Clear all

Power consumption  

Estimable Member
Power consumption

One thing I hadn't seen much discussion over (though it is generalized in the FAQs) is power consumption of our MK4s. This is basically "out of the box", though with 20 days+ of print time on it so far. This shouldn't be considered definitive.

My measurements should be better than +/- 0.5%, based on previous data compared to data captured via tens-of-thousands $ valued equipment. This is in the US, so 120v and it's on a 20A outlet, 12AWG, and is actually a really short run from the panel to the outlet (roughly 30' wire length). Measuring was done at the panel (see below for why I made this choice) for the specific outlet (it's the only outlet on the branch), so any loss should be accounted for, at least as much as I'm able to at home. I even entirely removed my raspberryPi here to get "just the printer" as much as possible. The Prusa supplied USB drive was inserted. The MK4 was plugged directly into the outlet, no extension cords, power strips, etc.

At a dead idle, my MK4 draws a mean of 13W. Methodology for this was: I turned the printer on and let it idle for one hour, sampling the power consumption every second, and applied a `mean` calculation to that. No pre-heating, homing, leveling, or anything was done, so I believe this would be the floor for MK4 power consumption.

First data point that I have is with Prusament PLA. Hotend was at 215°C and bed at 60°C. Speeds:
 Perimeters: 160
 Small Perimeters: 150
 External Perimeters: 100
 Infill: 120
 Solid infill: 120
 Support Material: 180

All other speeds default. Filament settings were Prusament PLA defaults. Infill style was Gyroid, 10%.

There was a small period of time which I missed the filament running out, so the graph reflects a brief leveling out almost smack in the middle of the print. This example print is a small, what should've been ~14hr print (but as I mentioned, I missed a filament change as I was not home when it happened), but ended up being ~16.5hr. I have data resolution down to the minute available, but for the sake of brevity, I bucketed the data into 1 hour sets. I have another PLA print going right now, and so far the data is almost identical hour-over-hour, so I am gaining confidence that this data is accurate and well representative. 

I'm almost done with another product design feature update, which I'll print in PETG to compare. I actually ran out of ABS, so I'll have to order some before I can do some data capturing with that.

So why measure at the panel, instead of at the outlet? To get an accurate assessment of what your utility would measure [and hence charge for], not what a kill-a-watt or similar would from the outlet. I'm sure that will cause some debate, rightfully so, but any loss or excess between the panel and outlet in this case should've been extremely minimal.

Posted : 12/05/2023 6:56 pm
TopherC liked
Estimable Member

On my test bench:

While running, my MK4 uses ~80-130W. 

While running, my MK3S+ uses ~80-140W


Both machines set to 195C\60C and both running 150% speed.

If I were to log it and average it out, they would probably both hovery very closely to 95W.

Posted : 18/05/2023 1:46 am
Active Member

I also wondered about this and added the power figure and energy consumption of the printer (MK3.9) to my data collection setup. Below is one example. The power and energy figures are collected every second from mystrom switch where the printer is connected. It was printing with PLA during this. My quick take home message from this is during heating particularly the bed power figure peaks (>200W), then during printing it is around 100W.

Posted : 24/10/2023 9:50 am