RoadTest: InnoSwitch 3-Pro GaN Controllable USB Power Supply
Author: wolfgangfriedrich
Creation date:
Evaluation Type: Power Supplies
Did you receive all parts the manufacturer stated would be included in the package?: True
What other parts do you consider comparable to this product?: Any 60W USB-C PD adapter with PPS
What were the biggest problems encountered?: Nothing because of the product itself. My logic analyzer could not deal with 1.65V signals to decode PD protocol messages and my power meter was not 240V compatible to measure efficiency at this AC input voltage.
Detailed Review:
First I would like to thank Element14 and Power Integrations to give me the opportunity to roadtest the InnoSwitch 3-Pro GaN Controllable USB Power Supply.
The system consists of a bare power supply module and comes with a universal power cord which has a 2-prong plug and jack at the 2 ends. There are 4 different AC wall plugs included; I think they are for North America, Central Europe, United Kingdom and Australia. Only the UK one is a 3-prong type and has an additional fuse built in. Without a protective housing, there are AC mains wires and soldering joints exposed, so an abundance of caution and safe handling procedures are not only recommended but mandatory.
The AC cable solution is very clever with the cable adapter and the wall adapter being identical and the AC cord basically being an extension cord with what I call the Mickey Mouse ears connector. This way the module could be packaged as a power brick or wall adapter. Kudos to the mechanical and ID designers. Only downside when using the AC cable is, that the wall plug is double the height, so it sticks out of the wall plug by about 10 cm( 4 “).
| {gallery}InnoSwitch3-Pro GaN PD Supply |
|---|
|
Content: Board and AC power plugs |
|
Board Top view: Top view (with added pin header for voltage measurement) |
|
Board Bottom view: Bottom view |
|
Wall Plug setup: No power cord needed |
|
Cable Brick setup: More distance and flexibility |
Plugging the AC cable into the power supply unit gave me a very uneasy feeling of cracking the board, as there is an isolation slot around the AC input with a very short PCB material bridge to the edge of the board. Together with a board thickness of only 0.8 mm it would be easily possible to crack and separate the board. So I avoided plugging the AC port as much as possible. A proper mechanical enclosure would give some needed rigidity to the area.
My test procedure involves looking at all possible output modes and to determine the power supply efficiency. This includes fixed voltage output modes (5V, 9V, 15V and 20V as per system data sheet and USB PD spec), Programmable Power Supply ( PPS ) modes from 3.3V to 21V and 5V passive mode). To perform those tests, I am using some support devices in the form of:
For all measurements, the voltage is taken at the USB-C connector of the InnoSwitch supply and at the load after going through a 1 meter cable including contact resistance of 2 USB connectors and the sink output connector and 2 isolation FETs on the sink with non-zero Rds_on. The bulk of the measured voltage drop is connectors and cable between the supply and sink.
The fixed voltage measurements are done at 5V, 9V, 15V and 20V with current settings of 1A, 2A, 3A and 3.2A which turned out to be the maximum current above specification before the supply shut itself off (most likely because of measured over-current, but over-temperature could be an option too). E.g. a 5V 1A PD request can draw up to 3.2A from the power supply. Most interesting result is the adaptive voltage increase with higher currents to compensate for higher voltage drops across the delivery path. The tracking at the load is always very close to the set voltage. Voltage increase tops out at around additional 500mV. The measured efficiency is above 90% across the whole range and topping out at 94% at 20V/3A.
