Antec Kuhler 920 (Bad Temps): Help?

[Mindsurge]

Finished installing the Kuhler 920 this afternoon, and after a disaster with the pump not running at first (I think the CPU fan control has to be connected for it to work, not just connected to USB; once connected it worked, but my i7 950 hit 100C!) and a mess with the drivers and being detected on the USB...

...went through all of the mess only to end up with temps worse than my Promilatech Megashadow heatsink/air cooling.

It's mounted with Arctic Silver 5, good coverage, solid contact and pump is working:

I'm running 4C+ hotter than I previously was, and don't understand what the problem is at this point.

I'm not using the fans that came with it, instead, using betting 120mm Antec fans, same config (push/pull) out the rear of my HAF-X.

So, I've a few questions:

1) Any chance of damage to my CPU, since it hit 100C for a few minutes?
2) Does the fan control (3-pin) need to be plugged in for pump to work? Or just in the USB (which is all it takes for the pump light to come on) and it was some fluke?
3) Why would I be getting worse temps with this than with my previous air cooling set-up?

Given Kyle's reaction to/review of this cooler, I wouldn't have expected such bad temps.

I'm running an i7 950 (stock), 6GB Corsair Dominator, Gigabyte EX58-UD5 (F7 BIOS), HX850, GTX 580 (stock; Zalman VF3000F cooler) in a HAF-X, with an extra internal 120mm fan I've mounted in the empty optical drive space to cool across my RAM etc.

I would hate to go back to air cooling, as the install is always an annoyance, and I also had plans on using the 920 for my wife's system.

Help/advice?

 

[Tsumi]

1. There is some chance, but unlikely. Intel chips automatically throttle themselves to keep it below 100 C.

2. No clue, someone perhaps can answer that.

3. Depends on your previous air cooling setup. Since you're coming from a roughly equivalent cooler, then it's expected you'll see more or less the same temps. As for why it's way worse, the only thing I can guess at is poor contact, pump not working properly, or fans not working or spinning fast enough.

 

[Mindsurge]

1. There is some chance, but unlikely. Intel chips automatically throttle themselves to keep it below 100 C.

Really? Unlikely? My cores were all maxed at 100C, didn't go over, but perhaps the down-throttling was why, but it still concerned me that any damage might have been done.

2. No clue, someone perhaps can answer that.

I found out that it's the 3-pin CPU connector on the pump that powers it, and the USB just allows you to see readings.

3. Depends on your previous air cooling setup. Since you're coming from a roughly equivalent cooler, then it's expected you'll see more or less the same temps. As for why it's way worse, the only thing I can guess at is poor contact, pump not working properly, or fans not working or spinning fast enough.

Fans are definitely fast enough, and contact is really solid, unless it's not lapped totally flat, but it seems to be. But TIM is spread perfect and contact seems completely solid, so I've no idea.

But... when you say "coming from an equivalent cooler"... what do you mean?

How is my Megashadow tower (air) "equivalent" the the 920 watercooling?

Pump seems to be working properly, going by the readings... it's running at the full 3000+ RPM right now.

I don't understand... should be running cooler, from everything I've seen in review etc. The self-contained unit is certainly more convenient than a giant 350g+ tower hanging off my mobo, but it should also be cooling better. I cant really identify a problem here.

 

[Tsumi]

A common misconception (which [H] sadly does nothing to rectify) is that watercooling is magically always better than air cooling. That is not always the case, the self-contained watercooling units are really only competitive with air coolers in their respective price range. Your Megashadow is a high-end air cooler, which makes it comparable to high-end all-in-one watercoolers.

 

[FrgMstr]

A common misconception (which [H] sadly does nothing to rectify) is that watercooling is magically always better than air cooling. That is not always the case, the self-contained watercooling units are really only competitive with air coolers in their respective price range. Your Megashadow is a high-end air cooler, which makes it comparable to high-end all-in-one watercoolers.

There is a high end air cooler thrown into the mix here and we see a 2 degree difference with the fans cranked all the way up.

I have stated over and over that the two biggest positives with water cooling are usually either a low noise profile and/or the ability to exhaust the heat outside of the chassis. So if there are some misconceptions, I would say a bit more reading is all that is needed.

 

[Mindsurge]

A common misconception (which [H] sadly does nothing to rectify) is that watercooling is magically always better than air cooling. That is not always the case, the self-contained watercooling units are really only competitive with air coolers in their respective price range. Your Megashadow is a high-end air cooler, which makes it comparable to high-end all-in-one watercoolers.

Well, that's... disappointing to hear.

Is the 920 really considered a "high-end" all-in-one cooler?

Shouldn't there still be some difference? At least not slightly worse temps?

As I've said, the 920 is certainly more convenient than having a giant heatsink hanging off my mobo, but I'd think that the 920 would, at worst, equal it's cooling. It's relatively close, but slightly (or slightly more than slightly) warmer.

Maybe I should try to re-seat it, just for the hell of it, but I don't think that's the problem. I've mounted countless coolers over the years, and never had a problem.

If the Megashadow is truly "comparable" to the 920 watercooling solution, well... that's just shocking... and disappointing. I'd think the 920 would be at least somewhat better, despite the Megashadow being a high-end air cooler.

I'm a little baffled.

