AirPlay Mirroring requires a 2011 or later MBP (with a Sandy Bridge CPU).
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
What have you done with an early Intel recently?
- Thread starter Project Alice
- Start date
- Sort by reaction score
Maybe you can stream media from an earlier Mac. Mirroring the whole desktop or using an AppleTV as additional monitor requires a 2011 or later (Sandy Bridge) Mac. But AirParrot might solve this.
As a winter weekend project, I wanted to try something with the cooling situation on my A1261 MacBook Pro.
It’s not really a secret how the aluminium MBPs have the tendency to run pretty hot at even the best of times. But I got kind of tired dealing with a really hot lap after using it for awhile — even to the point where it was slowly scalding my thigh.
This MBP was the one I only started using last year after sourcing a green-dot board for it. As with all my Macs, I use Noctua NT-H2 paste, which is consistently pretty good at doing what it’s supposed to do. Even so, the MBP was still warm to hot in very localized spots directly beneath the heat sink bridge, particularly on the left side (or, the side nearest the Magsafe adapter). I’m guessing a far-infrared camera would have shown three localized hot spots where the CPU, GPU, and memory controller are located.
So I thought about the underside of the heat sink bridge and took stock of how there are only small portions of the bridge which actually make direct contact with the bottom case. Since the bridge, as designed, can shunt only so much heat to the fans, there aren’t other ways for heat conduction and dissipation to take place.
Except, well, perhaps there might be!
Borrowing from user discussions on improving heat dispersion in current-gen Silicon MacBook Airs, I bought a 10x10cm sheet of 1mm-thick silicone thermal pad. I went ahead and cut out pieces of the thermal pad to fit the underside areas of the heat sink bridge (which otherwise don’t make contact with anything, aside from poorly circulated air) to help to move some of the heat generated to the aluminium case, rather than make the fans do nearly all of the work.
If all worked as planned, the overall temperature of the case would feel warmer (including around the palm rest), but this manner of heat dissipation would also take longer for the fans to rev up as high to move out heat and to “de-localize” the hot spots underneath the laptop. In other words, I’m using the aluminium case more directly to keep “hot spots” from getting so hot so often and to make resting it on my lap more tolerable.
Before:
After:
There were a couple of tiny spots I could have filled in, but for sake of testing, I left them alone. I didn’t use thermal pads in the zones next to the fans, as those are sealed by a thin foam rectangle factory-attached to the bottom case (and sealing away air within from the rest of the case). I might come back to those later.
(While I had the heat sink bridge out, I also opened the fan assemblies, cleaned out a year’s worth of dust, and added to the spindles a drop of the lubricant I’ve been using for refurbishing fans.)
I ran a screen cap of the temps before I put the system into hibernate, and about an hour after I woke the system from deep sleep. The ambient room conditions were the same:
Before: After:
The numbers to pay especial notice to are these:
I wrote this post with the laptop on my lap, and for a change, it isn’t slow-cooking me. The fans aren’t moving as quick as before, while the top case and overall bottom case are perceptibly warmer. The above data are the best I can do without access to an infrared camera.
tl;dr: Not so hot!
EDIT to UPDATE, late into day 2: Even as this is still pretty early, post-modification, there is still a noteworthy drop in overall temperatures around the areas of heat generation — namely, the CPU/GPU areas and their respective heatsinks. In the 7-day line chart on iStat Menus, the drop in overall temperatures at the typically hottest area — the GPU heatsink — is evident (the break in the line is when the system was in hibernate mode and while I was adding in the thermal pad modification).
Overall, the CPU die is about 12°C cooler; the CPU proximity is about 6.5°C cooler, and the GPU Heatsink(0) is about 7°C cooler; other generators of heat are between 6 and 7°C cooler. Meanwhile, the three heatpipe sensors are about 2–5°C cooler, and the two case-related thermal sensors — Palm Rest and Skin Proximity — are 6°C and 4°C cooler, respectively (i.e., not as much of a drop as at the areas where that heat is generated). What this suggests is the aluminium case is now taking on (and dissipating passively) more of the overall heat from the thermal generation, borne out by an overall warmer case, and lacking the localized scalding-hot spots.
Put another way: the difference between the GPU Heatsink(0) and Skin Proximity before the thermal pad modification was about 28°C; since the modification, the difference is about 25°C — or about 3°C (which is bang-on with the difference between the 7°C cooler GPU Heatsink(0) and the 4°C cooler Skin Proximity reading). In Fahrenheit, that’s about 5.5°F.
As I add this section via the A1261, I’ve managed to push the Skin Proximity temp up to 42°C under load (Interweb, Transmission, Console, Preview, and a few OS X lightweight utilities like Dictionary and Calculator). Before the thermal pad mod, this would have been closer to 46°C — the latter being above the threshold for risking second-degree burns (after over an hour of constant exposure) or third-degree burns (after over 90 minutes of constant exposure); in localized spots, such as where the cross-members of the heatpipe assembly met the bottom case), that temp was likely much higher). And this might finally explain how I received such deep burns over the summertime as I used this laptop on my bare leg lap.
Unloved orphan of tl;dr: This thermal pad mod is doing exactly what I hoped it would (better cooling, less hot to the touch, and hopefully extending life on the GPU and CPU with less heat lingering) — all for less than $10!
It’s not really a secret how the aluminium MBPs have the tendency to run pretty hot at even the best of times. But I got kind of tired dealing with a really hot lap after using it for awhile — even to the point where it was slowly scalding my thigh.
This MBP was the one I only started using last year after sourcing a green-dot board for it. As with all my Macs, I use Noctua NT-H2 paste, which is consistently pretty good at doing what it’s supposed to do. Even so, the MBP was still warm to hot in very localized spots directly beneath the heat sink bridge, particularly on the left side (or, the side nearest the Magsafe adapter). I’m guessing a far-infrared camera would have shown three localized hot spots where the CPU, GPU, and memory controller are located.
So I thought about the underside of the heat sink bridge and took stock of how there are only small portions of the bridge which actually make direct contact with the bottom case. Since the bridge, as designed, can shunt only so much heat to the fans, there aren’t other ways for heat conduction and dissipation to take place.
Except, well, perhaps there might be!
Borrowing from user discussions on improving heat dispersion in current-gen Silicon MacBook Airs, I bought a 10x10cm sheet of 1mm-thick silicone thermal pad. I went ahead and cut out pieces of the thermal pad to fit the underside areas of the heat sink bridge (which otherwise don’t make contact with anything, aside from poorly circulated air) to help to move some of the heat generated to the aluminium case, rather than make the fans do nearly all of the work.
If all worked as planned, the overall temperature of the case would feel warmer (including around the palm rest), but this manner of heat dissipation would also take longer for the fans to rev up as high to move out heat and to “de-localize” the hot spots underneath the laptop. In other words, I’m using the aluminium case more directly to keep “hot spots” from getting so hot so often and to make resting it on my lap more tolerable.
Before:
After:
There were a couple of tiny spots I could have filled in, but for sake of testing, I left them alone. I didn’t use thermal pads in the zones next to the fans, as those are sealed by a thin foam rectangle factory-attached to the bottom case (and sealing away air within from the rest of the case). I might come back to those later.
(While I had the heat sink bridge out, I also opened the fan assemblies, cleaned out a year’s worth of dust, and added to the spindles a drop of the lubricant I’ve been using for refurbishing fans.)
I ran a screen cap of the temps before I put the system into hibernate, and about an hour after I woke the system from deep sleep. The ambient room conditions were the same:
Before: After:
The numbers to pay especial notice to are these:
I wrote this post with the laptop on my lap, and for a change, it isn’t slow-cooking me. The fans aren’t moving as quick as before, while the top case and overall bottom case are perceptibly warmer. The above data are the best I can do without access to an infrared camera.
tl;dr: Not so hot!
EDIT to UPDATE, late into day 2: Even as this is still pretty early, post-modification, there is still a noteworthy drop in overall temperatures around the areas of heat generation — namely, the CPU/GPU areas and their respective heatsinks. In the 7-day line chart on iStat Menus, the drop in overall temperatures at the typically hottest area — the GPU heatsink — is evident (the break in the line is when the system was in hibernate mode and while I was adding in the thermal pad modification).
Overall, the CPU die is about 12°C cooler; the CPU proximity is about 6.5°C cooler, and the GPU Heatsink(0) is about 7°C cooler; other generators of heat are between 6 and 7°C cooler. Meanwhile, the three heatpipe sensors are about 2–5°C cooler, and the two case-related thermal sensors — Palm Rest and Skin Proximity — are 6°C and 4°C cooler, respectively (i.e., not as much of a drop as at the areas where that heat is generated). What this suggests is the aluminium case is now taking on (and dissipating passively) more of the overall heat from the thermal generation, borne out by an overall warmer case, and lacking the localized scalding-hot spots.
Put another way: the difference between the GPU Heatsink(0) and Skin Proximity before the thermal pad modification was about 28°C; since the modification, the difference is about 25°C — or about 3°C (which is bang-on with the difference between the 7°C cooler GPU Heatsink(0) and the 4°C cooler Skin Proximity reading). In Fahrenheit, that’s about 5.5°F.
