Announcement

Collapse
No announcement yet.

LED Resistor Calculations ...

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #61
    Re: LED Resistor Calculations ...

    My panel has LEDs in series and parallel, why wouldn't your design work ? No, it doesn't need to be soldered, but it will conduct heat far better soldered to a large copper pad.

    Comment


      #62
      Re: LED Resistor Calculations ...

      Originally posted by diif View Post
      My panel has LEDs in series and parallel, why wouldn't your design work ? No, it doesn't need to be soldered, but it will conduct heat far better soldered to a large copper pad.
      OK so, you're getting more confusing ... in your last post you said, "The thin bit you see is the only part of the board apart from near the edge there is no used copper, when i say used, I mean used for dissipating heat, power is coming top to bottom." (whatever 'Top to bottom' actually means...)

      I read that to mean that no copper was used on the board for dissipating heat.

      BUT NOW you say, "No, it doesn't need to be soldered, but it will conduct heat far better soldered to a large copper pad."

      So which is it? Do we use copper for dissipating heat or don't we?

      MY first run of this board had all of the center pins sharing a single copper plane. Little did I know at the time, was that the center pin on the LED was also the Anode. So they didn't work until I drilled out the vias that connected the center pin to the copper plane on the back of the board.

      Why can't I seem to get a straight and complete answer as to the proper way to design a fucking board for LED arrays? If there is a known correct way to do this, where is it documented? I'd probably get a lot more out of reading something specifically intended to instruct engineers as to how to properly design a PCB that contains a ton of heat-generating LEDs.

      My current design uses a copper plane as well as an aluminum backing to dissipate heat but not heat that comes from the soldered pin on the LED, but rather the heat passed from the LED onto the copper plane through the solder mask then from the cooper plane onto the aluminum backing, and I added vias to connect the copper plane to the aluminum and I'm waiting to find out if they can actually do that or not. I don't know how they construct vias so I'm not sure.

      Now, if I were to have each LEDs center pin soldered to the board, I would have to electrically isolate each "heat dissipation zone" for each LED so that none of the anodes are connected to each other. And for the life of me, I can't see how that would be a whole lot better than what I came up with because in that design, the heat would only be dissipating through a smaller piece of copper rather than sharing the entire size of the board for heat dissipation.

      But I might be wrong in my thinking where that is concerned perhaps it is better to have small areas of copper attached to the LED for heat dissipation but then that begs the question, how does the heat from those smaller copper regions get to the aluminum backing? It would have to be through standard heat transfer by way of some thin layer of material between the copper area and the aluminum.

      NOW, if they are able to connect the copper area to the aluminum with via's then I can guarantee you that my design of passing heat through the solder mask then to the aluminum by way of dispersed via's will run a hell of a lot cooler than soldering the center pin to smaller areas of copper then moving heat through an insulating layer ... as my design would allow the heat to move along connected metallic paths. Like this ...

      Comment


        #63
        Rim e: LED Resistor Calculations ...

        Appologies, I'm on my phone as my Internet is down and autocorect changed to "shouldn't"
        Of course it should be soldered.

        Comment


          #64
          Re: Rim e: LED Resistor Calculations ...

          Originally posted by diif View Post
          Appologies, I'm on my phone as my Internet is down and autocorect changed to "shouldn't"
          Of course it should be soldered.
          So I just received the aluminum boards which are the same as the pics I've been showing you in terms of the copper layer etc.

          And something I noticed about them right off the bat trying to solder components onto them is that I had to hold the heat gun on the board for a LOOOOONG time before the solder even melted but then once that first one melted the rest just melted instantly with only a second or two of applied heat and then of course, the entire board was hot as hell when I was done so I rinsed it under some cold water ...

          Then, then I used the iron to solder two wires to it on the pads that supply power, I could not melt the solder on the whole pad at the same time so there is DEFINITELY heat transfer happening in a serious way from each and every pad to the aluminum somehow ... don't know how exactly ... but there is no question that it's happening.

          I don't think I'm going to need to use that design I have where I connect the aluminum to the ground plane with VIAs ... I think it will be overkill and unnecessary.

          Comment


            #65
            Re: LED Resistor Calculations ...

            Also, with everything powered up for over 10 minutes, the aluminum back plate is at 117 degrees C ... that's pretty damn hot! And at 12 volts, it's drawing 1.3 amps.
            Last edited by EasyGoing1; 05-25-2021, 04:36 PM.

