Thermal compounds / pads and thermal conductivity.

[stj]

Re: Thermal compounds / pads and thermal conductivity.

you cant use toothpaste, Flouride is incredibly corrosive.
infact it's one of the few things that can eat through concrete!

i have read about someone using peanut butter once though.
i wouldnt try it - it's probably not a smooth as it says!

[diif]

Re: Thermal compounds / pads and thermal conductivity.

Why not buy 20g tubes of MX-2 or MX-4. Both exceed the spec of TC-5622 and you don't need to buy such a large tub.
Buy direct from their site so avoiding ebay fakes.
Whilst i'm having to save up money for certain items for fixing, thermal past isn't one of them.
MX-2 30g $19.99 65g $29.99
MX-4 20g $18.99
$7 shipping

https://www.arctic.ac/us_en/

I heard milk can dissolve it too in certain circumstances stj.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

Why not buy 20g tubes of MX-2 or MX-4. Both exceed the spec of TC-5622 and you don't need to buy such a large tub.
Buy direct from their site so avoiding ebay fakes.
Whilst i'm having to save up money for certain items for fixing, thermal past isn't one of them.
MX-2 30g $19.99 65g $29.99
MX-4 20g $18.99
$7 shipping

https://www.arctic.ac/us_en/

I heard milk can dissolve it too in certain circumstances stj.

Okay. I originally ruled MX-# out because of the stuff that customer brought me that was supposedly MX-2. I worry about the specs though. It seems most of the companies that were selling to consumers, not businesses, fibbed a good bit about their numbers.

But I think momaka brought it up originally (maybe someone else), after a certain point, the higher the thermal conductivity, it doesn't really matter, with CPUs at least. So if MX-2 or MX-4 is comparable to AS-5, even if it's not really as high as TC-5622, it'll work just fine. And if it's cheaper than TC-5622, that's perfect! I'm looking for something a bit cheaper than AS-5 that's equal or better than AS-5. To me, AS-5 is wonderful stuff, it's just a bit pricey, and now that we cannot buy it locally, if we have to order thermal paste online (which is now the case), I figure we should look at what's out there.

I'll probably order both the MX-2 and the MX-4 at first, and see what one I like best. I can't stand this IC Diamond stuff. It gets real messy and it's hard to get off.

I didn't realize Arctic made the MX-2 and MX-4. The three sets of thermal pads I bought are made by Arctic and they seem to be a lot better than what was originally in the cell phone that I needed them for.

[momaka]

Re: Thermal compounds / pads and thermal conductivity.

I think application might have a lot to do with this. I think depending on the application, maybe it might matter. Probably never with CPUs in a computer though, but perhaps other applications? I see stuff for sale that cost like 500$ / 200 grams and you gotta buy a minimum of 200 grams. But it's got a high thermal conductivity. High to the point where I don't think you'd ever use it on a computer's CPU. Like it'd just be essentially throwing money away. But Dow Corning makes it, so I think there's applications where that high thermal conductivity could be needed.

Now we are onto something here.

There are indeed applications where the thermal conductivity matters. But for a computer CPU or GPU, just about anything with decent conductivity will do. And you won't see that much of a difference in most cases. That's because the TDP per area is not that big.

The only major difference between the good quality compounds and the cheap ones is how fast they dry our or run off the CPU/GPU. Cheaper compounds tend to dry out quicker most of the time. Arctic Silver 5, on the other hand, is extremely stable and does not really dry out (I have seen AS5 applications that were original from over 10 years ago, and it was still okay). Which is why AS5 is so well received by everyone. But even cheaper pastes like MX-2 and MX-4 are decent in that regard.

As far as temperatures are concerned, the heatsink design and case airflow matter much more, as you noted yourself. And this is an area I'd like to discuss here.

