UPDATE:
I wish I'd done some temperature benchmarking beforehand, but instead sourced the community to help out. Check out the various results on XDA... there are also results from two TF700 owners (in Europe) who were gracious enough to help me out.
http://forum.xda-developers.com/showthread.php?p=28051518
First, it seems clear that my mod really didn't mitigate the heat issue enough... so don't bother.
Second, comparing the numbers between other A700's and the TF700... well, I'll let you draw your own conclusion on that one. :-(
---
Hi all,
As many of you have experienced, the A700 can get very hot under certain conditions. Mine got extremely hot to the touch after just an hour of gaming (Heavy Gunner, NFL Flick QB, and Cut the Rope), and crashed 2 or 3 times.
Well I decided to do something about it. Reading through the Service Guide (thank you paugustin!!!) & A700 teardown photos from another site, I came to some conclusions. First, from the looks of things, the back panel has some kind of metal plate lining, and 3 "pads" that closed the gap between the metal plate and several surfaces of the mainboard. My hope was that these pads were not simply foam pads but were actual thermal pads. I guessed that a bit of thermal paste might help the efficiency of those thermal pads... and if I got really ambitious, I could replace the thermal pads with larger ones (and paste those too).
Service Guide:
http://forum.xda-developers.com/showthread.php?t=1716922
I followed the instructions in the Service Guide & opened up my A700 earlier tonight. I'm pleased to report that as long as you take your time, it's super easy to open up.
And lo and behold, my conclusions were accurate... the back plate is copper and the pads were thermal pads!
Back Plate:
https://www.dropbox.com/s/um556g6vtl7y99e/C360_2012-06-26-20-14-30.jpg
MainBoard:
https://www.dropbox.com/s/jka8gooiyqn9r2d/C360_2012-06-26-20-14-46.jpg
Closer Look @ the Thermal Pads:
https://www.dropbox.com/s/0j4e7wjfodwryk9/C360_2012-06-26-20-15-06.jpg
I took some CPU thermal paste I had lying around & dabbed some on:
https://www.dropbox.com/s/kvv99k3ghx73a00/C360_2012-06-26-20-17-24.jpg
Afterwards, I slapped everything back together, fired up NFL Flick QB and played for a half hour. I'm pleased to report that the A700 got warm but nowhere near as hot as it did in my prior gaming session. The "level of warmth" was what I'd consider mildly warm, comparable to my fiancee's iPad3, so definitely within reasonable tolerances.
I still intend to put the A700 through some more intense gaming tomorrow, but I wanted to write this post up and share it with everyone first. Hope folks find this useful and insightful!
Nice discovery, but yikes! Thermal grease is meant to go on in a translucently thin layer or it ends up having the opposite effect! If you put that much between a CPU and it's cooler, you'd kill the CPU! :S
superawesome!
Yet still i'd like to wait for more people to evaluate the hardwaremodding before i go in there myself.
Still, kudos to you. :good:
FloatingFatMan said:
Nice discovery, but yikes! Thermal grease is meant to go on in a translucently thin layer or it ends up having the opposite effect! If you put that much between a CPU and it's cooler, you'd kill the CPU! :S
Click to expand...
Click to collapse
Except in this case, the thermal grease isn't going up directly against the CPU... it's going up against a thermal pad & large metal plate that is covering the entire mainboard.
Additionally, unlike a heatsink & CPU which are fitted together tightly, the back panel of the A700 & thermal pads do not contact anywhere near as tight to the mainboard plate. I fitted the two halves together, then pulled them apart again, to see how much the paste really spread out. This way I could better gauge how tight the contact was & remove any excess that got squished out. However, not much paste really did, telling me that the compression between the two isn't that great. That is why I'm toying with the idea of replacing the thermal pads entirely.
Hey, good stuff. Keep us in the loop on this....
Thicker thermal pads might be in order, or a better paste job under the plate on the CPU/GPU...
Bigger pads to spread the heat out more since the back looks to be one copper heat sink..
http://www.frozencpu.com/cat/l2/g8/c487/list/p1/Thermal-Thermal_Pads_Tape.html
Beknatok said:
Hi all,
As many of you have experienced, the A700 can get very hot under certain conditions. Mine got extremely hot to the touch after just an hour of gaming (Heavy Gunner, NFL Flick QB, and Cut the Rope), and crashed 2 or 3 times.
Well I decided to do something about it. Reading through the Service Guide (thank you paugustin!!!) & A700 teardown photos from another site, I came to some conclusions. First, from the looks of things, the back panel has some kind of metal plate lining, and 3 "pads" that closed the gap between the metal plate and several surfaces of the mainboard. My hope was that these pads were not simply foam pads but were actual thermal pads. I guessed that a bit of thermal paste might help the efficiency of those thermal pads... and if I got really ambitious, I could replace the thermal pads with larger ones (and paste those too).
Service Guide:
http://forum.xda-developers.com/showthread.php?t=1716922
I followed the instructions in the Service Guide & opened up my A700 earlier tonight. I'm pleased to report that as long as you take your time, it's super easy to open up.
And lo and behold, my conclusions were accurate... the back plate is copper and the pads were thermal pads!
Back Plate:
https://www.dropbox.com/s/um556g6vtl7y99e/C360_2012-06-26-20-14-30.jpg
MainBoard:
https://www.dropbox.com/s/jka8gooiyqn9r2d/C360_2012-06-26-20-14-46.jpg
Closer Look @ the Thermal Pads:
https://www.dropbox.com/s/0j4e7wjfodwryk9/C360_2012-06-26-20-15-06.jpg
I took some CPU thermal paste I had lying around & dabbed some on:
https://www.dropbox.com/s/kvv99k3ghx73a00/C360_2012-06-26-20-17-24.jpg
Afterwards, I slapped everything back together, fired up NFL Flick QB and played for a half hour. I'm pleased to report that the A700 got warm but nowhere near as hot as it did in my prior gaming session. The "level of warmth" was what I'd consider mildly warm, comparable to my fiancee's iPad3, so definitely within reasonable tolerances.
I still intend to put the A700 through some more intense gaming tomorrow, but I wanted to write this post up and share it with everyone first. Hope folks find this useful and insightful!
Click to expand...
Click to collapse
Fantastic for the XDA community, but the average buyer will be like the cheesey pop group Power Station and will "Feel the heat".
rushless said:
Fantastic for the XDA community, but the average buyer will be like the cheesey pop group Power Station and will "Feel the heat".
Click to expand...
Click to collapse
Then be happy that you're amongst the enlightened!
CAUTION!!!!!
OK maybe that's a bit strong, but as an engineer I have to question the logic employed.
First of all, the heat generated by the CPU, battery, etc. must exit the device. ALL OF IT!
Temperature of the back plate is determined by two factors alone: thermal resistance (e.g. insulation, heat pipes, air flow, fins, etc.) and the amount of heat transfer. I'm serious!
It appears that you have done nothing to affect the thermal resistance between the back plate and the air (e.g. adding cooling fins or increasing air flow). Therefore if your back plate temperature has decreased (and the air temp stayed the same), then we would conclude that heat transfer between the back plate and the air has DECREASED.
If hea
If heat transfer from the plate to the air has decreased, then that means the heat transfer from the CPU etc. to the plate must have decreased as well!
My guess is that your thermal paste is producing a higher thermal resistance, as predicted by FFM! And it's deceptive because it's insulating the back plate from the CPU etc.
So where's the heat going? (I hear myself asking... myself.)
My guess is that it's dissipating through the tablet, getting distributed and absorbed and ultimately emitted more uniformly from the device. This would raise the internal temperature of the device, including the CPU! Hence the CAUTION!!!!!
Your idea could have some benefit, though, and here's how i see it playing out:
If you decrease the thermal resistance between the CPU and the back plate, then the CPU would be closer to the temperature of the back plate (i.e. lower). But the back plate temperature won't really change because you still have the same total heat coming through and the same thermal resistance with the air.
As for spreading the temperature across the back plate more, I'm not sure you can do it more effectively than the copper plate that's already there. Not without insulating the most direct heat transfer path!
yeoldeusrename said:
CAUTION!!!!!
OK maybe that's a bit strong, but as an engineer I have to question the logic employed.
First of all, the heat generated by the CPU, battery, etc. must exit the device. ALL OF IT!
Temperature of the back plate is determined by two factors alone: thermal resistance (e.g. insulation, heat pipes, air flow, fins, etc.) and the amount of heat transfer. I'm serious!
It appears that you have done nothing to affect the thermal resistance between the back plate and the air (e.g. adding cooling fins or increasing air flow). Therefore if your back plate temperature has decreased (and the air temp stayed the same), then we would conclude that heat transfer between the back plate and the air has DECREASED.
Click to expand...
Click to collapse
Right, I'm aware of this.
The exact problem I was attempting to mitigate, is the transfer of heat between the mainboard cover plate and the back panel copper plate. I was theorizing that the thermal pads being used were not fitted well, thus not functioning optimally as designed to begin with. As a result, not enough heat was being transfered to the back panel copper plate for proper dissipation.
So either heat is being transferred to the back copper plate more efficiently now and dissipating more effectively... or the opposite as you proposed.
However, if heat transfer has decreased, wouldn't heat buildup would still occur within the small area between the back panel & mainboard? While it wouldn't be absorbing the same amount of heat directly from the thermal pad, heat would still build up over time and that should still be noticable no matter what, right?
I'd also note that the BETTER solution would be to NOT combine thermal paste with the thermal pads, since as you point out, that can adversely affect thermal resistance.... but rather to replace the pads with better pads.
But I decided to be a guinea pig and try it out anyway.
Sorry I got distracted before posting my conclusion:
So my conclusion (prior to your latest comments) is that you can effectively lower the temperature of the CPU but not of the back plate. (unless you're insulating it from the CPU, which will spread the heat around and raise the temp of the internals!!!!).
I will consider your latest comments and repost!
The way I see it, the heat was not effectively transferred to the radiator. We have to remember that the case heat-up experienced by touch is not only the radiator, but the case itself heated by both the radiator and the air.
I'd assume that lack of crashes indicates smaller temperature of the CPU\GPU, as the heat is conducted more effectively by the paste to the radiator and dissipates evenly, instead of building up in one place and transferred to the case by air.
It's times like this, that highlights the one minor missing flaw of the A700... there's no bloody temperature sensors that we can poll!!!
*bashes head against desk*
So unfortunately, all findings are really subjective unless yeoldeusrename's conclusions are correct and mine are totally incorrect, and my A700 decides to fry itself.
Beknatok said:
It's times like this, that highlights the one minor missing flaw of the A700... there's no bloody temperature sensors that we can poll!!!
Click to expand...
Click to collapse
Well, i wouldn't be that sure. The A510 has them (at least separately for cpu and battery) and can be read by certain apps from the store. (E.g system tuner and battery monitor widget) so i would guess the A700 has them too.
Perhaps that helps
Sent from my A510 using xda app-developers app
mearoth said:
Well, i wouldn't be that sure. The A510 has them (at least separately for cpu and battery) and can be read by certain apps from the store. (E.g system tuner and battery monitor widget) so i would guess the A700 has them too.
Perhaps that helps
Sent from my A510 using xda app-developers app
Click to expand...
Click to collapse
The two apps I tried, didn't show anything for temperature, but I just grabbed system tuner & it does register cpu temp! Yay! Will explore the app further to figure out how to record history for accurate metrics. Thanks for the hint.
Sent from my A700 using Tapatalk 2
Beginning to wonder if the heat and touch issues are why the 700 is priced the exact same as the 510. Nothing else apparently makes sense.
Anyone who is tracking this thread & wouldn't mind helping gather some additional data, please check out this thread:
CPU Temp Benchmarking - Request for Help
http://forum.xda-developers.com/showthread.php?p=28009068
Thanks!!!
I think I'm going to look into creating a heat shield to place against the rear cover.
2 to 3 inches across and 4 down on the right side should do it.
This is a great tablet otherwise and don't want to give up on it just yet.
Besides the Asus sound system is inferior.
Hi, I read with interest the posts about wobbly (back and forth) docks on the tf101.
My TF700T developed this also after just 1 month.
Aditionally and more seriously perhaps, I noticed it stopped docking securely. I examined both the dock and the tablet and eventually noticed the retaining pin (inside the hole) on the right side of the tablet seems to have broken off due to frequently docking and undocking the two.
Has anyone else seen this?
ive sent it back to Asus who are claiming customer induced damage. This make me very angry bcoz i looked after it like a baby. all i did was use it a lot both docked and undocked.it never left my house. I reckon its a design flaw and eventually all many more of you will notice this.
wailim_2002 said:
Hi, I read with interest the posts about wobbly (back and forth) docks on the tf101.
My Tf701 developed this also after just 1 month.
Aditionally and more seriously perhaps, I noticed it stopped docking securely. I examined both the dock and the tablet and eventually noticed the retaining pin (inside the hole) on the right side of the tablet seems to have broken off due to frequently docking and undocking the two.
Has anyone else seen this?
ive sent it back to Asus who are claiming customer induced damage. This make me very angry bcoz i looked after it like a baby. all i did was use it a lot both docked and undocked.it never left my house. I reckon its a design flaw and eventually all many more of you will notice this.
Click to expand...
Click to collapse
Hey, It happened to me too and this other guy. Go to page 2 of this section, and go to one with Broken lock .but seems like its only three of us til now.
There is no solution as is...I've tried to get a replacement parts (like iphone) but i had no luck.
No, it's not only the three of you. "Funnily", the locking pin snapped off on my dock yesterday. I haven't been this angry in years. As with most of the users with a dock, it hardly ever left the dock -- I think I have uncoupled it for at most 6 or 7 times, and as such, it's not even a question of docking it with force or gently...
The locking pin would seem to be unable to withstand keeping the tablet in place for such an extended period. I noticed that both locking pins have degraded considerably -- again, I hardly ever uncoupled the tablet from the dock, so it's 'slide damage'. This would qualify as a hardware design error in my opinion. What moron uses plastic clips on locking devices? That would need some nice metal engineering. This is not even close to the durability that I expect from a 700 euro device!
Once my blood pressure has gone down for a bit, I'll try and drop ASUS Netherlands a line and get my issue known. Then I'll see what their point of view is, and if it's "your own damn fault", I'll have to determine what to do: either proceed to professional help (a lawyer, to be more specific) or let it slide. I do not know what my route would be in this case -- I don't want to risk a stomach ulcer, but my lawyer does this for free (many Europeans have a lawyer plan so there is no financial impediment to asking support).
EDIT: I had secured the tablet to the dock with a few strips of heavy-duty doublesided tape to keep it from accidentally coming out of the dock. I cannot have the tablet's screen cracking from fall damage and ASUS having a stick to beat me with.
MartyHulskemper said:
No, it's not only the three of you. "Funnily", the locking pin snapped off on my dock yesterday. I haven't been this angry in years. As with most of the users with a dock, it hardly ever left the dock -- I think I have uncoupled it for at most 6 or 7 times, and as such, it's not even a question of docking it with force or gently...
The locking pin would seem to be unable to withstand keeping the tablet in place for such an extended period. I noticed that both locking pins have degraded considerably -- again, I hardly ever uncoupled the tablet from the dock, so it's 'slide damage'. This would qualify as a hardware design error in my opinion. What moron uses plastic clips on locking devices? That would need some nice metal engineering. This is not even close to the durability that I expect from a 700 euro device!
Once my blood pressure has gone down for a bit, I'll try and drop ASUS Netherlands a line and get my issue known. Then I'll see what their point of view is, and if it's "your own damn fault", I'll have to determine what to do: either proceed to professional help (a lawyer, to be more specific) or let it slide. I do not know what my route would be in this case -- I don't want to risk a stomach ulcer, but my lawyer does this for free (many Europeans have a lawyer plan so there is no financial impediment to asking support).
EDIT: I had secured the tablet to the dock with a few strips of heavy-duty doublesided tape to keep it from accidentally coming out of the dock. I cannot have the tablet's screen cracking from fall damage and ASUS having a stick to beat me with.
Click to expand...
Click to collapse
Hi, This is great news.... well erm... sorry for your trouble and all but it is good that the numbers who are experiencing/noticing this failure are going up as it lends weight to the design flaw theory. I agree 100% with you that the design does appear to be plastic clips inside the tablet and what I said in my RMA description is that if ASUS did not have an updated reinforced part, that I felt the unit was unfit for purpose and I did not want it back. I was prepared to give them 1 chance to repair it as this is the consumer law in Ireland but I was intending to test it thoroughly upon return. But now that they are trying to say I dropped it or something.... this is another story!
For me, I am still under 3 months and within the time scale to do a credit card charge back. I'd rather have them come after me if they feel they can stand over their design.
wailim_2002 said:
Hi, This is great news....
Click to expand...
Click to collapse
Heh. Thanks for your considerate reply. I was one of the first TF700 owners in the world, so I'm far past any warranty period in which they have to prove it my fault instead of me proving it theirs. I'm still too pissed to take up the phone right now, but I will try and send an e-mail to start with somewhere in the next few days.
When I get the parts delivered I'll be doing a touchscreen and battery replacement on my TF300T. Looks a lot easier than the phones I have repaired -- more room to work -- the several videos spell it out well.
But I ask: Is it a wise idea to replace the thermal paste on the copper plate? And if so any brand recommendations or are there any incompatible thermal pastes out there to be avoided because of their composition? I notice the back of the tablet gets quite warm at times.
As for the touchscreen: I can see the screen is badly marked after close to 7 years of use. The letter "i" on the virtual keyboard is dead, and I can no longer swipe properly in twrp recovery. Swiping the wipe page will cause other partitions to be selected. Fortunately the swipe never succeeds in launching the flush or format.
Could this be a symptom of another more serious hardware issue -- the fact that the touchcreen is totally useless in twrp but only a little broken in user mode?
Thanks
Cheers
diordnAMRZAR said:
When I get the parts delivered I'll be doing a touchscreen and battery replacement on my TF300T. Looks a lot easier than the phones I have repaired -- more room to work -- the several videos spell it out well.
But I ask: Is it a wise idea to replace the thermal paste on the copper plate? And if so any brand recommendations or are there any incompatible thermal pastes out there to be avoided because of their composition? I notice the back of the tablet gets quite warm at times.
Click to expand...
Click to collapse
Here we go!
Battery replacement: https://www.youtube.com/watch?v=6RF8Yp_xxcc
Touchscreen replacement: https://www.youtube.com/watch?v=DauysmMk-7c
Wish you good luck! Please take care of correct fitting touchscreen for your tablets serial number.
ebonit said:
Here we go!
Battery replacement: https://www.youtube.com/watch?v=6RF8Yp_xxcc
Touchscreen replacement: https://www.youtube.com/watch?v=DauysmMk-7c
Wish you good luck! Please take care of correct fitting touchscreen for your tablets serial number.
Click to expand...
Click to collapse
Yeah thanks. I have seen these youtube offerings , and others. Do you have any advice about redoing the thermal paste for the processor? The tablet can get pretty warm at times.....I do not want to bork the tablet with the wrong type of thermal paste.
diordnAMRZAR said:
Yeah thanks. I have seen these youtube offerings , and others. Do you have any advice about redoing the thermal paste for the processor? The tablet can get pretty warm at times.....I do not want to bork the tablet with the wrong type of thermal paste.
Click to expand...
Click to collapse
Please note, you don't need any clue because the thin copper heat sink, shown in step 1 of the video, is self-adhesive. Take off the heatsink carefully by a light spin. Some glue fixes it on the CPU. On the left and on the right site are black or grey tapes. Remove the tapes carefully. Keep all together safe during the repair.
As you can see on the video the last step will be to put the thin copper heat sink back again with a light press down. It will stick again like before.
Did you meanwhile get the touchscreen of your device replaced successfully?
The factory heatsing had thermal paste.
I got a new one from seller dince i had overheat problems at summer and i used thermal paste as well.
But the px6 is not flat. It has holes and such.
So i am wondering.
Maybe a thermal pad would be better than thermal paste? If it is what thickness?
Paste has always been better with uneven surfaces.
I can understand it exactly the opposite way. Even surfaces go for paste, non even with pad.
GnFgr said:
I can understand it exactly the opposite way. Even surfaces go for paste, non even with pad.
Click to expand...
Click to collapse
You may be thinking of EPOXY thermal compound, which in my opinion is garbage.
On the other hand any silicone based grease is far better at filling gaps than pads are. Most factories tend to use pads though because it's cleaner and there is less maintenance to perform on the factory assembly machines.