            Comment


              #66
              Re: LED Resistor Calculations ...

              Thanks for sharing this mate! It helped me a lot

              Comment


                #67
                Re: LED Resistor Calculations ...

                Originally posted by BadmanRasta View Post
                Thanks for sharing this mate! It helped me a lot
                Which part, specifically?

                Comment


                  #68
                  Re: LED Resistor Calculations ...

                  Is it really 117c not 117f?
                  If it is that hot I wonder what the junction Temp of the LED is.
                  Typical 1W white LED has junction temp max of around 110c.

                  https://www.photonics.com/Articles/M...erature/a34316
                  Last edited by budm; 05-25-2021, 09:18 PM.
                  Never stop learning
                  Basic LCD TV and Monitor troubleshooting guides.
                  http://www.badcaps.net/forum/showthr...956#post305956

                  Voltage Regulator (LDO) testing:
                  http://www.badcaps.net/forum/showthr...999#post300999

                  Inverter testing using old CFL:
                  http://www.badcaps.net/forum/showthr...er+testing+cfl

                  Tear down pictures : Hit the ">" Show Albums and stories" on the left side
                  http://s807.photobucket.com/user/budm/library/

                  TV Factory reset codes listing:
                  http://www.badcaps.net/forum/showthread.php?t=24809

                  Comment


                    #69
                    Re: LED Resistor Calculations ...

                    Originally posted by budm View Post
                    Is it really 117c not 117f?
                    It's running a little cooler right now cause it's cooler in the room than it was earlier...

                    Comment


                      #70
                      Re: LED Resistor Calculations ...

                      Originally posted by budm View Post
                      Is it really 117c not 117f?
                      If it is that hot I wonder what the junction Temp of the LED is.
                      Typical 1W white LED has junction temp max of around 110c.

                      https://www.photonics.com/Articles/M...erature/a34316
                      The thing with these boards ... and I can't seem to get a straight answer from the sales reps at PCBWay ... but there is definitely some kind of layer between the pads and the aluminum so that no matter which pad I put heat on, that aluminum has to heat first before the pad will be hot enough to even put solder on. Im gonna make a video to demonstrate what I'm talking about ...

                      But I'm convinced NOW that the proper design for this kind of board would be an isolated copper area for each anode, since heat transfers out to the aluminum immediately without needing to physically connect the copper with the aluminum ... I just don't know how they do it.

                      Comment


                        #71
                        Re: LED Resistor Calculations ...

                        Originally posted by EasyGoing1 View Post
                        But I'm convinced NOW that the proper design for this kind of board would be an isolated copper area for each anode, since heat transfers out to the aluminum immediately without needing to physically connect the copper with the aluminum ... I just don't know how they do it.
                        NO, the PROPER way is as I showed and detailed above, definitely no large ground plane.

                        Comment


                          #72
                          Re: LED Resistor Calculations ...

                          Originally posted by diif View Post
                          NO, the PROPER way is as I showed and detailed above, definitely no large ground plane.
                          The problem is ... nothing you've said so far makes any sense. So your method is still a mystery.

                          After working with the aluminum boards, I think this layout will be the most efficient.

                          Dark red is power in as well as heat dissipation for the anode's of the first LEDs in each row.

                          Dark green is heat dissipation for the anodes of the next two LEDs in line, and the gray area is ground.

                          Comment


                            #73
                            Re: LED Resistor Calculations ...

                            I even included a photo to aid with the explanation.

                            If you want to improve your design further, then add another couple of vertical breaks so you can remove the trace to the resistor. All anodes and cathodes with then be on large copper pads.

                            Comment


                              #74
                              Re: LED Resistor Calculations ...

                              Originally posted by diif View Post
                              I even included a photo to aid with the explanation.
                              But the actual traces were not clear in the photo. Compare your photo to the ones I've been providing ... there is no guesswork involved in my photos, your photo was clear as mud.

                              Originally posted by diif View Post
                              If you want to improve your design further, then add another couple of vertical breaks so you can remove the trace to the resistor. All anodes and cathodes with then be on large copper pads.
                              I would agree with that, if there were some kind of precision in the actual transferring of heat out to the aluminum backing. But the fact is, it doesn't matter which pin is producing heat, the aluminum will dissipate the heat from any pin universally throughout its entire surface area ... so additional dividers are essentially useless.