First off, I have seen way too many posts in the past where people say they switched from thermal compound X to thermal compound Y and got Z many degrees lower temperature. But that's hardly saying anything if the tests weren't performed in some controlled manner. By controlled, I mean you have to take into account not only the load you are putting on the CPU/GPU in question, but also the room temperature and how long the computer has been running. All of these *will* impact the temperature readings.

When I do my tests, I always check the thermometers in my room to see what the ambient temperatures are (I have two: a mechanical one with a rolled spring, and an alcohol tube one). Moreover, I prefer to do all thermal experiments only in the late spring or mid-fall, as those are the times when we often have both the AC and the heating OFF in the house, thus leading to much more stable room temperature (typically 28C/83F). I've found that the house AC/heating can have an impact of around 1C in temperatures, even though my room thermometers will not show a change.

I also prefer to do my tests with an open case - that way, the room temperature is the ambient case temperature and it only takes about 10-15 minutes max before the PC comes to a stable idle and load temperatures (vs. a closed case, which can take a while to heat up and also cool down, depending on how open or closed it is).

And finally, there is repeatability - how many times were the tests performed before concluding that the temperatures changed. Surely, doing any experiments just one time may not give the most accurate results. Also, was the PC loaded the same way? That will impact the temperatures too.

I like this and I'm going to try it with some of the cheap / counterfeit thermal pastes I got laying around. I think the only fake stuff I got is the MX-2. That's practically water. It'd be nice if I could use it for something though instead of just throwing it away, even though I didn't pay for it or anything.

Just watch your temperatures and don't experiment on a chip that does not have thermal throttling protection (i.e. Athlon XP or Pentium II, or just about any GPU).

A guy brings me a PC that has an aftermarket CPU cooler that cost around 15$. Someone just took a cheap fan and screwed it into a heatsink. The stock is an AMD Wraith cooler. He upgraded the APU to the fastest APU the system could handle. The stock cooler for that new APU is still the AMD Wraith cooler, but his aftermarket cooler simply cannot cool the new APU.

I suppose I should have been more specific about the aftermarket coolers.

Obviously getting a cheap one will not yield good temperatures.

There are some aftermarket coolers that are phenomenal, but I don't think 20$ is going to get you anything like that. I think 20$ might you get something that can cool around what a stock cooler can cool. And think in order to get something better than the stock, you're going to be looking at spending more than 20$, in my opinion.

For CPU heatsinks, $15-20 can get you a decent cooler that cools slightly better than the stock while being more quiet (clearly the one that the guy above brought you must have not been like that).
The Cooler Master Hyper T2 comes to mind here. Typically, it is around $20, but may be found for $15 with rebate. Will keep a 65W TDP CPU cool while being much more quiet than a stock. Can even be used up to 95W TDP CPUs.
Then there's its bigger brother, the Hyper TX3. Has 3 copper heatpipes and performs even better than the T2. Can be found for $20-25.
And finally, there are the CM Hyper T4 and Hyper 212 Evo. These will cool down just about anything without a problem. Up to 130 Watts TDP no problem. Prices on these vary from $25 to $35.

That said, I'm not trying to promote or "push" Cooler Master heatsinks here. It's just that we always had them stocked at the place where I was just working. Thus, we used them more than anything else when it came to custom builds. If you find a cooler from a different brand with a similar design to the above mentioned coolers, you can expect it to perform more or less the same.

Maybe a 20$ cooler would really cool a CPU, better than stock, but maybe it's got a really cheap fan and that fan fails much sooner than the stock. That's what I'm getting at. I think a lot of the cheaper aftermarket fans really aren't any better than stock.

That, you are right about. Stock coolers typically use good fan brands, like Delta, Nidec, and similar... or at least used to. Seems like the quality of the fans on the stock coolers has fallen down a bit too. I saw quite a few Intel stock HSes with seized fans in the last 6 months (some even melted! )

As for aftermarket coolers, the fan quality can obviously vary quite a bit. That said, the above Cooler Master heatsinks seemed to have decent quality fans. I saw quite in customers' computers that were caked in dust, but the fans were still spinning fine and did not sound bad/dry. So whatever CM is using for their fans aren't that bad. On the other hand, the cheap OEM-like replacement aftermarket coolers often have very cheap fans that tend to seize fairly often.

