Announcement

Collapse
No announcement yet.

Strategy for Soldering Skills validation on fully working iphone 5s's.

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

  • Strategy for Soldering Skills validation on fully working iphone 5s's.

    13/05/2020 5:01 PM

    Any advice on maximising the use of a fully working 5s motherboard to validate soldering skills? I've acquired 5.

    BTW Would love to know the little components ie caps, coils, resistors that cause big problems on the 5s when they go missing. They would be GREAT to validate my soldering skills on.


    My approach.

    STRATEGY
    Identify IC by function
    Select non criticall IC first eg sensor IC, audio, RF comms Ics, WIFI first
    Practice on critical ICs LAST…..
    1. NAND/CPU/PMIC/Charging IC any IC required to make the phone work
    2. display ICs (need to see what I'm doing)


    PRELIMINARY
    Measurements… establishing the normal state of a phone ie voltages/resistances on a working board.
    Taking measurements
    BTW Are seated resistance measurements a valid method of measuring the 'normal' condition of the power rails on a working phone?

    Practice on non working board first (identify sure soldering disasters before I dive in!) then for a fully working board….

    PROCEDURE
    1. Remove one IC at a time.
    2. Reassemble, test for loss of function. Full retest not just the IC I removed.
    3. Restore IC / check for shorts on all relevant power rails BEFORE powering up(Don't get trolled - the natural condition of some rails appears SHORT!!!). main power rail should not be short. Other power rails?
    4. No shorts. Retest for restoration of function. And retest ALL other functionality still ok!
    That's all I can think of. Any further suggestions?

  • #2
    That all sounds fine. Your number one job is to protect the fun of board repair. If this sounds fun to you—go for it!

    I would take something pretty far away from important stuff, but something I could measure. Filters near the camera connector for example. Then the easiest chip—5s tristar. Make sure the phone can charge before and after. Then tight clearance chips, sage or meson, then chestnut.

    I would leave all the baseband chips alone because it would be hard to test baseband function

    Comment


    • #3
      A problem I found in the past as part of skills validation was that I would remove an IC eg accelerometer,
      solder it back on, do a retest then notice something else was not working.
      Eg rear camera. I would then reflow an IC near to where I was working (ie associated with the camera eg SUB PMIC)
      and camera function would return. I can only conclude that heating the board(the wrong way? too quickly?) creates dry solder joints on neighbouring components.
      Is that a reasonable assumption?

      Comment


      • #4
        I don’t think I’ve ever removed an accelerometer ic—not a failure point in any of the devices I work on, so I don’t have any experience base with the challenges of that repair. Does working in that area have anything to do with camera? Or is there some other explanation. I don’t know. Most of what you can mess up when working on a practice board is baseband rf signaling and that is challenging to test.

        But yes, for some chips working in one area can bother underfilled chips on the other side.

        Comment


        • #5
          Sorry. I neglected to mention that this particular set of skills validation was done on a number of Samsung S7/S7 edge motherboards not iPhone 5s. Rear camera, unlike on the iPhone 5s is powered from a sub PMIC (not main PMIC) which is located in the vicinity of the S7's accelerometer. The IC is located on the same side of the motherboard. Anyway,I was wondering if this was a preheating issue(the lack of) but I guess time will tell. I was inspired by the lead you gave of practicing on ICs with tight clearances so have digressed to practice on cumulus of a dead iPhone 6 I have. I KNOW I have tweezers with about the finest tips available but honestly can't find any good spot on that chip to apply them to. Awkward.

          Comment


          • #6
            cumulus. (see pics)The only points to stick my tweezers in. Hold the hot air in my right hand and aim away from the board?!
            Attached Files

            Comment


            • #7
              I would use hot air to remove that cap in the middle on the left. Then take cumulus nice and easy.

              Comment


              • #8
                PREVENTING part problems on the iphone 5s…….
                No microsoldering done today unfortunately. In the process of 'validating my soldering skills'
                I managed somewhere along the way to destroy 2 perfectly good WIFI flexes in the frame, a good battery, destroy a good screen and destroy a good motherboard. Which sent me into a spin.
                Lessons learnt…..
                WIFI flex - disconnect - fingers only - NO spludger. Spludger's uncontrolled flick
                Causes the paper thin WIFI flex to richocet and tear. So finger nails only. No tiny components near the antenna to get flicked off so feel free to use the fingernails to pull off the connector controlled fashion!
                Battery flex - do NOT bend at the base(it's a weak point that is easily damaged) - bend back the end connector (where it connects to the PCB) And now I tape and straightjacket the battery flex cable to the side of the battery!
                Constant disconnects/reconnects induces stupification. Struggling with LCD connectors and forgetting to disconnect the battery before reconnecting. Faux Pas. Retest screen IMMEDIATELY. NEVER force connectors. NEVER allow LCD cables to be stressed/flexed during reconnects(disconnects aren't an issue). Pin the top connectors firmly against the screen between thumb and index finger of hand one while using the hand two take each cable, one by one, and connect to the motherboard.

                And finally I did the following.
                3 batteries tested fully working

                3 charge port flexes tested fully working

                3 screens tested fully working

                1 fully assembled frame/battery/motherboard/screen tested fully working as fully assembled parts to test against.

                With parts issues in order, 3 lots of tested goods to validate against I'm ready to hit the microsoldering validation again tomorrow!

                And if I f*ck those up I've got two fully tested, fully functional iphone 5s's to disassemble!

                I'm going to nail this once and for all!

                Comment


                • #9
                  Those are all good observations.
                  Use fingernails for everything, or a fine general
                  purpose tweezer. Black plastic spudgers are for things like helping to separate something like an iPad mini LCD from the backplate. I rarely use them. Never use the blue plastic board breakers for anything.

                  you’re right to protect screen and battery flexes. Never let the screen “lean” on the flexes. Always support the screen and keep it upright.

                  my advice would be to work under the microscope. I avoid “dog damage” by taking off home buttons and sensitive parts under the microscope.

                  My second advice is to consider separating “screws and glues” from board repair. Yes you should be proficient in all the disassembly reassembly you described here. Great work! But for Microsoldering eliminate variables. Work on the board alone on the table and test it with a fresh charge port, screen split so you can see image without needing backlight, and a dc power supply and then battery. Nothing else. Then work on taking off tristar and replacing it with a new one being careful not to bother audio ic (underfilled) or baseband cpu (opposite side) and pull off all stickers before you change the chip so they don’t curl and pull components by charge port.

                  A 5s iCloud locked board has a value of maybe $10-$20? So think of them as disposable and get several. Chuck through them discarding it as soon as one no longer charges on usb ammeter after changing tristar and grabbing a fresh one to try again.

                  Once you can reproducibly and reliably change tristar and have the device still charge, then worry about “test all the functions” stuff.

                  last tip—from the kinds of problems you described here, I can predict that you are going to rip the antenna cable that runs along the side of the board when you go to remove the shields, or forget to take it off and have a hard time with the shields. Don’t do that. Be careful of the glue and take it off with fine tweezers under a microscope.

                  good luck and have fun!

                  Comment


                  • #10

                    Thanks for the advice. Just want to be sure I understand.
                    Always support the screen and keep it upright
                    Upright in my hand at all times vertical (90 degrees) to the motherboard which is lying on the bench?

                    screen split so you can see image without needing backlight
                    Scratching my head. Don't understand. Could you clarify?

                    Motherboard with screen and charge port flex connected to DC Power Supply(via squid cable)
                    Observed behaviour…..
                    Nigglies and confiscation
                    prompt to boot from pins
                    Are the following behaviours consistent with your experience?
                    Observed behaviours
                    (1) 1% battery level then after a while goes up to 15% WTF???21(Tried this with two fully working motherboards) It's connected up to DCPS for heavens sake… surely it should read 100% and be done with it?
                    Further Inconsistent behaviours
                    Challenges getting motherboard to boot reliably, touching the pins the right way ?consistently shorting BUTTON_TO_AP_HOLD_KEY_CONN_L to Ground
                    Adhoc reboots(hate this one) The phone doesn't stay reliably booted/stable from DCPS! I had this problem with Samsung S5s(though far worse) and went through 2 other power supplies, one like the type sold on the ipad rehab store and another high end Rigol digital DCPS revered by Dave Jones on EEVblog & otherElectronics enthusiasts BUT with the same problem. BUT, the phone's operation is rock solid when connected to it's battery! Haven't been able to resolve the DCPS problem

                    When the iPhone boots to the lock screen. Press the software home button(as provided by accessibility assistant, I've deliberately set up accessibility assistant for this and powering off the phone), screen goes blank and the phone reboots. If instead the phone boots directly into to the home screen I breathe a sigh of relief.