                              Comment


                                #75
                                Re: LED Resistor Calculations ...

                                Originally posted by BadmanRasta View Post
                                Thanks for sharing this mate! It helped me a lot
                                Originally posted by EasyGoing1 View Post
                                Which part, specifically?
                                These kind of posts are usually by a spam bot to look legitimate and fly under the radar, but I usually wait untill he makes the actual spam post and then nuke from orbit...
                                "The one who says it cannot be done should never interrupt the one who is doing it."

                                Comment


                                  #76
                                  Re: LED Resistor Calculations ...

                                  Originally posted by EasyGoing1 View Post
                                  But the actual traces were not clear in the photo. Compare your photo to the ones I've been providing ... there is no guesswork involved in my photos, your photo was clear as mud.



                                  I would agree with that, if there were some kind of precision in the actual transferring of heat out to the aluminum backing. But the fact is, it doesn't matter which pin is producing heat, the aluminum will dissipate the heat from any pin universally throughout its entire surface area ... so additional dividers are essentially useless.
                                  There's no precision but it is science getting the heat out, bad deign is bad design and that will create hot spots at the base of those LEDs.

                                  I don't know the size of your panel but I have one here 24cm square, 150 0.5W LEDs, run at 60W it's warm not hot to the touch. It doesn't use crappy no name LEDs though, there are no traces on my panel. Just + and - pads the full width of the board with no dividers.

                                  Comment


                                    #77
                                    Re: LED Resistor Calculations ...

                                    Originally posted by diif View Post
                                    There's no precision but it is science getting the heat out, bad deign is bad design and that will create hot spots at the base of those LEDs.

                                    I don't know the size of your panel but I have one here 24cm square, 150 0.5W LEDs, run at 60W it's warm, not hot to the touch. It doesn't use crappy no name LEDs though, there are no traces on my panel. Just + and - pads the full width of the board with no dividers.
                                    The final panel will be 120mm X 60mm. They had two options for the aluminum thermal conductivity. The default selection was 1.0W/mK (whatever mK is ... ) I chose the other option which is 2.0W/mK which meant to me that it can move more heat than the first option.

                                    These boards that I've been testing with are only 90mm x 40mm with the 1.0W/kM aluminum and they simply get too damn hot. Worst-case scenario, if the final panels are too hot, I can leave out a couple of rows of LEDs ... these things are damn bright as it is ... one or two rows less won't be that noticable.

                                    Comment


                                      #78
                                      Re: LED Resistor Calculations ...

                                      After a little research i have found that heat affects the Vf of the LEDs in turn drawing more current, that's why the pads need to be the same size so there is no temperature differential, that trace to the resistor has to be changed to another pad.
                                      I think a panel your size should run warm at 7.5W it's an 1/8th the size of my 60W.

                                      Comment


                                        #79
                                        Re: LED Resistor Calculations ...

                                        Originally posted by diif View Post
                                        After a little research i have found that heat affects the Vf of the LEDs in turn drawing more current, that's why the pads need to be the same size so there is no temperature differential, that trace to the resistor has to be changed to another pad.
                                        I think a panel your size should run warm at 7.5W it's an 1/8th the size of my 60W.
                                        I've been looking at how these kinds of PCBs are manufactured, and I've also been looking at the designs from professionals who put on youtube, their own high-powered LED custom-designed circuit boards, and I've really learned a lot from those videos.

                                        I guess CREE makes the most popular high-powered LED that people seem to use ... one of them, in particular, can run a SEVEN AMPS, but the guy ran his at only 700 milliamps with damn good light output ... I'm thinking that doing something like that might provide me with the ability to get the same amount of light from an LED that would hardly kick out much heat at all because it's being driven WAY below its tolerance levels...

                                        Also, I saw one video where a guy compared the actual thermal output from an aluminum PCB vs. a standard PCB and the aluminum PCB didn't perform any better really than the standard PCB ... which I found interesting ... if you click here to see the video, I set the time index to the relevant spot on the video.

                                        I need to look into it more and look at the costs of the LEDs etc., but I might be looking at a complete re-design of this panel.

                                        Comment


                                          #80
                                          Re: LED Resistor Calculations ...

                                          But are the thermal via boards electrically isolated like the aluminum boards?

                                          Comment

                                          Working...
                                          X