When are you going to update this article to cover the new load tester???
The power draw meter also has the ability to measure the power factor, which is what it appears to be drawing compared to what is is really drawing. It is presented as a decimal number between 0 and 1. For example, a device with a power factor of 0.5 is really only drawing 0.5 times (or half) what it appears to be drawing, so you will pay for twice the amount of power you really are using. Of course, the closer that number is to 1, the better. By nature, switching power supplies often have a low power factor.
Explain what apparent power is (RMS volts x RMS amps, but in a way that someone new to electronics would understand). Homeowners are only billed for real power, though.
The drawback is that as [resistors] get hot, their power draw goes down.
In the specifications for Stackpole KAL series (presumably that or similar aluminium-cased power resistors are used), the temperature coefficient for values of 0.1~9.9Ω (higher values are better) is ±50ppm, giving a potential variation of ±1.25% from 25°C to the operating limit of 275°C, which doesn't sound like a big deal to me. (By the way, what is the tolerance of the resistors you used??? Anywhere from 0.1% to 5% is available in that series.)
The WCB series of axial leaded ceramic cased wirewounds is as good in that regard. RNF (metal film) comes in ±10/±25/±50/±100 (the better ratings being unavailable in low values). Carbon-based resistors are the real suckers but their temperature coefficient is
negative - Stackpole states -400ppm for CF series in 10Ω~9.99k; this time higher values are worse, except for <10Ω which are ±400ppm (why both ways???) - which would result in increasing power draw. (RNF and CF can only withstand 155°C, though.)
I'm pretty sure a load tester could be made quiet to even 900W, without using ludicrously large fans (I'm thinking dual 120mm, as used on both the old and new units here) or putting the resistors in danger, and even in a hot environment, as long as a fair temperature gradient was allowed between intake and exhaust. (
NMB says that 1.76CFM of flow gives a temperature rise of 1°C for each watt dissipated, so you could get away with 76.3CFM at 1300W allowing a 30°C rise (which in a PC would be unacceptable, but in a load tester it should be fine, as long as you're careful with the hot air
).)
Information is far more fragile than the HDDs it's stored on. Being an afterthought is no excuse for a bad product.
My PC: Core i3 4130 on GA‑H87M‑D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, Kingston SA400S37120G and WD3003FZEX‑00Z4SA0, Pioneer BDR‑209DBKS and Optiarc AD‑7200S, Seasonic G‑360, Chenbro PC31031, Linux Mint Cinnamon 20.3.
by c_hegge » March 7th, 2013, 5:31 pm 
