People are enthralled by the USB-C Power Delivery (PD) specification and charging graphs.....right guys?....surely that can't just be me.....? Either way, I'll assume you want to read about how we test and visualize mobile device charge time.
We've been including the mobile charge test in ShortCircuit videos since December 2024, but we've recently created a new visualization that adds much more context on the charging process. In this article we'll cover the test that we currently conduct, our new visualization (along with some of its limitations), and an interesting finding from the Motorola razr ultra 2025.
Test Setup
The block diagram below shows the test setup, with the Quarch QTL2843 serving as the sole measurement device. This device serves many functions (connections, power delivery, measurement, monitoring) and allows the setup to be fairly simple. The QTL2843 is placed between the mains grid (wall outlet) and the device being charged. It monitors and logs the electrical characteristics to the USB or LAN connected computer.
The QTL2843 is able to record at a 8 kHz sampling rate on three channels, up to 400V and 10A. We don't use this setup for anything particularly high power (we have a fancy setup for that) so we're often well below these limits. We've also reduced the sampling rate to 1 kHz which is plenty for our testing, and merciful to our data storage.
One of the benefits of this system is the integration with Quarch's Quarch Power Studio(QPS). This software allows for simple real-time monitoring of the measurements as well as some light analysis. Quarch also provides a Python library for interface, control, and automation of the Power Analysis Modules (PAMs), but we haven't required that yet.
The standard USB-C charging brick (colloquial term for an AC adapter) that we use is the Anker Nano II 45W. This provides enough power to allow devices to charge quickly while still representing an accessible charging solution. If the device is packaged with a charging brick (increasingly rare), then we will also conduct the test with the manufacturer provided charger as it may enable special features or increased charging rates.
Testing
The main steps of the charge test are as below:
Ensure the charging settings are in the default or out-of-box configuration. Some devices have modes that will sacrifice battery health for high wattage (and faster) charging. We don't enable those initially, but time permitting, we will also conduct a charging test under those conditions.
Disable "optimized charging"[1] that will slow or halt charging.
Ensure the device is fully discharged by attempting to turn it on. If it is not fully discharged then we will run benchmarks until charge depletion and let the device cool down to room temperature before conducting the test.
Configure and begin recording with the QTL2843 followed by plugging the AC adapter and device into one of the monitored outlets.
Keep the device turned off (or turn it off if it turns on automatically) to minimize inconsistencies and interactions during the test. Turning the phone off can sometimes limit the charging rate, so if we're not reaching the maximum rate then we will try the test with the device turned on.
When charging has completed - usually indicated by a step change to zero watts in the charging rate - the recording is stopped and we note the total duration as well as download the full results.
So far, the measured and published result from the test has been the total duration from when the device is plugged in, until the device stops drawing power. This is informative and reflects the way most devices are advertised. However, presenting the entire test as a single duration has limitations as it doesn't show the full charging curve.
While a phone may enable 100W charging, how long is it actually pulling 100W, and does that significantly speed up the typical charging process, or is it only useful to get from 0% to 5%? Additionally, if we are only presenting the time to 100% charge then it biases against more conservative charging schemes. Many devices significantly slow down charging toward the end, so it may take 60 minutes to reach 80% charge, but an additional 30 minutes if you want the remaining 20%.
Our New Graph, and How It's Lacking
In an attempt to more comprehensively capture and communicate the charging rate and duration, we're now experimenting with graphing the wattage over time. This test also still has limitations, but despite them, I think it can provide some interesting information.
First, let's cover some of the known limitations of this testing, then we'll get to evaluating the results.
Measurement Point
We are measuring the wattage 'at the wall', and not at the USB-C cable. This means that our measurement also includes the energy that is 'wasted' (converted to heat) by the charging brick (AC adapter). Therefore if the measurement shows 50W, the device is likely only charging at around 40-45W, depending on the efficiency of the AC adapter.
We have tried integrating the Quarch QTL2928 to measure the USB-C power, but the device is limited to 16.5V so it is incompatible with many of the higher charging rates. (requiring 20V) We're looking at other USB-C measurement solutions so let us know if you have a favourite.
Reported Battery Percentage
The reported/displayed battery percentage of the device isn't monitored. This would likely require a monitoring service, or to keep the device (and display) on so that we could point a camera at it, affecting the test.
Instead, the area under the graph of wattage against time is calculated (using the trapezoidal rule), representing the total energy delivered to the AC adapter. This can then be analyzed to find the times when 10%, 25%, ..., 100% of the total energy has been delivered.
AC Adapter Efficiency
The efficiency of the charging brick will be dependent on the output wattage. Like computer power supplies, efficiency can vary depending on how heavily it is loaded. For the purpose of our test we assume constant (wattage independent) efficiency, but in the future we will more systematically evaluate the efficiency of power bricks to determine how variable the efficiency is.
USB-C is Complicated
USB-C is complicated, involving a lot of communication and handshakes between the device and the charging brick. Sometimes the handshake 'fails', or a charging brick isn't completely compatible with a device. This may limit the charging rate or cause two seemingly identical charge tests to produce different results.
It may be neat to eventually have a way to monitor and publish the supported USB-C Power Delivery (PD) modes.[2]
Results
Returning to our results and graph, the main trace/line displays the wattage delivered to the AC adapter for the duration of charging. With the caveats above, this can be considered the charging rate, and can communicate how 'aggressive' manufacturers are being with the charging rate and battery health.
Additionally, there are vertical dashed white lines accompanied by a percentage and duration. These are the times at which it reaches 10%, 25%, ..., 100% of the total energy delivered to the AC adapter. This is not the percentage that the phone will report. There are many layers between the actual amount of energy consumed and the reported battery percentage. (including: AC adapter efficiency, heat lost in the phone battery while charging, determining the charge level of the battery, battery charge level to reported battery percentage mapping, and likely more)
I believe that the wattage trace and energy delivered percentages provide a more complete illustration of how a device charges, though there is certainly more work to be done here so this isn't the final state of the graphing or testing.
As always - especially with these orange highlighted "under development" graphs - we'd love to hear any feedback about the testing or visualization.
Bonus Motorola razr ultra 2025 Results[4]
While testing the Motorola razr ultra 2025, I initially conducted the charge test with our 'standard' 45W charger and as expected it drew the maximum of 45W for a brief period before gradually stepping down the wattage. However, repeating the test with a higher wattage charger and "Charge Boost" activated still didn't get us to the claimed 68W charge rate.
For fun, we tested with a few other AC adapters - from various brands - which produced similar (non-68W) results. With some further digging, we found that the 68W charging is only enabled on the Motorola razr ultra 2025 with the use of their specific 6.5A charger and cable.
It is entirely possible to deliver 68W(even up to 240W with EPR) within the USB-C Power Delivery(USB-C PD) specification while adhering to the 5A limit of the standard.(by increasing charging voltage to roughly 14V) This would enable 68W charging with any USB-C PD equipped charger[3] instead of requiring the separate purchase of the Motorola charger and cable that are cable of 6.5A.
It's not completely clear to us why Motorola chose a proprietary solution over adhering to the standard. The simplest explanation is that they already possessed the 6.5A charging technology (developed for the Motorola Edge 40 Pro) and it was relatively simple to port it over to this device. Another possibility is that the design changes to the phone to enable greater than 11V charging were larger than the design changes to enable 6.5A charging. Especially considering that the phone will only be charging at 6.5A - and generating excess heat - for a short period at the beginning of the charge cycle.
Regardless of the reasoning, regular charging of the Motorola razr ultra 2025 will be limited to around 50W.
[1] Or similarly named technologies. in 2025 most of them now claim that AI is optimizing that for you.
[2] In case this is too subtle, we're definitely looking into this.
[3] ...any USB-C PD equipped charger that supports APDO(Augmented Power Delivery Object). This is a fairly common feature for modern chargers.
[4] I have a compulsion to put objects in our Lumafield Neptune CT scanner, I scanned this device.
