Framework 16 Charging Efficiency-60W/100W/140W/240W

I remember somebody posted the link to this one - https://www.digikey.com/en/products/detail/delta-electronics/ADP-240KB-BA/23026419

But can’t find who it was

2 Likes

Thanks for all the numbers and discussion this far!

Already now: sorry for the wall of text. :slight_smile:

I thought I will chip in as a brand new owner of a FW16 and put some rather unscientific measurements into this thread.

I have opted for a setup with measuring the power at the wall socket with a meter [1] instead of a USB-C one. The main reason was that I found the only USB-PD 3.1 compatible device [2] in my region too cheap to be trusted to handle 100+W loads :sweat_smile:

So… I had the following at my disposal:

  • A 65W HP charger block [3]
    • Even though it’s nowhere stated that the charger supports USB-PD, I trusted that HP puts out a compatible device, based on the listed, officially compatible devices at the page and also that I have already charged smartphones with this charger.
  • A 100W docking stationg from i-tec [4]
    • I love this docking station, I can only wholeheartedly recommend it (no affiliation)
    • These measurements have been done with everything else disconnected from the docking station.
  • A 100W GaN charger block from Baseus [5]
    • I assume that 100W is the total available power on all 4 ports, I have assumed that 1 port can sustain around 60-65W. Look at the table for results :wink:
  • A 112W charger block from Club 3D [6]
    • I assume that 112W is the total available power on all 4 ports, I have assumed that 1 port can sustain around 60-65W. Look at the table for results :wink:
  • The FW 180W charger block

For good measure for the blocks I used a cable with an integrated power display [7], and even though according to the not-existing specifications of this no-name-product it supports up to 240W, I don’t trust it with the FW charger :upside_down_face:

Originally I wanted to separately stress CPU and GPU on a Live USB of Linux Mint 22. For CPU I wanted to use y-cruncher [8], but there is a bug where the software does put 100% load onto all cores of the CPU but it does not want to pull more than 10W on the CPU, even if the FW charger is plugged in, and thus it puts out benchmarks comparable with Core i3 8121U… (low power laptop CPU, which is not a surprise as it does not want to draw power, bug opened at GitHub [9]). So, at the end I have only tested with Superposition [10].

Method:

  • Put the monitor to the lowest brigthness possible
  • Plug the selected charger into the wall socket power meter
  • Plug in the laptop
  • Wait for 100% battery
  • Start Superposition
  • With high scientific accuracy (also known as “eyeballing it”), look at the power meter and record outliers and get a grasp of the stable load (if there is any)
  • Record results on paper
  • Start the next test as soon as possible (just after jotting down the results to paper and selecting the next highest benchmark; one benchmark runs for around 2 minutes, the whole cycle with one charger is thus around 10 minutes with short breaks)

Power in Wh min/stable/max; Points reported by Superposition, FPS is min/avg/max (as reported by Superposition), GPU temperature is ºC min-max (as reported by Superposition).

Note about temperatures: the i-tec measurements are pure, all the others were on top of a Cooler Master Ergostand III [11] going full blast (providing -5ºC when idling both for CPU and for GPU, getting it down to 42.9ºC CPU/39º GPU from 47.4ºC CPU/44º GPU as reported by sensors).

Scenario↓/Charger→ HP 65W i-tec 100W Baseus 100W† Club 3D 112W† FW 180W
Suspend¹ 2.2-2.3 Wh 4-4.3 Wh 3.5-3.6 Wh 2.4-2.8 Wh 1.8-2.1 Wh
Baseline² 18 Wh 25-26 Wh 15-16 Wh 28 Wh 21-22 Wh
Max light³ 26 Wh 32 Wh 27-28 Wh 34-35 Wh 30-31 Wh
720p 56/62/63.5 Wh
12431 points
14.11/92.98/142.60 FPS
31-66ºC
78/85/91 Wh
12579 points
14.16/94.09/149.35 FPS
32-67ºC
67/70/72 Wh
12090 points
13.91/90.43/139.63 FPS
31-62ºC
67/68.7/71.8 Wh
12048 points
13.62/90.12/139.62 FPS
27-58ºC
90/95/101 Wh
12850 points
13.74/96.11/158.24 FPS
31-75ºC
1080p medium 62.7/63.0/63.7 Wh
5067 points
11.75/37.90/48.56 FPS
31-71ºC
79/85/92 Wh
4962 points
11.35/37.12/48.59 FPS
32-75ºC
68/69.5/70.7 Wh
4917 points
12.2/36.78/47.23 FPS
32-65ºC
68/71.4/72.6 Wh
4797 points
11.70/35.88/49.26 FPS
29-63ºC
99.8/101.5/106.5 Wh
5301 points
11.42/39.65/51.79 FPS
31-80ºC
1080p high 62.5/63.2/63.9 Wh
3533 points
13.88/26.43/32.34 FPS
32-70ºC
80/86/92 Wh
3472 points
9.91/25.97/62.62 FPS
32-74ºC
68.7/69.5/69.7 Wh
3376 points
10.22/25.25/30.98 FPS
32-66ºC
70.8/71.5/73.8 Wh
3386 points
9.83/25.33/31.15 FPS
30-65ºC
96/101.7/108.8 Wh
3643 points
10.51/27.25/33.51 FPS
31-80ºC
1080p extreme 62.5/63.3/63.9 Wh
1632 points
10.06/12.21/15.44 FPS
32-72ºC
79/86/93 Wh
1575 points
7.39/11.78/15.32 FPS
33-78ºC
67.3/68.1/69 Wh
1556 points
6.89/11.64/14.66 FPS
32-66ºC
68.5/70.9/71.5 Wh
1533 points
7.04/11.47/14.68 FPS
31-65ºC
101.2/106.8/107.6 Wh
1690 points
7.52/12.64/16.09 FPS
31-83ºC

¹ Laptop suspended and waited for a stable reading on the meter
² Running without any load, with lowest brightness on the screen, baseline value
³ Running without any load, with highest brightness on the screen
† Note: based on the measurements with the two charger blocks, I have seen a difference of -4Wh to +10Wh reported on the cable vs. the power meter. Both have a 1 second tick for the measurements, which weren’t in sync.

