Single page Print

Now we're not-cooking

Finally, let's talk numbers, or at least number collection. For some reason, I continue to find myself reviewing products for which there are not standard performance metrics. I hope you can tolerate my improvised test method, I was unable to find an ASTM test that was suitable, and my makeshift lab is surely short on the necessary equipment anyhow. I did however have an immersion circulator and a cooking pot.

Call it a synthetic benchmark if you wish, but what I'm about to describe minimized variables, was easy to repeat, and produced remarkably consistent results over multiple test runs. I started by using my immersion circulator to bring the temperature of a pot of water up to 180ºF. Then I connected to the immersion circulator via an app on my phone and used the app to set the temperature to 70ºF. That meant the heating element would turn off, but the circulation would continue. From there, I immediately moved the pot of water on to the room temperature MasterMeal Maker and started my timer.

All I had to do after that was sit in my chair and record the temperature as displayed on my phone every minute for 15 minutes. I made sure that the volume of the water in the pot was at the same level before each test and that the ambient temperature of the house and heat sink were consistent as well.

Finally, a legitimate use for Android.

I only snagged photos during my last run of tests, but I tested the MMM in six different configurations. I also did one extra run as a control where the pot of water cooled without being placed on the MMM first. Of all my tests, only two are considered official configurations by Cooler Master, the standalone bare aluminum MMM and the copper-topped Cu Edition with its pair of fans. This is The Tech Report though, so I dove a bit deeper by additionally testing the plain aluminum model with the Cu's fans as well as the Cu Edition without its fans.

Like I said earlier, you'll barely even noticed the MMM on your counter.

Fish, you idiot, that only adds up to five different test runs. Yeah, yeah, I'll get to the last two in a bit. They were a product of an accidental discovery and I want you to see the numbers first before I explain the circumstances that produced them.

Side note: a USB battery pack will run a pair of 60mm fans for a ridiculously long time.

The numbers

My apologies for the non-zero origin of the y-axis in the graph below. I'm not trying to fool anyone, it just makes it easier to visualize the small amount of separation between some of the configurations. Also, yeah, these measurements are in Fahrenheit, not exactly the preferred unit of measure for heat sink testing, or you know, the rest of the planet, sorry about that. It's all relative anyway.

I can't confirm that a watched pot never boils, but I can conclusively say that a near-boiling pot always cools.

Surprising no one, our control, the hot pot than never got placed on the heat sink, cooled the least amount during the 15-minute time limit. Its rate of cooling isn't linear, but it's as close to linear as we see. This all makes sense, the greater the temperature differential, the more readily energy transfers. The rate of cooling slows as the temperature drops.

Let's talk about those two lines that start out dropping way faster than the rest. I noticed a strange result in one of my normal tests, before I had added the copper plate to the MMM. The initial drop was extremely quick compared to my previous test and I didn't believe the fans I'd just added were enough to explain it. Sure enough, I had spilled some water on top of the MMM when I moved the pot over to it. That water filled the tiny air gaps between the pot and the heat sink, vastly improving thermal transfer. I threw out those initial results, but I knew two additional test runs would be required to tell the whole story. That's what those two lowest lines are.

Yep, water is a better TIM than air.

I'm posting my raw data below to further help with breaking down the graph above. Some of the results are just too close to call without seeing the actual measurements. Check out the race between the MMM and the MMM Cu without fans. They are never more than about a degree apart. I found that pretty surprising given the extra mass of the copper sheet, but it appears that advantage was canceled out by the heat's need to transfer though the MasterGel and into the main body of the heat sink from there.

With the thermal dissipation of the fans added to the equation, both configurations of the MMM perform better. However, the Cu Edition actually performs slightly worse than the stock MMM. At this point, I couldn't help but think about the rather poor job I did applying the MasterGel and the impact that might be having on performance. I also started wishing that that copper plate was soldered to the aluminum body. There was nothing I could do about it, though, and it's likely to be representative of other real-world assemblies, so I didn't sweat it too much. Technically, the stock MMM isn't even intended to be used with fans, so the Cu Edition beating the fanless MMM is performing as intended.

Click on this one if you need a bit more precision from your kitchen benchmarks.

The most interesting results are the ones I stumbled upon by accident, though. To formalize that test, I used 5ml of water from the heated-up pot to draw an X on the surface of the MMM Cu before moving the pot over to the heat sink. Obviously, the test that included the fans was the best performer, with the airflow allowing for steady dissipation of the energy building up in the heat sink as it was removed from the water. The test without the fans saturated the heat sink so quickly that at the end of 15 minutes its curve had leveled off enough that the non-Cu MMM with fans actually caught up to it. Physics!