Ultra High-End PSU Roundup
by Christoph Katzer on October 22, 2007 6:00 AM EST- Posted in
- Cases/Cooling/PSUs
Tagan ITZ 1300W Efficiency, Temperatures, Fan Speed and Acoustics
The efficiency is similar to what we have seen from the other two power supplies. The difference we can see here is the nicely even curve which is constantly above 80% with any load above 260W, with a slight drop near the 100% load mark.
The temperatures increase slowly during room temperature testing which is a good sign for a well arranged airflow. Just with overload the temperature is suddenly making a big step of over 10°C within a few minutes. Under the hotter heat chamber conditions the temperature begins to increase at medium load and beyond. Naturally, the fans are also spinning faster at this point, but they are nearing their limit. The temperature of both heat sinks mostly stabilizes by 80% load, beyond which point heat is only increasing slowly.
The push-pull fan configuration is quite effective, and having a more powerful fan on the exhaust also works quite well. In this way there is a kind of vacuum established which lets air come in through the venting holes on the side. The result is that the components are cooler but this kind of power supply still gets very hot and needs proper airflow to function. This can only be done with high speed fans, and with up to 4000 RPM this might be enough. Also with this power supply we see the full speed already reached at half of the full load while in the heat chamber. Under room temperature conditions the fan is steadily spinning faster and only reaches maximum RPM at full load. It's again clear that the amount of airflow necessary to cool a 1300W power supply cannot be achieved without producing noise, and therefore we see up to 45dB(A) under hot conditions. With low speeds at lower loads you can still hear the fan quite easily, but in a PC system this noise will not be as audible as when we had it open and running in front of us.
The efficiency is similar to what we have seen from the other two power supplies. The difference we can see here is the nicely even curve which is constantly above 80% with any load above 260W, with a slight drop near the 100% load mark.
The temperatures increase slowly during room temperature testing which is a good sign for a well arranged airflow. Just with overload the temperature is suddenly making a big step of over 10°C within a few minutes. Under the hotter heat chamber conditions the temperature begins to increase at medium load and beyond. Naturally, the fans are also spinning faster at this point, but they are nearing their limit. The temperature of both heat sinks mostly stabilizes by 80% load, beyond which point heat is only increasing slowly.
The push-pull fan configuration is quite effective, and having a more powerful fan on the exhaust also works quite well. In this way there is a kind of vacuum established which lets air come in through the venting holes on the side. The result is that the components are cooler but this kind of power supply still gets very hot and needs proper airflow to function. This can only be done with high speed fans, and with up to 4000 RPM this might be enough. Also with this power supply we see the full speed already reached at half of the full load while in the heat chamber. Under room temperature conditions the fan is steadily spinning faster and only reaches maximum RPM at full load. It's again clear that the amount of airflow necessary to cool a 1300W power supply cannot be achieved without producing noise, and therefore we see up to 45dB(A) under hot conditions. With low speeds at lower loads you can still hear the fan quite easily, but in a PC system this noise will not be as audible as when we had it open and running in front of us.
27 Comments
View All Comments
EatSpam - Monday, October 22, 2007 - link
PC Power and Cooling will do a custom wiring harness for you, so if you need a different connector configuration, they'll do it for a small fee.NicePants42 - Monday, October 22, 2007 - link
I understand that most machines don't need anywhere near this much power, but those acoustic graphs are important. There aren't many 500-600W PSUs that will remain quiet at 100% load.For my last build, I got the CM Real Power Pro 1000 Watt PSU for $179 after $50 MIR - a far cry from the usual ~$350 price range for 1000W+ PSUs. While I may not use all of the power capacity, I do get very high efficiency and zero noise at an (estimated) 500W load, and the PSU will last that much longer because it's not running at 100% capacity.
After 6 months with the PCP&C 'Silencer' 750 (which cost $200, btw), this 'overkill' PSU is a welcome change, and definitely worth $179.
For those interested, I think svc.com is now offering a $50 MIR, although I got mine from ZZF.
EatSpam - Monday, October 22, 2007 - link
I can't imagine needing this much power. My server is running a lowly Silencer 610 and has 2 Clovertowns, 12gb of FBDIMMs, and 16 HDs. No problems at all.flipmode - Monday, October 22, 2007 - link
1300w PSUs - for the folks that drive tandem dump trucks to get their groceries.AssBall - Monday, October 22, 2007 - link
NEXT UP!AT does a review of a mental facility and finds that in fact, 95% of the patients are actually retarded!
........
magreen - Monday, October 22, 2007 - link
That was so random I couldn't stop laughing...sprockkets - Monday, October 22, 2007 - link
Not quite sure if it was that high, perhaps 100c? If it was 1000c I would be worried :)
Christoph Katzer - Monday, October 22, 2007 - link
Actually is was way over 1k. I just attached new cables and it was running well again. Sorry by the way for not including these ripple tests. There is one prob after another. I actually don't have internet since 5 weeks now which is why there was no review last month. The e-scope function needs an IP address and this is provived by the router from the ISP. This company has just the lamest service I've seen in my life and doesn't seem to have the need to send a new box.jenli - Monday, October 22, 2007 - link
I am curious if two lower wattage PSU gangedtogether would be cheaper, quieter, and more
robust ?
eetnoyer - Monday, October 22, 2007 - link
So when do we see the budget power supply roundup that you suggested when you started this project? You know, the stuff that the other 96% of your readers care about.