off the leds from 13 of the boards. I need to get ordinary leds, some SMD soldering stuff and I guess I'm alright.
See Page 2 - it's clearly a 6- pin, 3 individually pinned LED package.
and a PCB Footprint (in Easy EDA)
which is consistent.
There's little doubt in my mind that they should be able to comprehend that. Good luck, though, the representatives I've interacted with over there are challenging.
C
I finally got a confirmation from LCSC that the datasheet was wrong. And the datasheet is written by Xinglight, the manufacturer.
Now I have 15 boards, each of which has 3 of these wrong components. I could still use them, but I have to rethink the circuit and the functionality of my product.
And I'm still disappointed of how JLCPCB handled my case. I'm the one who found the error in the first place. I'm the one who later found out that the error was in the datasheet and that the component was totally wrong, not only backwards pictured. I'm the one who started the process to get the info on JLCPCB and LCSC pages corrected, as well as the Xinglight datasheet corrected. JLCPCB is the one who got my 73 $ and sent me faulty boards. And JLCPCB compensates me with a 40 $ coupon, which I have to use on my next order from them, to be able to benefit from the compensation at all.
And JLCPCB would totally agree that my description is correct, we've discussed it long enough. This is how it went, and this shows what their policy is.
Looking at your board from this thread: Resoldering SMT components I'd probably consider making R38, R40 etc physically larger and hand soldering all the components. That way you also ensure that you have a working prototype and could then consider outsourcing the manufacturing if you can't face hand soldering all 15 boards.
Omitting the prototyping stage, that is working only from data sheets, can be risky if you are not familiar with the components (as you have discovered) especially if the data sheets are not easy to interpret or otherwise misleading.
I'm sorry for letting this become a cross posting thing. My intention was in this thread to point out the customer policy of JLCPCB and in the other thread only discuss a matter of how to do correction work on SMD components. 6v6gt referred tot he other thread. I responsed here, my bad.
I warned you in #4 that I might delete this thread. I hope no one get upset if I do. Right now I chose to edit my post #9.
