[GUIDE] Want to measure power consumption but don't have a Kill-A-Watt?

[autonomous42]

It's easy! (provided you have the supplies)
It's also quick-and-dirty, but it works

All you need are:

An old power cable with the leads stripped
A multimeter
Some alligator clips
Balls of steel (optional)


Disclaimer and warning: This has the potential to be very dangerous as it involves powered, exposed wires. Please be careful and treat any wire like a live one. My table is a bit messy but don't be like me; do this in a clear area. You may also want to try wrapping all exposed contacts with electrical tape for safety and peace of mind. If possible, set up the circuit on an independent circuit breaker/fuse so you can disable the wall power while assembling. If you are unsure of your abilities, do not try this at home. I nor these instructions are to be held liable for personal injury or damage to your multimeter or mining rig. If safety precautions and common sense are followed, this method will provide a quick reading of your unit's power consumption.


1. Attach one alligator clip to one lead of the disembodied power cable
2. Attach clips to your multimeter leads
3. Turn off miner
4. We are testing for the current (Amperes) the unit draws on the mains so the multimeter must be in series in the circuit we're making
5. Clip one multimeter lead on to the exposed cable wire and clip the other on to one of the PSU's prongs
6. Clip the alligator on the other cable on to the other prong
6a. Ensure no unintended contact is made between clips, wires, or leads.
7. Set your multimeter to AC Amps (provided it can measure ≥10 A)
8. Plug in the old power cable and turn on your mining machine
9. Let the gears turn for a bit and read the Amperage. Multiply your Amps by your Volts to get Watts. For me, 4.85 * ~116 ≈ 562.6 W. I like that estimate because it includes some overhead. For a more accurate calculation, multiply the Watts by your PSU's efficiency (90% in my case): 526.6 * .9 = 473.94 W
10. Buy a new card (optional)


https://i.imgur.com/zXlTLuf.jpg

[1714547633]

1714547633

[1714547633]

1714547633

[swordfish6975]

Balls of steel (optional)

not sure its optional... especially in Australia where its 240v....

[autonomous42]

Bump so more people can see

[pand70]

Hmm.

That setup looks a little dangerous...
I think there should be a safety warning somewhere in your OP.

[autonomous42]

Hmm.

That setup looks a little dangerous...
I think there should be a safety warning somewhere in your OP.

Thanks, I've done that now. I'm not going to re-take the picture, however, because I've since given the multimeter back to my friend.

Please, everyone: if you try this be careful! I wore rubber gloves and stood back while the power was reestablished.

[bitcoiner49er]

You can also fill up your auto with gas/petro by filling your mouth and spitting it into your tank; but it's not a good idea. 

[bcpokey]

A kill a watt is about $10. More than worth it.

[Hephaestus]

This doesn't actually provide the most accurate result, because you aren't calculating the power factor.  Volts times amps works fine for purely resistive loads (where the power factor is 1), but very few things are purely resistive, and switching power supplies certainly aren't.
http://en.wikipedia.org/wiki/Power_factor
The Kill-A-Watt actually does measure power factor.

[ripple]

Placing a APC Back-UPS (Uninterpretable Power Supply), or similar, between the mains and your system gives a really simple and accurate measure of load on the battery backup in watts, i.e. your total power usage.  You can read from the desktop monitor the kWh, kW per day and even cost per day with the supplied software. To measure the power usage of cards you are using for mining you can just start and stop mining and examine the numerical difference in readings.

The cost of the unit is quite high so this method is only recommended if you already have one around the office or home, though if you are a regular power user a battery backup unit such as made by APC is highly recommended to avoid disruption to your system in the event of a power failure.  The unit will also hibernate or shut down your system with system commands in the event of prolonged power failure in unattended mode a designated number of minutes before the battery backup expires.

I find this set-up is very useful tweaking cost calculations and finding occasions when increases in hashing speed in a system may have a lower or higher than exactly proportional power usage.