Good read here:
Thermal Pads vs Thermal Paste: The Best Choice for Mounting Your Heatsinks
I used to work as a construction assistant, where my primary responsibility was to level properties prior to the heavy concrete trucks laying down a level foundation. It was a tedious and taxing job, but something that is quite crucial for a building in order for it to withstand years of use and...
resources.altium.com
I never said about epoxy based. Never.
The px6 is not only uneven but has a metal shield on top. The metal shield acts as an interference shield ( like a faraday cage maybe?) And has holes. So the surface in not only not even it is full of holes.
I am not asking about even and not, and you stand correct, in surfaces that are either even or almost paste in my mind is the way to go.
But when the gap is bigger that x what is better?
So today i cleaned the paste with isopropyl alcohol and used a good pad from gelid i think. It is rated as 15W/mK and is 0.5mm thickness.
With external temp of 5 degrees Celsius the cpu was at 78 to 80 degrees after an hour. That is with thermal paste.
With the thermal pad the temperature after an hour is 65 degrees.
The usage is exactly the same at the one hour usage.
Maybe the metallic shield has no contact with the cpu and the pad manages through the holes to have at least some contact and achieve better heat transfer?
Well... some kind of cooling is better than none, that is for sure, but heat sinks are supposed to make full and direct contact with the device and if it can't then I suggest you have mechanical issues to correct rather than thermodynamic ones.
Damn.
I am not asking what is better as a general concept. I am asking what is better for a specific use that there is a shield above the cpu and there is no direct contact.
Anyway mate forget it.
GnFgr said:
Damn.
I am not asking what is better as a general concept. I am asking what is better for a specific use that there is a shield above the cpu and there is no direct contact.
Anyway mate forget it.
Click to expand...
Click to collapse
I doubt you can find many to answer a question on mounting a heat sink in such an improper fashion. Not withstanding, you seem to be convinced that the pad is the right way to go, so I'm not sure why you are asking the question in the first place.
I asked because i wanted to learn what is better. Haven't test both when i asked and thought asking in a forum might find someone that had data on it.
Sorry if i overwhelmed you.
Anyway just forget this thread to the void.
Keep on
PLEASE DO NOT TRY THIS UNLESS YOU'RE PERFECTLY FINE with possible fire / you know what you're doing .
Yes ik this sounds random asf but yes, I put some shampoo in my phone, let me explain my story:
I was just existing really looking at random videos of yt until I found a random series of using random objects as thermal paste, & then one supeising result came when he tested shampoo multiple times & concluded that well yes, "this is just as good as thermal paste" & so I wanted to see, would this apply to phones? Can u somehow put a alt to thermal paste onto a thermal pad & for no reason on the back & get away with it without anything happening?
I'm short terms, it depends on who you are.
Anyway....
I the great papyrus I then decided to do a teardown of my hlte note 3 & answer that question! After taking it apart (& halfway through my nano / grandma came in to watch), I had some honey shampoo on some tissue, applied it above the 2 yellow sheets in assuming are thermal pads for the CPU & whatever the tiny one was for XD
For fun, I also decided to apply it onto the removable backcase & on the screen to clean later with a wet tissue & put it back together & I turned it on
Aww Nice, I haven't accidently turned my phone into a s7 bomb!
Now that we know that if you're careful you can apply shampoo to your phone, does it benefit the device, or does it attract ants & smell nice because I used honey shampoo?
Yes and no, I'm not sure yet as it seems to be OK whether it's unlimited or limited to a temp, the CPU will be a bit more obedient somehow but hey I'll take that as a huge plus & also, the temp it sits at the most is 49'C, before, it used to sit at 54'C so again, a plus! But it doesn't mean it doesn't go extreme.
From launching it on twrp or just booting it, it seems more comfortable going to 60'C & a lot of heat likes to transfer to the front, aka, the screen.... Which may cause burns but will mean the heat can go off easier so its a case of which component you prefer to not take risks on. (well ngl you'll risk both doing this)
So in conclusion, if you didn't have all the control over your device, then it doesn't make a difference.
But if you have that control, then go for it as it's way more OK with limiting its temp without performance drops somehow I still need an answer to how lol.
Hope you enjoyed this story & ye.