When I buy an aftermarket cooler for a custom build (which I haven't done in a while, I will admit), I look at the fan type, the heatsink material, the fan size, the fan airflow, the heatsink dimensions, the fan life expectancy, max TDP, etc. I also try to find user reviews.

For the worthwhile coolers, there are actually reviews out there similar to how JhonnyGuru, HardwareSecrets, and Xbit Labs test PSUs. And typically, they will show how the cooler stacks up against other coolers. To me, that's probably the most useful information in terms of seeing how the cooler will perform.

about the break-in period.

it was as4 that first mentioned that, with a description that made sense.
it said that when it's hot it gets thinner, and combined with the pressure and vibration it would work it's way into the pores and scratches on the metal better.

Exactly.
And most of that break-in happens the first time the CPU/GPU heats up and "melts" the thermal compound. Any subsequent power cycles don't matter as much.

you cant use toothpaste, Flouride is incredibly corrosive.
infact it's one of the few things that can eat through concrete!

I never said it was practical to do it. Just meant to say that it can work... for a little bit.

By the time the Flouride corrodes the aluminum/copper material of the heatsink, the toothpaste will have long ago dried up and allow the temperatures to go high enough to fry the CPU/GPU.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

I want to purchase a thermal paste where I can use it not just for CPU / GPUs. I've been working on TVs a bit. I'd like to find one that's cheaper than AS5 that I can use on just about everything electronic related. The heatsinks on the MOSFETs, the northbridges / southbridges, various microcontrollers, GPUs, CPUs, etc. In this case, would the AS-5 / MX-# be good enough or should I go for something that has a much higher thermal conductivity like the Dow Corning stuff?

For that guy's cooler, it is a joke. But the whole computer is a joke. He saw the words "Gaming PC" and spent 300$ to buy it. 300$ for a custom built gaming PC? He saw the speed of the processor and how many cores it had and just thought it was insanely fast.

It has a freaking APU for cripes sake! APUs aren't what I expect to find in a gaming PC! The person who built it (he bought it new off e-bay) had to have paid less than 300$ for parts. The APU cooler is a bit like the stock, in looks. It's got a printed AMD sticker that was obviously not made by AMD. It looks a little bit like the Wraith (the one it's supposed to come with) but no pipes. And the fan, oh man, it's the cheapest thing I've ever seen!

The thermal pastes here did make a difference. The counterfeit MX-2 (or maybe it was MX-4) that he brought, that was causing it to overheat. Well, I mean, the cheap fan was causing it to overheat. Putting AS-5 on it fixed this. Not really sure why he took the cooler off and put the thermal paste on anyway, but he thinks he knows a lot about PCs and likes to tinker. I guess it brings me money though, so I can't complain too much.

So, assuming AS-5 and MX-4 have similar thermal conductivity, and ignoring what the company that makes it claims they have, assuming they're both around 0.95 W/mK, using his extremely cheap cooler that someone built, would that Dow Corning prevent the new APU from overheating I wonder? Right now, that cooler was good enough for the original APU that came with it (if we were using AS-5 or if he was using the original compound that came with it), but now, AS-5 doesn't help with the new one. It still overheats.

Granted, the proper fix is to replace the cooler, which we're doing, but if we were to put something like that had a real thermal conductivity that was around 4 or 5 times higher than AS-5, do you think it'd prevent the overheating?

There's a bottleneck (or more than one) in his system. The APU never reaches 100% load. The hard drive does. It runs slow as shit, even with the new APU. It's one of the cheapest systems I've ever seen. I told him for 300$, we could have built a much faster system, especially if we went with used parts. He essentially just threw his money away.

[momaka]

Re: Thermal compounds / pads and thermal conductivity.

Granted, the proper fix is to replace the cooler, which we're doing, but if we were to put something like that had a real thermal conductivity that was around 4 or 5 times higher than AS-5, do you think it'd prevent the overheating?

A good cooler alone will stop the overheating.

You would have to use some seriously crappy, dried, fake thermal compound to get bad temperatures with a good cooler.

Like I said, even cheap compounds work fine on CPUs/GPUs/APUs with big metal heat spreaders - as long as the cooler is adequate, of course.

At most, you may see 2-4°C difference in temperatures between AS5 and something cheap. Depending on how close you are running to the max rated temperature of the CPU, 2-4°C may or may not matter.

I have a tall cooler with four heatpipes on my AMD Athlon FX-57 - a CPU rated for 105 Watts TDP. Even in the summer now with my room at 30°C and the PC in which this is in has poor case ventilation, I still don't see my CPU temps go higher than 47°C. And that is also with a 120 mm fan pushing air through the cooler at a whisper-quiet 800 RPM - load or no load. The thermal compound I am using is some cheap eBay no-name 10 gram tube of watery gray stuff (I believe that's the correct technical terms to describe it ). I've been using that same compound on a bunch of other computers too - some of which close to 5 years ago.

I'm sure AS5 would be better. But for some applications, it is just not needed.

I want to purchase a thermal paste where I can use it not just for CPU / GPUs. I've been working on TVs a bit. I'd like to find one that's cheaper than AS5 that I can use on just about everything electronic related. The heatsinks on the MOSFETs, the northbridges / southbridges, various microcontrollers, GPUs, CPUs, etc.

That's a tricky one then.

For chipsets based on flip-chip technology, I would suggest using only AS5 or something with similar thermal stability that won't dry out over time, as some chipsets run quite hot.

For transistors in TVs, are we talking about ones mounted to a heatsink? If so, for those you can use AS Ceramique. Many PSUs use similar white ceramic-based compounds.

And where there is a thermal pad, go with a thermal pad again. Only exception is flip-chips, where you may want to make a copper or aluminum shim.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

Yes, a better cooler will fix the issue, but I'm saying if we weren't replacing the cooler (which we are), would the Dow Corning actually make a difference where it would not overheat? In this case, if we're using an inadequate cooler that is physically incapable of removing the heat that needs to be removed in order to prevent the APU from overheating, would something that has a thermal conductivity of 4 or 5 times greater than AS-5 actually have a positive effect to such a point where we might be able to use inadequate coolers? Do you see what I'm asking?

For the TV question, yes, ones with a heatsink. I know they use white ceramic-based compounds, but would something like AS-5 work or is there some special qualities that the ceramic based stuff has that other compounds simply don't have?

Thanks!

[momaka]

Re: Thermal compounds / pads and thermal conductivity.

Yes, a better cooler will fix the issue, but I'm saying if we weren't replacing the cooler (which we are), would the Dow Corning actually make a difference where it would not overheat? In this case, if we're using an inadequate cooler that is physically incapable of removing the heat that needs to be removed in order to prevent the APU from overheating, would something that has a thermal conductivity of 4 or 5 times greater than AS-5 actually have a positive effect to such a point where we might be able to use inadequate coolers? Do you see what I'm asking?

Yes. I'll try to explain again, as sometimes my writing is not so clear written.

Even if you have the chip's core soldered to the heatsink/cooler, it still
wouldn't make a difference if the heatsink is too small for the application. In other words, if the heatsink/cooler is too small for the application, nothing will help you keep your CPU/GPU/APU from overheating.

On the other hand, if the cooler's performance is just marginal and the CPU/GPU/APU is running just a hair away from its throttling temperature under load, a good thermal compound may lower the temperatures down by 2-4°C, which would be enough to make a difference and possibly have the chip not throttle.

But the main point is, if the chip is running close to max temperature, you are better off investing into a better cooler than a more expensive compound.

For the TV question, yes, ones with a heatsink. I know they use white ceramic-based compounds, but would something like AS-5 work or is there some special qualities that the ceramic based stuff has that other compounds simply don't have?

From what I know, AS5 is slightly conductive, so you would have to be very careful how you apply it there, as it may conduct enough to have something arc. For that reason, I don't suggest AS5 be used on non-isolated power components.

Ceramic compounds on the other hand, are only slightly capacitative on the other hand - typically not enough to cause any issues even when spilled onto other components.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

We put the Thermaltake T4 cooler in, and it's better, but still not super cool. Idling around 56 degrees C. A lot better than overheating. The case only has one intake fan in the front. It's a small case as well. I don't think that helps much.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

Yes. I'll try to explain again, as sometimes my writing is not so clear written.

Even if you have the chip's core soldered to the heatsink/cooler, it still
wouldn't make a difference if the heatsink is too small for the application. In other words, if the heatsink/cooler is too small for the application, nothing will help you keep your CPU/GPU/APU from overheating.

On the other hand, if the cooler's performance is just marginal and the CPU/GPU/APU is running just a hair away from its throttling temperature under load, a good thermal compound may lower the temperatures down by 2-4°C, which would be enough to make a difference and possibly have the chip not throttle.

But the main point is, if the chip is running close to max temperature, you are better off investing into a better cooler than a more expensive compound.


From what I know, AS5 is slightly conductive, so you would have to be very careful how you apply it there, as it may conduct enough to have something arc. For that reason, I don't suggest AS5 be used on non-isolated power components.

Ceramic compounds on the other hand, are only slightly capacitative on the other hand - typically not enough to cause any issues even when spilled onto other components.

Okay, that's what I needed to understand. Thank you.

I see why they use ceramique now. Thanks!

[ratdude747]

Re: Thermal compounds / pads and thermal conductivity.

you cant use toothpaste, Flouride is incredibly corrosive.
infact it's one of the few things that can eat through concrete!

No, no, no. Fluorides are Florine salts. Not corrosive; although if dissoved in water I'd think it would rust steel like salt (NaCl) water.

You're thinking of elemental Florine (F2) Gas. It's the most reactive nonmetal in the periodic table and is indeed nasty stuff. Reacts with everything minus a few Noble gases (Helium and Neon for sure, I can't recall if they've gotten Argon to react or not).

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

Yes. I'll try to explain again, as sometimes my writing is not so clear written. ...

Just wanted to say the question I asked was just stupid. I didn't get a lot of sleep yesterday and when I don't get sleep, my mind don't seem to work well. Now, looking back at the question, it's obvious thermal paste, regardless of how good it is, wouldn't help with a situation like this. When I had asked the question, I wasn't thinking about the purpose of thermal paste. I was thinking the purpose of thermal paste was to help transfer heat from the CPU to the heatsink. And although it might do that, I believe the real reason we use thermal paste is to fill in those little holes.

If the two surfaces (ie, the CPU and the heatsink) were perfectly flat, I don't think we'd need thermal paste at all. Thermal pastes tries to help connecting the two pieces of metals together, where they normally wouldn't touch.

Some thermal pastes are better at transferring that heat than others though, and this is where the thermal conductivity comes into play. Just like how different metals are better at dissipating heat.

[stj]

Re: Thermal compounds / pads and thermal conductivity.

how is airflow in the case??

the best heatsink in the world wont work if airflow is poor.

it's a 3 stage process.

1: the heat is transfered from the chip to the heatsink via the paste.
2: the heat is distributed to the surface of the heatsink - surface-area is more important than volume.
3: heat is removed from the surface of the heatsink by convection or fans.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

I replaced that cheap heatsink in that "gaming system" where the guy upgraded the APU. Airflow is a problem there, but using AS-5 and a proper heatsink, we were able to get the CPU to run at acceptable temps. It still should have a case fan though. It has one in the front, but there should be one in the back. It is a very small case. Not a micro, but I wouldn't call it a mid. I'd call it a mini-tower. Very little room. The heatsink we put on was actually a bit too tall. We able to put the side on, but there's a very slight (hardly noticeable) bend. Thankfully, this didn't affect the fan at all.

It was sooo cramped, the Thermaltake T4 was big and we had to move the RAM. We had to put the two sticks in the two slots furthest from the cooler. That was the only way to get the cooler in there. Otherwise, the fan itself was resting on the RAM and wouldn't seat properly. I know that's not optimal because the RAM should be in every other slot on this system, but it was that or not use the cooler the customer bought.

I think from now on, if a "computer person" brings me a PC to fix, I'm not going to let them help. This guy insisted, but I'm just going to say if you have to help, you have to take it somewheres else. I would have checked the physical size of the fan before purchasing it, just to make sure it fight properly. This guy, he brings me a lot of business, because he's constantly breaking shit. I think he repairs stuff for other people, but doesn't know what he's doing. After he fucks it up, then it he brings it to me and pays me to fix it.

[momaka]

Re: Thermal compounds / pads and thermal conductivity.

We put the Thermaltake T4 cooler in, and it's better, but still not super cool. Idling around 56 degrees C.

56°C is quite high for idle temperature. What model is the CPU/APU. I know the new Ryzen chips supposedly report 20°C higher than actual temperature. But if this is not a Ryzen chip, I would be a bit more worried.

As I mentioned, I like to keep my temperatures below 60°C, especially on AMD/ATI and nVidia hardware as they seem to fail a lot more. Intel CPUs on the other hand, I trust up to 70°C - at least the old generation of chips (i.e. 4th gen i3/i5/i7 and older).

It was sooo cramped, the Thermaltake T4 was big and we had to move the RAM. We had to put the two sticks in the two slots furthest from the cooler. That was the only way to get the cooler in there. Otherwise, the fan itself was resting on the RAM and wouldn't seat properly.

Is that really a ThermalTake heatsink? Or is it a Cooler Master (Hyper T4)?

Reason I ask is because most Cooler Master Hyper coolers have dual fan mounts - that is, you can have the fan "push" air through the heatsink or you can place it on the other side of the heatsink and have it "pull" air through it.

I've had to switch the fan for a few builds like that, where it was hitting the RAM. Add a fan duct behind the fan if the case has an opening, and the CPU fan will also double as a case exhaust fan.

Dell used that design quite a bit in the Pentium 4 era, and it was very effective at getting rid of heat from the case.

The heatsink we put on was actually a bit too tall. We able to put the side on, but there's a very slight (hardly noticeable) bend. Thankfully, this didn't affect the fan at all.

Make sure the case side isn't pushing the heatsink sideways or something like that. This can make it loose or not make good contact with the CPU.

The 4 heatpipe heatsink I am using is also too tall for my micro-ATX case by almost 1 cm. Thus, I just have the case side hanging only by the top metal locks.

I think he repairs stuff for other people, but doesn't know what he's doing. After he fucks it up, then it he brings it to me and pays me to fix it.

Aw man, those are the worst.

The only thing worse than a failed part due to bad design is a failed part due to bad design that someone tried to fix and messed up even more.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

56°C is quite high for idle temperature. What model is the CPU/APU. I know the new Ryzen chips supposedly report 20°C higher than actual temperature. But if this is not a Ryzen chip, I would be a bit more worried.

This is the A10-7890K and yes, I meant the Cooler Master Hyper T4. The baby's been sick. I gotta take her back in. There hasn't been a lot of sleep around here. Over the weekend, I got a little break, because my wife was able to watch her a bit, but it seems I still gotta watch my wife while she watches the baby.

Anyway, APUs I think are a bit hard to read, temp wise. I think that's why AMD released a special program to read them. Maybe the BIOS isn't properly reporting the APU temperature? Also, being an A10-7890K, I'd think it'd run a bit hot. He took it gaming yesterday and I guess it worked well, in the sense it never overheated.

I think to get cooler temps, we'll need to add a rear case fan.

Reason I ask is because most Cooler Master Hyper coolers have dual fan mounts - that is, you can have the fan "push" air through the heatsink or you can place it on the other side of the heatsink and have it "pull" air through it.

Yes, this one you can switch the fan, but it doesn't help. The only thing that would help would be if we could rotate the cooler 90 degrees, which we cannot do. But then we would have to deal with other issues (air flow being stolen from the PSU fan, etc). I think the only option to get this to run cooler is to an exhaust fan. It's a very small case, everything is cramped together. It's a very hot processor. We literally have one fan in front of the hard drives pulling air in and then the PSU fan sucking air out of the case and then blowing it out the back. I just don't think there's enough airflow at this point.

Make sure the case side isn't pushing the heatsink sideways or something like that. This can make it loose or not make good contact with the CPU.

The 4 heatpipe heatsink I am using is also too tall for my micro-ATX case by almost 1 cm. Thus, I just have the case side hanging only by the top metal locks.

It's not pushing the fan or heatsink. I had already checked that out when I installed it. I looked through the back, where the extra case fan can go. It's too bad about the RAM though. It'll never run in dual channel mode now and he's limited to two sticks. But I guess that's what he gets for buying his own parts.

The cooler itself isn't that bad of a cooler. It has a very large case fan (120mm I think) that cools a very large heatsink. There's a cooper plate at the bottom that's connected to the fins with the pipes. It runs fairly quite, especially for being so big.

[momaka]

Re: Thermal compounds / pads and thermal conductivity.

This is the A10-7890K and yes

95 Watt TDP chip.

I don't think it should run that hot with that heatsink. My Athlon 64 FX-57 is 105 Watts TDP max, and it never goes above 47°C, even now in the summer with 30°C/86°F room ambient and a cramped small case.

But I suppose there is also a chance that AMD rated the TDP of those A10 CPUs as "nominal", and the maximum could be higher. Intel did that with the Pentium 4 Prescott, saying they were 89W TDP. In reality, their maximum consumption would hit about 100 Watts.

Either that, or temperatures are not reported correctly.
I don't know what AMD did with their temperature sensors, but pretty much since the AM2 era, they've been very inconsistent with temperature reporting.

I think to get cooler temps, we'll need to add a rear case fan.

It would probably help, yes. Or just tell him to keep once side of the case open. (poor man's cooling)

Yes, this one you can switch the fan, but it doesn't help.

How so? Switching the fan should put it on the other side of the heatsink, so it never interferes with the RAM. Or is his motherboard one of those where the RAM is above the CPU?

It's a very small case, everything is cramped together. It's a very hot processor. We literally have one fan in front of the hard drives pulling air in and then the PSU fan sucking air out of the case and then blowing it out the back. I just don't think there's enough airflow at this point.

Yes, more intake and/or exhaust fans would help, though I also don't like it when they are overdone. (I've seen some people's builds with fans in every slot in the case. It's not bad to cool your PC like that, but there are some ways to cool it just as well with much less fans. One of them is just a big side fan blowing air directly on the motherboard. That alone will allow to the PC to run without any other case fans.)

The baby's been sick. I gotta take her back in. There hasn't been a lot of sleep around here.

Sorry to hear that. Hope she gets better soon. I know from my sister that the first year with baby is the toughest.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

I feel it's too hot, but there are multiple problems that would need to be addressed to get cooler temps. One, the case is just simply too small. There is no way to add a side fan. The whole PC was 300$ new, listed as a gaming PC. There's no window, just a solid case. To add a side fan, we'd need to drill vent holes, mounting holes, etc. I guess it's possible, but not probable.

If I have the PC laying flat on the workbench, with the cover off, where as to where I'm looking down and I can see the inside of the PC with the bottom towards my stomach and the top towards my head, the power supply is directly above the CPU. The RAM is directly to the right of the CPU. The RAM is so close to the CPU. The CPU heatsink can physically only be mounted one of two ways, but the way it's designed, changing the way the heatsink is mounted doesn't affect anything.

The fan can only mount on the heatsink on one of two sides. Either the side closest to the power supply, or the side further from the power supply. It cannot be on the side of the heatsink. So we cannot put it opposite side of the RAM. Regardless of where the heatsink fan is attached, the right hand side of the fan always rests on the two slots. This is why we needed to move the RAM to slots 1 and 2, instead of 2 and 4. For his board, the proper configuration for two sticks is 2 and 4.

There is a very good chance that the board is incorrectly reporting the temps. I believe this particular APU runs very hot to begin with, but if a person where to sell a new computer that they built for 300$, we can assume that the same person that built that PC spent less than 300$ to build it. He has to make a profit. If I where to build this system myself, I think with new parts, I could probably get away with building it for maybe 150$.

Essentially, if he wants a gaming PC, I feel the proper course would be to save up the money and purchase the proper hardware. APUs aren't really associated with gaming PCs to my knowledge. In fact, I think it's quite the opposite, isn't it? Aren't they generally for low-end systems that are made to surf the net, check e-mails, watch youtube videos and are usually found in tablets and similar devices?

If I recall (my memory is all jacked up, sorry), I believe AMD makes a special program to read the APU temps. I'm guessing this program probably reads the same sensors in the APU that the BIOS reads, but it might be worth a shot to have him download it and check what the temps say with the official AMD tool.

The biggest problem is this is just a horrible design! I wish so bad now that I had taken pictures of everything so you could critique it with me. If it comes back, I'll take some pics and we can nit pick what we see wrong. It could help other people who might come here looking for info on how to build a lower end PC. It could teach them the do's and don'ts. I know I wouldn't be able to sleep at night if I built a PC like this and sold it to anyone, let alone as gaming system!

The PSU needs to be replaced. I tried telling him that. It's going to destroy the motherboard and probably a bunch of other stuff along with it. He won't, not until his PC doesn't start anymore. But by then, it'll be too late, the damage will be done.

He saw the numbers of cores and the speed of the process, the amount of RAM it had and made an impulse buy without properly doing his research. Then when it arrived and was slower than dog shit, he decided to upgrade the APU to the fastest APU the board supported, again, without properly doing his research. When the new APU was too hot for the current heatsink and it started locking up, that's when he came to me. It gets a bit old I guess. He takes cell phones and tries to root them and put custom firmware on them, and then when he bricks them, he brings them to me. I don't mind helping people, but I haven't been charging him. His girl lost her job or something and they're struggling for money, but how many times do you have to burn your hand on the stove before you learn not to touch the stove, right?

[stj]

Re: Thermal compounds / pads and thermal conductivity.

adding a fan just needs a drill and a jigsaw.

[Spork Schivago]

Re: Thermal compounds / pads and thermal conductivity.

And an available header on the board or a splitter / adapter to hook into the PSU to provide power to it. I feel a back case fan might be sufficient though. I don't think I want to go through all that work for free. If he wants to pay, I could do it. I got a drill press. But I don't think I'm doing anymore freebees.

The cell phones, he pays me to fix, but this computer, I don't remember ever getting paid. I think he brings it over and asks for advice, and we end up just fixing on it. Unfortunately, usually he's there the whole time we work on it, so everything takes like 10x as long. I start to do something, he doesn't understand what I'm doing, but he thinks he does, and says he already did it, and then I have to stop and explain what I'm actually doing, then he wants to know why I'm doing it, and I just don't got the time to teach him stuff like that, especially for free.

I did suggest he take the computer repair program up at the college though. That's a pretty good one, albeit it's a 2 year program, not a 4 year. The main professor is extremely thorough though (sometimes a little too much!)