I tried to draw some parallels based on the public Superposition measurement and came to the following:

Charger Roughly equivalent to
HP 65W CPU: i7-4770/Ryzen Z1 Extreme
GPU: GTX 960/ASUS Phoenix
i-tec 100W CPU: i5-4690K/Ryzen Z1 Extreme
GPU: GTX 1050 Ti/GTX 1080
Baseus 100W CPU: i7-6700HQ/Ryzen 7 2700X
GPU: GTX 1050 Ti/GTX 970
Club 3D 112W CPU: i7-4770/Ryzen 7 2700X
GPU: GTX 1050 Ti/GTX 970
FW 180W CPU: i7-6700/FX-8320
GPU: GTX Titan/Radeon RX 580

(Links broken up, because as a new user I’m not allowed to post more than 2…)

[1] www.brennenstuhl[dot]com/de-DE/produkte/reise-steckdosenadapter/primera-line-energiemessgerat-pm-231-e
[2] www.amazon[dot]de/Tester-Messger%C3%A4t-Multimeter-Strommesser-Spannung/dp/B0D9YKHP79/
[3] www.hp[dot]com/de-de/shop/product.aspx?id=671R3AA&opt=ABB&sel=ACC
[4] i-tec[dot]pro/en/produkt/c31dualdpdockpd-2/
[5] eu.baseus[dot]com/de/products/gan3-pro-4-ports-desktop-charger-100w
[6] www.club-3d[dot]com/en/detail/2562/power_charger_4_ports_2x_usb_type_a_2x_type_c_up_to_112w_power_delivery(pd)_support/
[7] www.digitec[dot]ch/en/s1/product/onit-usb-c-usb-c-1-m-usb-40-usb-cables-39380775?tagIds=77-532
[8] numberworld[dot]org/y-cruncher/
[9] github[dot]com/Mysticial/y-cruncher/issues/52
[10] benchmark.unigine[dot]com/superposition
[11] www.amazon[dot]de/Cooler-Master-Notebook-K%C3%BChler-Einstellbarer-R9-NBS-E32K-GP/dp/B00PC2X1WU/?th=1

4 Likes

Framework themselves did x.com

I actually found a listing on eBay for that charger shipping from China but I’m not willing to spend $200 to see if it’s real.

3 Likes

That article sounds more meaningful than it actually is.

All certified passive (ie. without internal chips) USB4 v1.0 40 Gbps cables (as well as passive Thunderbolt 3 and 4 cables) have been retroactively updated to USB4 v2.0 80 Gbps. USB4 v2.0 80 Gbps has only a minor increase in signal integrity requirements over the prior generation, as a result the designers decided against increasing the requirements for the cables.

That article is about the first cable available to be certified for USB4 v2.0 80 Gbps from the get-go, however IMO that’s meaningless as all passive USB4 40 Gbps cables are certified to meet the same standard and have had 80 Gbps support retroactively enabled. The first USB4 40 Gbps 240W cables were certified nearly 2 years ago.

4 Likes

Digikey has stock available now for $126 if you are interested.

1 Like

I like that the price dropped but it looks like those don’t ship until December. I’m keeping an eye on Mouser though because they’re supposed to get 30 in stock next month.

It looks like maybe DigiKey chargers will be shipping at the end of the month.

5 Likes

Oh that’s awesome. I’m ordering one now. Thanks for putting the effort into checking with them.

4 Likes

Just got a charge from DigiKey, and the website has updated to show the chargers can ship immediately.

5 Likes

Same for me here ^^

5 Likes

Just got an email that it shipped.

1 Like

Just got the delivery estimated time, I hope you don’t mind it being in french lol
It ships in only two days ! (fingers crossed)

1 Like

Any plans on posting a review / photos when you get it?

Somewhat interested in what that charger looks like but am not in the position to blindly purchase at the moment!

1 Like

Mine just shipped out. Should arrive Monday. I’ll do some testing to see if it makes any performance difference overall but unfortunately I do not have/can’t afford to buy a USB-c multimeter that actually supports 240w safely/accurately.

1 Like

I can take some photos and do a bit of a comparison. I really want to take the new charger out for a spin once I have the thermal issue with my laptop resolved, so it may be a few days before I have some numbers. I don’t have any hardware tools like the PD measurement either, but I should be able to pull everything such as computer sensors.

3 Likes

Would https://www.amazon.de/dp/B0B5W5NKKN be an option? I’ve got it and looks good, but can’t test the full range due to lack of 240W adapter and not having a dGPU.

That would be awesome. I’m waiting to confirm it has a replaceable usb cable, just like the Framework adapter, before I purchase. I’m not wasting money on any more power adapters that become useless after the usbC end gets damaged.

Looking at the datasheet available on Mouser, it looks like an attached USB-C cable 1.6m long and no included wall plug/cable, though it looks to be the same style wall cable as the 180W Framework one.

1 Like

Got the charger in the post today, I’ll have some photos up later on after I finish some work around house.

4 Likes