| Voltage | Current | V(C15 Inno) | Load Voltage | Current | DC Power | AC Power | Eff |
| Set | Measured | Measured | at 110V 60Hz | ||||
| [V] | [A] | [V] | [A] | [W] | [W] | % | |
| 5 | 1 | 5.27 | 5.09 | 1.0 | 5.27 | 5.8 | 90.9% |
| 2 | 5.41 | 5.03 | 2.0 | 10.82 | 11.8 | 91.7% | |
| 3 | 5.51 | 4.96 | 3.0 | 16.53 | 18.1 | 91.3% | |
| 3.2 | 5.55 | 4.94 | 3.2 | 17.76 | 19.4 | 91.5% | |
| 9 | 1 | 9.17 | 8.98 | 1.0 | 9.17 | 10 | 91.7% |
| 2 | 9.29 | 8.92 | 2.0 | 18.58 | 20.1 | 92.4% | |
| 3 | 9.4 | 8.85 | 3.0 | 28.2 | 30.4 | 92.8% | |
| 3.2 | 9.41 | 8.84 | 3.2 | 30.112 | 32.4 | 92.9% | |
| 12 | Inno PD Supply is not specified for 12V fixed output, compliant with spec | ||||||
| 15 | 1 | 15.26 | 15.07 | 1.0 | 15.26 | 16.7 | 91.4% |
| 2 | 15.36 | 15 | 2.0 | 30.72 | 33.2 | 92.5% | |
| 3 | 15.47 | 14.93 | 3.0 | 46.41 | 49.9 | 93.0% | |
| 3.2 | 15.49 | 14.92 | 3.2 | 49.568 | 53.1 | 93.3% | |
| 20 | 1 | 20.21 | 20.04 | 1.0 | 20.21 | 22.2 | 91.0% |
| 2 | 20.33 | 19.99 | 2.0 | 40.66 | 43.8 | 92.8% | |
| 3 | 20.42 | 19.91 | 3.0 | 61.26 | 65.2 | 94.0% | |
| 3.2 | 20.45 | 19.92 | 3.2 | 65.44 | 69.9 | 93.6% | |
| 3.25 | 20.48 | 19.92 | 3.3 | 66.56 | 71.1 | 93.6% |
The 5V passive mode is entered by using 2x 5.1 kOhm pull-down resistors on the CC1/2 signal lines with the PD Sink controller. Here the current is ramped from 50 mA to 3.25 A with a (expected) constant voltage of 5.0 V. Again, an adaptive voltage increase with higher current is noticed.
| Current | V (C15 Inno) | V (Load) | Current | DC Power | AC Power | Efficency |
| Set | Measured | Measured | Measured | |||
| [A] | [V] | [A] | [W] | [W] | % | |
| 0.00 | 5.07 | 5.06 | 0.00 | 0.0 | 0.0 | |
| 0.05 | 5.16 | 5.15 | 0.05 | 0.3 | 0.3 | 86.0% |
| 0.10 | 5.17 | 5.14 | 0.10 | 0.5 | 0.6 | 86.2% |
| 0.20 | 5.16 | 5.13 | 0.20 | 1.0 | 1.2 | 86.0% |
| 0.50 | 5.20 | 5.12 | 0.50 | 2.6 | 2.9 | 89.7% |
| 0.75 | 5.25 | 5.11 | 0.75 | 3.9 | 4.3 | 91.6% |
| 1.00 | 5.27 | 5.10 | 1.00 | 5.3 | 5.7 | 92.5% |
| 1.25 | 5.30 | 5.09 | 1.25 | 6.6 | 7.2 | 92.0% |
| 1.50 | 5.34 | 5.08 | 1.50 | 8.0 | 8.7 | 92.1% |
| 1.75 | 5.37 | 5.07 | 1.75 | 9.4 | 10.2 | 92.1% |
| 2.00 | 5.40 | 5.05 | 2.00 | 10.8 | 11.7 | 92.3% |
| 2.25 | 5.43 | 5.04 | 2.25 | 12.2 | 13.3 | 91.9% |
| 2.50 | 5.46 | 5.02 | 2.50 | 13.7 | 14.8 | 92.2% |
| 2.75 | 5.50 | 5.00 | 2.75 | 15.1 | 16.4 | 92.2% |
| 3.00 | 5.52 | 5.00 | 3.00 | 16.6 | 17.9 | 92.5% |
| 3.25 | 5.54 | 4.97 | 3.25 | 18.0 | 19.6 | 91.9% |
The PPS mode allows setting voltage in 20 mV steps and current in 50 mA steps. Current setting is not really relevant with the unit as it always tops out at ~3.2A independent of the setting. Most noticeable is the fact that the adaptive voltage increase is only active between the standard voltages (5V, 9V, 15V and 20V) and standard voltages + 40 mV. All other values are output as requested with a tolerance of a few tens of millivolts. The limits of 3.3V and 21V are strictly enforced, any value higher or lower are simply rejected, which is the correct behaviour. The table also shows the set voltage at the source and sink at 0A with the load turned off.