As I add this section via the A1261, I’ve managed to push the Skin Proximity temp up to 42°C under load (Interweb, Transmission, Console, Preview, and a few OS X lightweight utilities like Dictionary and Calculator). Before the thermal pad mod, this would have been closer to 46°C — the latter being above the threshold for risking second-degree burns (after over an hour of constant exposure) or third-degree burns (after over 90 minutes of constant exposure); in localized spots, such as where the cross-members of the heatpipe assembly met the bottom case), that temp was likely much higher). And this might finally explain how I received such deep burns over the summertime as I used this laptop on my bare leg lap.
Unloved orphan of tl;dr: This thermal pad mod is doing exactly what I hoped it would (better cooling, less hot to the touch, and hopefully extending life on the GPU and CPU with less heat lingering) — all for less than $10!
Last edited:
Nice job! I wish I had the courage (and knowledge) to do surgery like this on my machines.
Is this on a unibody MBP? If so, that's disappointing, as the non-unibody MBPs have IMHO an utterly terrible system for managing heat -- from what I could tell from the time I opened up a 2006 MBP 1,1, Apple designed it such that the entire bottom case acted as one giant heatsink for the GPU/CPU/Chipset heatpipes. I would have hoped Apple designed something better for the succeeding generation.
I'm glad that the mod was effective! (Or at least it seemed that way from the numbers I was reading in your screenshots.) Can you give a link to where you bought the thermal pads/what brand you got?
In the aluminium MBPs the entire bottom case isn’t functioning as a giant heat sink, given the way the heat sink assembly actually makes very little direct contact with the bottom of the case.
In the above pics, those little cross-members which are about a millimetre higher than the pair of pipes beneath are the only portions of the heat sink which make physical contact with the bottom of the case. This is, I hypothesized, why cooling is so poor: they designed it so that the fans — and those minor cross-member contact points — are the only active and passive means, respectively, of moving heat away from the processors. Also, these contact points might lend to why the hot spots get so hot: all the heat which can be shunted passively through those tiny cross-members are creating especially hot areas on the bottom case where they make direct contact.
By forcing more direct contact between the heat sink assembly and the bottom of the case, more of the heat could, as part of that hypothesis, be passively transferred to the entirety of the aluminium case foremost, leaving the rest of the cooling work to be handled actively by the fans (which would spike to maximum speed less quickly and, hopefully, less frequently).
Since reassembling everything, I’m finding the bottom (and top) of the case to be, overall, warmer, but those localized hot areas are no longer so hot to the touch. When the CPU/GPU get put under high load, the active cooling is still kicking up the fans as expected, but they don’t seem to be running as long or as hard to bring down temperatures. Even now, on my lap, when I’d expect full fan speed from the GPU side, I’m only seeing about 5700rpm (instead of the usual 6000rpm or higher). It’s subtle, but there’s a moderate difference.
Again, the above observations are provisional — a few hours old, really — but if time bears out with consistency, then getting the heat sink assembly to make substantially more direct contact with the aluminium case is actually helping to move away heat more efficiently by making the aluminium case do more of the work.
So far, yes, it appears to be effective. It might be more so if I add some additional thermal pad in the fan zones, but for now, I’m going to leave it with thermal pads as shown above.
You can probably use any silicone-based thermal pad. I went with a generic thermal pad from Aliexpress which purported to have a higher thermal efficiency than others on the same site. Specifically, this was what I bought, but any of the “silicone plaster” type thermal pads available ought to work. I went with a 1.0mm thickness, though I think a 1.5mm thickness might have worked even more effectively here.
Because of the “plaster” texture (protected by differently-coloured backing on either side), it’s easy to cut with an X-acto blade, but it’s also pretty soft and delicate (i.e., pliable) once both sides of the plastic backing are removed.
Last edited:
Ah yes, I skimmed through B S Magnet's original post and missed the model number reference.
Ah, that makes a lot of sense. I'll have to look more closely when I ever get around to taking apart one of my A1211/A1150 MacBook Pros. I thought that the heatsink pipes rested directly on the bottom of the case, but I must have missed those cross members. Also, thanks for the link. I've been curious in trying out thermal pads on MacBooks ever since I saw a video from LTT comparing them vs. conventional thermal paste. I might even try out your mod on my A1226.
AirServer also works. I have that on my Intel/PowerPC Macs and none of these Macs are newer than 2009.
I can send audio from any Mac to any Mac and video from my iPhones to any Mac as well. Works pretty well.
I’m going to continue to use a paste-based thermal grease (like Noctua’s) for any time when direct contact between CPU/GPU/memory controller and heat sink plate meet. I bought the thermal pad to try on places where air is currently the “heat-conductive” medium and to use the pad as a bridge between heat source and passive cooling surfaces which Apple didn’t pursue.
Next up in this thermal pad-as-an-ancillary approach: redressing overheating on the backside of the Power Mac G5, particularly the memory controller. After that, provided I have any left, I’ll do the same with my two PowerBook G4s.
