Tech Reaction

Continuing along from Part #1

A few problems and sprung up that will effect the rest of the tests. The USB Temp monitoring probe is some what slow in response and it shows in some of the graphs. It also appears to be off in reading temps.

The main problem with the temp probe is you can only use the probe they provided with the USB attachment, from what they told me its calibrated to each probe. So I cannot use any other probes that I have here on hand, the readings go way off if I use one of my other probes.

The next problem is this probe does not want to placed on bare metal. So by taping the probe to a bare metal Evap I would get really crazy readings. When the probe was removed from the Evap it would read correct. The answer to this problem was to put a small amount of black electrical tape around the probe and then tape that to the Evap.

Although this fixed that problem, it also made the probe read slightly off from actual temps and it would become even slower in reading temps.

So I will have to look for a different way to read the temps.

I did continue on with testing to see what the results would be like. You can see in these next graphs that temps are some what warm in the readings and appear to be about be about 5c – 8c off from actual temps measured with a Fluke meter.

Here is the CPUz screen of the following tests.

First test was the Jinu SS Phase, AMD 965 C3 at 4.0ghz running wPrime 1024.

This is one of the harder tests to run for a Phase unit as it stress all cores of the CPU to 100% use and puts out a heavy continues heat load to cool. The power reading matches a Kill-O-Watt meter very closely so the graph is spot on for power. As you can see it jumps over 100watts for this test and the Phase temp readings show that the Phase goes under a heat load and temps warm and then slow down as the Phase unit catches up to the load.

Second test will be 3DMark06 at 4.0ghz

This test shows the power usage going up as the GPU is working harder, but over all temps are lower because the CPU is not working as hard. There are points in the test where a higher heat load output is shown and that’s around CPU Test #1 & CPU Test #2, the Phase than gets colder as the Phase unit recovers from the load of those tests.

Third test will be 3DMark Vantage at 4.0ghz

Here you can see pretty much the same thing as 3DMark06, all though 3DMark06 uses a little more power than 3DMark Vantage does.

So I will continue to look for a better temp monitoring setup so I can get more accurate readings for this.

I also have a load tester close to being finished and I will run that to get a load curve of the Phase unit. I am looking forward to seeing actual heat loads on the Phase and what readings I will get.