1. 34

  2. 6

    You can check if your MacBook is performing thermal throttling like this:

    pmset -g thermlog

    Leave it running for a bit while you work and if throttling is happening you’ll see more messages appear. The first one appears to be a lie.

    My MacBook Pro 15” 2015 was throttling on a hot day last week and at times was running at 20% max processor speed.

    With the anti-bad-stuff software I’m obliged to run on this company laptop, this left less than zero cycles for anything else. This meant missing keystrokes, several minutes (!) to launch an app (like Terminal) and some subsystems not starting up (like WiFi).

    There were only two solutions that worked: Hold the laptop directly in front of the air con vent, or move to a different room with a lower temperature.

    I tried pointing a massive fan at it but this didn’t seem to bring the temperature down enough.

    Any other ideas appreciated! (Dust level inside is low)

    1. 3

      Try replacing the thermal paste it can dry out and conducts far less heat.

      1. 3

        Definitely try cleaning off & repasting the CPU. It’s not too hard to get to: https://www.ifixit.com/Guide/MacBook+Pro+15-Inch+Retina+Display+Mid+2015+Heat+Sink+Replacement/55922

      2. 5

        I have a 2019 MBP and it definitely has thermal issues.

        What sucks is that it needs all 100W allowed by the USB-C/Thunderbolt spec…and most hubs/monitors output at most 85W, including the hub mentioned in TFA.

        I have a Thunderbolt monitor that provides 85W, which I’m saddened isn’t enough. It would be great to only have on cable.

        Anyone got any recommendations for a Thunderbolt dock that outputs 100W and supports a Thunderbolt monitor (most of the ones I’ve seen only have DP/HDMI out). I know I can get an adapter cable but I’d prefer not to.

        Really though I’m just waiting for the ARM MBP 13”, which shouldn’t require 100W…

        1. 4

          I really like the author’s suggestion that computers should detect when their fans are ineffective and offer actionable advice. It wouldn’t be difficult or expensive to add the hardware, would help computers last longer, and would give people better insight into their computers. I’ve never heard of any computer shipping with this feature, though.

          1. 3

            You probably don’t need any custom hardware. Macs can already control their fans through the firmware interfaces (most computers can, though there’s a depressing lack of standardisation, so figuring out which fan control is for the CPU is tricky sometimes). They can also retrieve CPU temperature at any given time. It would be fairly simple to have a tool periodically spin up the fans and detect the rate of change in CPU temperature. If you do this straight after a boot / resume when the machine has been off for a while, the baseline CPU temperature can give you an indication of the ambient temperature. If the fans aren’t cooling adequately, tell the user to unblock the air vents and then run again. If that still doesn’t work, tell them to take it in for fan cleaning.

            1. 1

              One advantage of having a dedicated airflow sensor is that you could gain some insight as to why the fans aren’t cooling the CPU. If air is moving but the temperatures are consistently high, it’s probably the best sink or thermal paste. If the fans are consistently ineffective, it’s probably dust. If they are usually effective but aren’t right now, the laptop might just be on a soft surface that’s blocking the airflow.

          2. 3

            I’m curious if one could achieve similar results by blowing air in through the hinge (where the vents are located). But then you’d miss out on the satisfaction of seeing those squeeky clean fans at the end.

            1. 2

              I would think that might cause the fans to spin. I seem to recall hearing somewhere that this can be a bad thing, since it might damage the fans (somehow? maybe my spinning them in the wrong direction, or at an RPM that is too high?), but maybe that was just some unfounded rumor?

              1. 3

                Ah, I didn’t mention this in my post. I’ve read similar concerns that making the fans spin at too high of a RPM can damage them. I don’t know how valid the concern is, but just in case, I did use a toothpick to stop the fans from spinning while I used the compressed air to clean them.

                1. 1

                  It depends on the machine. I doubt it’d be a problem on Macs because the Apple Tax allows them to cut fewer corners than a budget box, but (AFAIK, IANAEE etc) the issue is that you generate voltage in the motors of things like fans or 3d printers by moving parts around via external force, and this can hurt (sensitive | cheap | poorly isolated) circuitry.

                  1. 1

                    Yeah i would not spin the fans up… overspeeding them has caused mine to not sound good a while back so now i just hold them stationary while using the air… seems to work much safer. also, considering a fan is basically a generator in reverse i would think it could create some overvoltage on the board? Maybe apple thought of that too?

                    1. 2

                      I remember a buddy of mine who fried his PC mainboard by using compressed air from a generator to remove dust.

                      1. 2

                        Though I have no doubt this could happen, I cleaned out an old ThinkPad T61 with a leaf blower and it continued to run well for 3 more years.

                        1. 1

                          Might’ve been around 2005 or earlier. They probably have circuitry now that can prevent this kind of damage now that hasn’t been available in cheap motherboards from yesteryear.

                          1. 1

                            Compressed air from a generator normally has condensation that shows up at the destination side as water vapor condensed onto the pcb. Which probably means you can end up shorting things.

                            Compressed air is the best option, you can’t really protect against bridging circuits outside of the pcb.

                            The leaf blower probably just didn’t condense air, not something I would personally use. Seems too good a way to let the magic smoke out of circuits.

                2. 3

                  This is where the bad-keyboard MBPs are nice: they come with a retroactive four-year keyboard replacement guarantee, and because everything is glued together you don’t just get a new keyboard, you get a fresh battery and I’m sure they blow out the dust too. I’ve had the keyboard replaced twice now on my 2017 laptop, and there’s still time for one more refresh once the keys stop working.

                  1. 1

                    I just heard that the local Macbook repair company has been given stricter rules from Apple: apparently it used to be enough to say “the keyboard doesn’t work right”, but now you have to list the keys that aren’t working, and they’ll first try to clean them with pressurized air. There also can’t be any signs of intentional damage. I guess they suspect people have been abusing the replacement program to get new batteries.

                    1. 1

                      Yep, I had this repair done last week and the service report includes:

                      Quote and Assessment Summary
                      Verified issue reported: YES
                      Symptom/s: Sticky > key
                      Ran Apple diagnostic MRI: Passed
                      Removed, Cleaned, and Replaced effected keycap, issue persists
                      Visual inspection shows no evidence of liquid contact or corrosion
                      Isolated to faulty keyboard
                      Resolution: Replace Top Case
                      Warranty Status: Service eligible under Apple Limited Warranty, AppleCare, and/or Australian Consumer Law.
                      Quote to Repair device: $0.00 Customer Approved: Yes
                    2. 1

                      I didn’t know that, cheers!