Can it also use an AppleTV as an extended or mirrored screen, which is what AirPlay Mirroring is?
Got the Early 2006 Mini streaming Drone Zone on iTunes 11.1.5 to the BT loudspeaker. Had to slip in one of my Belkin BT adapters for it to work right though. Looks like I'll be picking up another Sena Parani UD100 adapter for this Mac though as it's only supporting one connected device at a time. That's the same one the Early 2009 Mini has (because I damaged the internal BT antenna on that Mini). 1000 meter range, supports multiple connections.
Oh, what the heck, why not?!
Second SL only 2006 Mini up and running headless.
So that makes four Minis in the garage and one inside.
Second SL only 2006 Mini up and running headless.
So that makes four Minis in the garage and one inside.
So my 2006 macbook has been collecting dust because the charger wore out. It would charge the battery when powered off, but didnt have enough juice to power the thing on. A few months back i ordered a new charger. Unfortunately i didn't pay attention and bought a magsafe2. So last night i took the magsafe1 cable from the original charger and wired it in to the new charger. Problem solved. Today i spent the day doing a fresh 10.6 install and setting it up to my liking.
No. Neofetch is just a bash script. Nothing special needed. I just downloaded the binary and plopped it in to /usr/local/bin. I've attached the latest version below.
Impressive!
Did you visually document this conversion work? If so, I'd ask you to share it for us to see as I'm definitely intrigued.
It was the easiest rewiring i've ever done honestly. 2 wires with + and - labeled on both boards, with 1 white (+) and 1 black/shielded (-) wire. Just desolder, and resolder.
Hardest part was cracking the cases open. ;-)
Hardest part was cracking the cases open. ;-)
As a
Below is a screen capture of the 7-day tracking of temperature differences of GPU Heatsink(0) (next to GPU Heatsink(1)) at a mid point in time between before the modification (before the gap) and after — i.e., 84 hours, or 3.5 days later. (The gap, as mentioned earlier, was the period when I had the MBP hibernating and disassembled.)
(The low point of 57°C at the end is a byproduct of being in a cool room while I wasn’t home; since finishing the modification, the idle point rests between 59 and 62°C.)
CONCLUSION
The goal for this modification was to improve thermal dissipation efficiency for the 2006–08 aluminium MacBook Pros — to make the underside less a risk for getting so hot as to slow-burn skin — and this goal is being met! If I decide to go back in there later to add in a slight bit more of that thermal pad to fill in the areas where the fans live, one can expect to see even more improvement of evened-out thermal dissipation overall, as the fans will only need to move up gradually, not suddenly, to move the once-sudden spike in heat.
FINAL NOTES
Dirty internals. To dispel one last outstanding question one might have about the change in overall temperatures from the hottest heat generator sensor on the MBP — at GPU Heatsink(0) — the temps on my MBP, prior to the modification, were from internals which were cleaned out completely a year ago (when I added a fresh, working 2.6 GHz Penryn logic board with a working, green-dot GPU), and whose thermal paste (for the CPU/GPU-to-heatsink-plate contact points) was the same, high-quality thermal paste I used this time (Noctua NT-H2). It was the same batch/tube.
While there was some dust which had entered the case following about 16 months of use, I don’t think there was enough to dramatically alter the overall cooling of the internals prior to this modification, as the fans worked and they were able to push out air.
Generously speaking, it’s possible a year and change’s worth of dust amounted to an overall uptick of one degree Celsius, at most.
Net difference. Overall, I’m seeing a net drop across all heat-generating sensors of about 3–4°C. Where the 3–4-degree difference in heat is now going is to the aluminium case itself, in a more widespread, linear, broader way than prior to the modification. This makes the feel of handling the MBP a bit warmer overall (by letting the whole of the case do more passive cooling work), but the upside is there aren’t those isolated hot spots underneath which previously led to burning one’s thigh or knee.
Apple, wyd? The only reason I can suss why Apple didn’t design the heatsink pipe assembly to make broader, direct contact with the aluminium case was to shave away any possible mass from the overall weight of the laptop (3.08kg), as any uptick in weight, over the PowerBook G4 DLSD predecessor (3.1kg), would have received unwelcome flak in reviews — despite the MBP being faster, more feature-laden (built-in iSight), and wider (due to inclusion of said iSight).
Other notes. I use Macs Fan Control for managing the fans. For the “left side fan”, I have MFC monitor the “GPU diode” (which appears to be the same thing as what iStat Menus reports as “GPU Heatsink(0)”), and to begin increasing fan speed at 50°C and max out at 68°C. For the “right side fan”, MFC monitors the “CPU diode” (which is what I think iStat Menus reports as the “CPU Heatsink”); to start increasing fan speed at 55°C and maxing out at 65°C. These settings are unchanged from before the mod.
Thanks for reading!
Last edited: