Why are WD hard drive specs inaccurate?

Question

After doing my own testing with multimeters, I've edited this question from the original title "Why does WD Caviar Green use more power than Caviar Black?".

I have two old 1TB Western Digital hard drives laying around, and I want to use one in a low power Raspberry Pi system with a tiny PSU. Looking at the power requirements printed on the drives, I see this:

╔════════╦═══════════════╦═══════════════╗
║        ║ Caviar Black  ║ Caviar Green  ║
╠════════╬═══════════════╬═══════════════╣
║ 12VDC: ║ 0.55A         ║ 0.55A         ║
║ 5VDC:  ║ 0.68A         ║ 0.70A         ║
║ Date:  ║ 2009          ║ 2011          ║
╚════════╩═══════════════╩═══════════════╝

The green series is known for its low power consumption, so why would it require more max power? It's also 2 years newer than the black, so I would expect they made efficiency improvements in that time.

Western Digital's spec sheet for the green drive shows:
12VDC: 1.671A peak
It makes no mention of 5V.

The spec sheet for the black drive does not mention voltage at all, but it says:
Average read/write power requirement: 6.8W

Now I measure voltage and current with 2 multimeters, and get quite different results. I recorded the maximum observed current, and the simultaneous voltage.

╔════════╦════════════════╦════════════════╗
║        ║ Caviar Black   ║ Caviar Green   ║
╠════════╬════════════════╬════════════════╣
║ spinup ║ 1.88A @ 11.35V ║ 0.59A @ 11.67V ║
║ write  ║ 0.48A @ 11.77V ║ 0.24A @ 11.92V ║
║ idle   ║ 0.49A @ 11.74V ║ 0.20A @ 11.76V ║
║        ║                ║                ║
║ spinup ║ 0.44A @ 4.95V  ║ 0.45A @ 4.94V  ║
║ write  ║ 0.50A @ 4.93V  ║ 0.56A @ 4.92V  ║
║ idle   ║ 0.36A @ 4.96V  ║ 0.12A @ 5.00V  ║
╚════════╩════════════════╩════════════════╝

The listed specs are nowhere close to the real-world measurements I got. Why? And how can low-power system builders get enough information to spec power supplies?

Answer 1

What you did with testing the voltages / currents of your hard drives ( especially when you are building a low-power system ) is very exemplary, and many enthusiasts / amateurs do not consider this important step. Or maybe simply believe what the seller ( spec sheet ) says.

The spec sheet for WD Caviar that you have posted does include some specifications, but its purpose is being more a commercial brochure rather than proper specification. A more proper spec sheet would be this one. I don't know which exact model is the one you own, but it won't matter so much for the focus of this answer.

The key to the question of voltage / current deviation from specified values is tolerances. The WD HD spec sheet that I have posted doesn't include that information ( perhaps for a good reason ? ). But just for example lets take a look at this Barracuda spec sheet.

In the document section 2.8.3 ( as well as in Table 1 and Table 2 ) we have this information on tolerances :

+-5% and +-10% aren't small numbers at all in our case. And doing some quick math gives us these intervals for voltages :

5V interval = [4.75 - 5.25]V
12V interval = [10.8 - 13.2]V

Now, although these specifications were not from your hard drive, you can safely tell now what is the main reason behind your measurements. Your measured voltage values are therefore very well within the expected scope ( tolerances of your HD cannot be very different from those on Barracuda ).

As for current value, this is an even more unpredictable parameter, and as you can see there is no tolerance value even in the Barracuda spec sheet ( this time definitely for a good reason ). As far as I can tell, your measured value is not alarming at all. So the only problem now is if it does not meet your low-power requirements.

So the bottom line now :

Why are WD hard drive specs inaccurate?
They are not inaccurate, they are just missing the very important piece of information, which is tolerances.

And how can low-power system builders get enough information to spec power supplies?
Perhaps I misunderstood your intention behind this question, but my answer would be to buy a sufficiently able&stable power supply, like you would for any other desktop PC. Since you are building a low-power system, I suppose it will be running most of the time. Therefore don't cut on wattage here, and make sure it is regarded as a reliable/stable power supply. As you have seen so far, giving accurate ( and particularly stable ) numbers on this topic is a tough job. But some power supply unit calculators go as far as giving options for different types of PSU ( by their RPM and size in inches ). You can check Newegg PSU Calculator and eXtreme Power Supply Calculator for reference. Also, if you will be doing some calculations on your own be sure to include the maximum values given by tolerances, and of course some safety margin due to possible random peaks.