| Voltage | V(C15 Inno) | Voltage | Current | DC Power | AC Power | Eff | V(C15 Inno)@0A | Voltage@0A |
| Set | Measured | Measured | at 110V 60Hz | Measured | Measured | |||
| [V] | [V] | [V] | [A] | [W] | [W] | % | [V] | [V] |
| 3.28 | PD Buddy red LED on shows that no good contract could be negotiated! | |||||||
| 3.30 | 3.33 | 2.49 | 3.00 | 10.0 | 11.1 | 90.0% | 3.37 | 3.36 |
| 3.32 | 3.35 | 2.55 | 3.00 | 10.1 | 11.2 | 89.7% | 3.39 | 3.38 |
| 4.90 | 4.92 | 4.42 | 3.00 | 14.8 | 16.1 | 91.7% | 4.96 | 4.96 |
| 4.92 | 4.94 | 4.44 | 3.00 | 14.8 | 16.2 | 91.5% | 4.98 | 4.98 |
| 4.94 | 4.96 | 4.46 | 3.00 | 14.9 | 16.2 | 91.9% | 5.00 | 5.00 |
| 4.96 | 4.98 | 4.48 | 3.00 | 14.9 | 16.4 | 91.1% | 5.02 | 5.02 |
| 4.98 | 5.00 | 4.50 | 3.00 | 15.0 | 16.4 | 91.5% | 5.05 | 5.04 |
| 5.00 | 5.53 | 5.03 | 3.00 | 16.6 | 18.1 | 91.7% | 5.17 | 5.16 |
| 5.02 | 5.53 | 5.03 | 3.00 | 16.6 | 18.1 | 91.7% | 5.17 | 5.16 |
| 5.04 | 5.53 | 5.03 | 3.00 | 16.6 | 18.1 | 91.7% | 5.17 | 5.16 |
| 5.06 | 5.08 | 4.59 | 3.00 | 15.2 | 16.7 | 91.3% | 5.13 | 5.12 |
| 5.08 | 5.10 | 4.60 | 3.00 | 15.3 | 16.7 | 91.6% | 5.15 | 5.14 |
| 5.10 | 5.13 | 4.63 | 3.00 | 15.4 | 16.8 | 91.6% | 5.17 | 5.16 |
| 6.00 | 6.02 | 5.52 | 3.00 | 18.1 | 19.6 | 92.1% | 6.06 | 6.05 |
| 7.00 | 7.01 | 6.51 | 3.00 | 21.0 | 22.8 | 92.2% | 7.06 | 7.05 |
| 8.00 | 8.00 | 7.51 | 3.00 | 24.0 | 26.0 | 92.3% | 8.05 | 8.04 |
| 8.98 | 8.98 | 8.48 | 3.00 | 26.9 | 29.1 | 92.6% | 9.03 | 9.02 |
| 9.00 | 9.40 | 8.91 | 3.00 | 28.2 | 30.5 | 92.5% | 9.05 | 9.04 |
| 9.02 | 9.40 | 8.91 | 3.00 | 28.2 | 30.5 | 92.5% | 9.05 | 9.04 |
| 9.04 | 9.40 | 8.90 | 3.00 | 28.2 | 30.5 | 92.5% | 9.05 | 9.04 |
| 9.06 | 9.06 | 8.56 | 3.00 | 27.2 | 29.3 | 92.8% | 9.11 | 9.10 |
| 10.00 | 10.00 | 9.50 | 3.00 | 30.0 | 32.4 | 92.6% | 10.06 | 10.04 |
| 11.00 | 11.00 | 10.50 | 3.00 | 33.0 | 35.5 | 93.0% | 11.06 | 11.05 |
| 11.94 | 11.93 | 11.44 | 3.00 | 35.8 | 38.6 | 92.7% | 12.00 | 11.98 |
| 11.96 | 11.96 | 11.46 | 3.00 | 35.9 | 38.4 | 93.4% | 12.02 | 12.00 |
| 11.98 | 11.98 | 11.48 | 3.00 | 35.9 | 38.6 | 93.1% | 12.04 | 12.02 |
| 12.00 | 12.00 | 11.50 | 3.00 | 36.0 | 38.7 | 93.0% | 12.05 | 12.04 |
| 12.02 | 12.00 | 11.51 | 3.00 | 36.0 | 38.7 | 93.0% | 12.07 | 12.06 |
| 14.96 | 14.93 | 14.43 | 3.00 | 44.8 | 48.1 | 93.1% | 15.00 | 14.98 |
| 14.98 | 14.95 | 14.45 | 3.00 | 44.9 | 48.1 | 93.2% | 15.02 | 15.00 |
| 15.00 | 15.47 | 14.98 | 3.00 | 46.4 | 49.8 | 93.2% | 15.14 | 15.13 |
| 15.02 | 15.47 | 14.97 | 3.00 | 46.4 | 49.9 | 93.0% | 15.14 | 15.12 |
| 15.04 | 15.47 | 14.98 | 3.00 | 46.4 | 49.8 | 93.2% | 15.14 | 15.12 |
| 15.06 | 15.03 | 14.53 | 3.00 | 45.1 | 48.5 | 93.0% | 15.10 | 15.08 |
| 19.98 | 19.89 | 19.39 | 3.00 | 59.7 | 63.9 | 93.4% | 19.96 | 19.94 |
| 20.00 | 20.41 | 19.91 | 3.00 | 61.2 | 65.5 | 93.5% | 20.10 | 20.07 |
| 20.02 | 20.41 | 19.92 | 3.00 | 61.2 | 65.6 | 93.3% | 20.09 | 20.07 |
| 20.04 | 20.42 | 19.93 | 3.00 | 61.3 | 65.6 | 93.4% | 20.09 | 20.08 |
| 20.06 | 20.00 | 19.49 | 3.00 | 60.0 | 64.1 | 93.6% | 20.06 | 20.05 |
| 20.90 | 20.82 | 20.33 | 3.00 | 62.5 | 66.7 | 93.6% | 20.90 | 20.87 |
| 20.92 | 20.84 | 20.35 | 3.00 | 62.5 | 66.7 | 93.7% | 20.92 | 20.89 |
| 20.94 | 20.86 | 20.37 | 3.00 | 62.6 | 66.8 | 93.7% | 20.94 | 20.91 |
| 20.96 | 20.88 | 20.38 | 3.00 | 62.6 | 66.9 | 93.6% | 20.96 | 20.93 |
| 20.98 | 20.90 | 20.40 | 3.00 | 62.7 | 67.0 | 93.6% | 20.98 | 20.95 |
| 21.00 | 20.92 | 20.43 | 3.00 | 62.8 | 67.1 | 93.5% | 21.00 | 20.97 |
| 21.02 | PD Buddy would not allow voltage setting of 21.02 V | |||||||
With the roadtest deadline looming and whooshing past me, my initial roadtest review is ending here. I have plans to continue testing the unit, decoding the PD packets with a logic analyzer and PulseView/Sigrok software. I have seen decoded PD messages this way before but could not get consistent results with the existing setup, which showed high signal levels of only 1.65V which was low for my logic analyzer. A little voltage translator should do the trick. Also I am awaiting the delivery of a Power-Z or TC66C USB tester, which I want to use to confirm the test unit’s capabilities.
Overall I am really impressed by the efficiency and the adaptive voltage settings in a small package with some clever interface design elements.