 

[Tsumi]

Quieter? Likely. Cooler? Depends on the fan speeds you set them at. But to the point where you're hitting 100C? That's not normal, it should be approximately equivalent (maybe 5-10 C difference at most).

 

[Mindsurge]

Quieter? Likely. Cooler? Depends on the fan speeds you set them at. But to the point where you're hitting 100C? That's not normal, it should be approximately equivalent (maybe 5-10 C difference at most).

Heh, no... the "100C incident" was when the pump wasn't actually turned on, so that's irrelevant (thankfully!) at this point.

As for fan speed, I'm using two better Antect fans, TriCools, on medium in push/pull. Definitely enough airflow. With only one fan on my Megashadow, my temps were a few degrees cooler, which again, I'll say sort of baffles me.

Ironically, the rear exhaust fan is actually making more noise, as there was a slower 140mm fan on the HAF-X initially, but with the 120mm pushing faster/more air through the mesh, my case is now louder.

So, as desperately as I want to love the 920 and continue to use it, even on my wife's system, I'm a little iffy about the situation right now.

It's seeming as if the 920 water is no better a solution than the Megashadow air, which is both surprising and disappointing. But, I'll probably stick with it, unless it gets worse, since it's better than a monstrous heatsink hanging of my mobo, which I've always hated.

I just don't get why I cant duplicate [H]'s temps from the review, especially due to the fact that Kyle lives in Texas. I'm running 10C hotter both idle and load than the review, on basically the same CPU.

 

[Tsumi]

I'm not really sure about Antec TriCools being better than stock fans, but if you used the exact same fans as on the Megashadow and the Kuhler, I'm not sure what's causing the Kuhler to be warmer.

Are your settings the same as in the review? Chips vary a lot, especially Intel chips. Some just run hotter than others, you can't use other reviews to tell where your chip should be at.

Besides, [H] uses a temperature probe in the CPU's IHS, whereas your temps are read from the core temp sensors. There's no way these would match up.

 

[Mindsurge]

I'm not really sure about Antec TriCools being better than stock fans, but if you used the exact same fans as on the Megashadow and the Kuhler, I'm not sure what's causing the Kuhler to be warmer.

Are your settings the same as in the review? Chips vary a lot, especially Intel chips. Some just run hotter than others, you can't use other reviews to tell where your chip should be at.

Besides, [H] uses a temperature probe in the CPU's IHS, whereas your temps are read from the core temp sensors. There's no way these would match up.

Right, but what I would think I should have been able to expect would be temps that are cooler than what I was running, which would be closer to their results.

Only big difference in my settings, is I'm using an i7 950, stock, where-as they used a i7 920, both of which I found ran at the same temps for me.

Yep, same fans as I used on the Megashadow (actually, only used one on the air cooler) so, yeah, they're better than the stock fans.

Again, I'll ask the question two questions that are going to help me the most:

- 1) Are these temps, on my set-up, from what you've seen in the screens (idle @ around 40-42C, load around 52-54C) what should be "expected" from this Kuhler 920?

- 2) Is the Kuhler 920 somehow only about "equal" to the Megashadow in cooling?

The worst watercooling solutions are only as good as the best air cooling solutions, but from their review, and others, I'd have expected better temps.

It's odd to see the temps of others dropping by 10C using this (even in a lot of customer reviews, on many varying CPUs), or even the Corsair H50, yet I'm running exactly the same as on the Megashadow.

If people on the i7 920 are seeing drops in temp into the 30C range, I should be seeing similar. On reviews of this and other Corsair coolers, I'm seeing even many people posting they "dropped 10C" in their temps on comparable CPUs... yet, I'm at the same temps?

 

[Mindsurge]

Going back over [H]'s reviews...

Prolimatech Megahalems
Antec Kuhler 920

...the coolers do run nearly the same, which means anyone claiming to have dropped 5-10C from a similar air cooler to the 920 Kuhler is completely full of shit.

It also means that the 920 Kuhler is really no better than the Megashadow, which means it's no better than the highest-end air cooling solution.

I can see that from my own results, but looking through more reviews to confirm, there... it's confirmed.

People should know that before purchasing the Kuhler when coming from a high-end air cooling solution.

There's a few advantages, I suppose, such as not having a giant tower hanging off your mobo which, between the weight on the mobo, and the interference with case airflow, just sucks. Also, exhausting hot air directly out the rear of the case.

In terms of noise, that's debatable... about the same between the two, as a 120mm fan running faster, blowing through mesh as the exhaust fan, is louder than a slower 140mm fan.

Of course, I don't know why [H]'s temps always seem to be 10C cooler than mine... maybe they test in a cold basement, heh... but, there it is, my questions answered, I suppose.

Still, I'm surprise... what would I have to do to drop my temps further?

 

[Tsumi]

Like I said, [H] uses a temperature probe that's stuck into the IHS between the CPU and the heatsink (look up their testing methodology, they mill a small groove into the IHS, put in a temperature probe, and read those readings). You on the other hand are using the coretemp sensors, which is by no means the same.

As I have said numerous times, the only real reasons to go with a all-in-one watercooler vs a top-end air cooler is to take the weight off the motherboard, leave the area around the CPU clear for RAM installation and heatsinks, and if you don't have enough room in your case to put in a large air cooler.