                    Comment


                    • #11
                      Screen--yes, support it so that it is always somewhat upright, never lying flat open.

                      Split screen. This isn't necessary, it's just a suggestion. When a phone can turn on with nothing on the screen, how do you know if the problem is no image ,or no backlight? Separate the test screen into two pieces by pulling the backlight piece away from the LCD, then you can shine a light from behind very easily.

                      Re: 1% battery level then after a while goes up to 15% WTF??
                      You are booting a phone on DCPS. The DCPS has no gas gauge, so the phone has no way to know battery percentage.

                      Challenges getting motherboard to boot reliably, touching the pins the right way ? It is challenging to boot a phone from DCPS, connect the pin under the microscope or use charge port to reliably boot the phone.

                      Re: "The phone doesn't stay reliably booted/stable from DCPS!" Well you don't want to leave a phone on dcps for more than a few minutes. It doesn't have gas gauge and phones don't like that. DCPS simply asks "can you boot yes or no" If yes, then put it on a battery. A phone on DCPS will drop after three minutes due to lack of gas gauge. It sounds like you are trying to ask the phone to stay on DCPS for too long. Any cheap power supply is fine for iPhone work.

                      The phone should always boot to the lock screen with no trouble after an effective prompt to boot. if that is your problem, then resolder your squid leads.

                      Comment


                      • #12
                        Ok thanks for that. I've pretty much mapped out all the ICs except the RFs on the bottom of the board see pictures(oh and I didn't label CPU but I think that one is prtetty obvious!) but I can't seem to find any equivalent to the Tigris charging IC on this motherboard. It's driving me crazy.
                        Attached Files

                        Comment


                        • Jessa_the_Professa
                          Jessa_the_Professa commented
                          Editing a comment
                          tigris function before iPhone 6 is part of the job of the PMIC

                        • Jessa_the_Professa
                          Jessa_the_Professa commented
                          Editing a comment
                          Sage and cumulus are reversed. Chestnut is the "display pmic" backlight driver is just called backlight driver

                      • #13
                        Thanks for that. So Tigris is a power IC and Tristar is a USB logic IC? Uploading corrected version of bottom board IC identification
                        Attached Files

                        Comment


                        • #14
                          Today I desoldered SAGE off a fully working iphone 5s. I took the motherboard minus SAGE and reassembled it back into the frame. Of course there was no touch but I was able to confirm that the nearby underfilled WIFI IC was still working - well there were still signal bars(this is a big bear bug of mine - sometimes, in the past, when I'm working on area of the board nearby stuff stops working)*. However, I was surprised to see that the reassembled screen had vertical lines running down it. At first I assumed I had not properly connected the display and attempted 3 times with the same result. After taking the screen and connecting to another fully working motherboard the vertical lines disappeared first go. Back to the original motherboard and, after reballing SAGE and hot airing back on, touch functionality was restored AND the vertical lines had completely disappeared. (all other functions of the phone tested fully working). Definitely was a motherboard issue, originally I thought maybe the hot air impacted Chestnut and I intended to do a reflow… but now? Any ideas?


                          *I do this because I am very much concerned about the way I am applying hot air to the board. Of course I employ the strategy(where I can) of aiming the hot air out and away off the board) but also need to know that the way I'm aiming that nozzle, the way I approach the board with it eg am I coming in too fast? is not having detrimental impact on other areas
                          Attached Files

                          Comment


                          • Jessa_the_Professa
                            Jessa_the_Professa commented
                            Editing a comment
                            yes that is normal for a board with no sage or no meson

                        • #15
                          As I said the vertical lines on the screen disappeared when I restored sage. Is this expected behaviour when removing and restoring Sage?

                          Comment

                        Working...
                        X