Saturday, July 19, 2014

GAW Miners Fury Mods - Power Supply [MOD]

The Fury's power supply has received the ire of miners since GAW launched the product and shipped the first units. There has been some speculation about what might be gained from powering the Fury with a higher quality supply. I've spent a bunch of time testing and trying to discover what can be gained from using a better PSU and I have results.

All testing was done with the original power supply included with the Fury. The original power supply is a 12-volt 5-amp, unbranded, power brick that was manufactured in China. Later orders included a different 12-volt 5-amp brick manufactured by Shenzhen Lianyunda Electronic Company. Perhaps GAW listened to the complaints about the original PSU and changed their supplier, but I can only speculate.

The original PSU pulls less than 1-watt with no load. Powering the Fury at idle, the combination consumes 24-watts at the wall. The Fury at idle draws 1.66-amps at 12.58-volts for about 21-watts. The fan itself pulls around 2.3-watts. While hashing at the recommended 328 clock speed, the stock PSU and Fury pull about 49-watts as measured by a Kill A Watt meter.

We can't make power without heat and the stock PSU makes its share of heat. During testing, the top of the power supply ranged from 135° Fahrenheit to 152°F. The bottom, when place on a solid surface got as high as 179°F and averaged 167°F. These temps were recorded at a relatively cool ambient temperature of 68°F to 70°F.

Stock  Fury PSU

The ATX power supply used with the Fury is a Corsair AX860i certified 80 Plus Platinum. With nothing connected to  it, the AX860i costs 8-watts of power just to have it on. Paired with the Fury running at idle, they pull 29-watts together.

The  ATX power supply uses more energy than the stock PSU at all of the tested clock rates, while powering just one Fury. There's no improvement there.

Fury w/Corsair AX860i ATX PSU

Looking at the numbers, the better, at the least much more expensive, Corsair AX860i power supply had no effect on hash rates, reject rates, or hardware errors. It appears that the Fury's OEM power brick adequately supplies 12-volt power at the tested clock rates.

The temperatures at which the stock PSU operates have had some people concerned. A quality ATX power supply seems like the clear winner in the "possibly NOT burning your house down" category. I didn't measure any temperatures on the ATX PSU because resting my hand on the unit during operation didn't alarm me at all. It's a dull warm and the big ass fan in the top is designed to turn on under higher loads to keep the PSU cool. The stock unit was very warm to the touch. When using the stock PSU, many miners have stood the brick on it's side to maximize the exposed surface area and help keep it cool.

I guess fire is important, but in addition to not dying, miners are looking for lower costs and increase profits. The real benefit of a, high quality-energy efficient, ATX PSU is when powering multiple Fury's. With the clock rate set to 328, 340, and 350, three Fury's, all running off the same AX860i, pulled a total of 135-watts, 139-watts, and 143-watts respectively. That's 45-watts, 46-watts, and 48-watts per Fury, or a savings of 12-watts, 14-watts, and 16-watts at each of the clock rates.

If you have more than one or two Fury's, you might want to consider running them on an ATX power supply. It will save you a bit on electricity, but don't expect much more than a little piece of mind.

Up next... Fury Cooling Mod

Fury Test Bench
Criteria: Watts, amps, volts, hash rates, rejects, and hardware errors are an average of several samples. Data was recorded after running for several hours to allow for the numbers to stabilize. Measurements were made with the following tools: Kill A Watt Electricity Usage Monitor, Fluke 62 MAX Plus Infrared Thermometer, Fluke 87-V Digital Multimeter, Extech EX330 Autoranging Multimeter.

No comments: