Diagnostics Power Draws (Nintendo Switch)

Once we have confirmed we have a working BQ24193 it is time to hook up our bench power supply to the Switch for the remainder of all our diagnosis and repair.

Without a bench power supply telling us the current draw we are flying blind with nothing more than a multimeter and the hopes of a full working LCD and console. However, with a bench power supply we can tell if a system is booting fully into the OS with 100% certainty without even having a screen connected.

We can tell if the eMMC is bricked/bad or stuck in first-stage boot not going to second stage boot. We can tell if we have dead shorts, if we are failing to go to sleep mode, and so much more. All from the current draw of the bench power supply. So if you are going to be repairing Switch consoles I cannot recommend enough a good bench power supply.

We sell a nice priced mid-range bench power supply with accurate current readings enough for the task. A high-end one will cost you several thousand, but these mid-level accurate bench power supplies are good enough. Cheaper ones are of no use, their current measurements are nowhere near accurate enough to rely on for any diagnostics.

It is simple to power up the console from the bench power supply. Install the 10K bypass resistor between the two test pads above the battery connector, remove the battery, and connect your bench power supply up to the ground of the USB-C connector and the VBAT pin of the BQ chip as shown below.

Set your bench power supply to 4.2V and 2A current limit. Any less than 2A and you will fail to get into second-stage boot.

In the spirit of bypassing as much as possible to keep potential issues like a bad ribbon cable or power button affecting our results, we can turn the Switch on with nothing more than a pair of tweezers.

Short out the top 2 pins of the power/volume ribbon cable to simulate pressing the power button.

Once you do this you should observe the bench power supply current draw increase.

The current will depend on the current operational state of the console, and if you have an LCD connected or not. We will base lots of our diagnostics on these specific current draw values.

For example a working console without an LCD connected would typically draw 200mA at first-stage boot, and jump to 400-600mA during second-stage full boot.

Here is a table of the power draws at the stages of booting. Starting with first-stage boot, then second, then sleep.

As we find more specific cases of failed components indicated by specific current draws we will update this table.

A fully working boot will power up at 100mA, jump to 150mA, stay a little while, then jump to 500-800mA (with an LCD) or 400-500mA (without an LCD).

After a short while the console sleeps and it goes to 8mA (without an LCD) or to 340mA then 13mA (with an LCD).

When you turn on (150mA), see the Nintendo Logo (375mA), then black screen and stays stuck at 100mA, it is typically a corrupt eMMC specifically the boot flags.

A great trick is to install a PicoFly / HwFly and boot into the hacked Atmosphere. This fixes the boot flags. Then remove the mod and it should boot perfectly fine.

If you have a dead short, you will see excessive current draw, the inductor on BQ usually overheating (as all the system current passes through it).

If your console is fully booting, but the backlight is bad, the current draw will act like full boot, but be slightly less.

So it will rise to 100mA, then 150mA, stay a little while, then jump to 400-500mA. After a while it will got to sleep at 8mA.

If this is the case, and you don't see anything on screen, you likely have a bad backlight or bad backlight driver.

A bad P1USB can stop boot at second stage if failed open, or stop boot at first stage if shorted.

A common symptom of a damaged USB-C port is to send the 15V into the P1USB killing it.

When the P1USB is shorted, it will pull the 3.3V rail fully shorted, sinking around 300mA when you turn on, and no other movement on the current.

This is easy to check by measuring if the 3.3V rail is lower or 0V when this happens.

If this is the case, inject 1V into the 3.3V rail and check for shorts and overheating to find the component that has failed.

If the fuel gauge is bad, damaged or resistors are missing on it, you will get initial power up and then stopping pretty fast at 100mA.

Sometimes it will boot up to 160mA then drop to around 120mA and stay.

The general rule is if it boots in under 2 seconds to around 100-120mA, its usually the fuel gauge.

Very similar to the fuel gauge, a bad MAX77621 will often boot straight to 100mA and stay.

If your battery, bench power or traces for power supply are limiting the current capability of the system, you will often get 0mA then current draw of a few hundred milliamp, then back to 0mA and a constant boot loop.

A bad M92T can cause multiple issues.

The first is if your USB-C charger only works one way. This can be the USB-C port also, but it can be the M92T.

The second cause can be booting to 200mA, then going to second-stage boot drawing 400-800mA, and then either constantly bouncing between 400-800mA or boot looping back to 200mA.

If your power draw goes to around 180-200mA but the device never then enters second-stage boot, it is usually a bad eMMC.

Check for clock and data activity on the eMMC to confirm and check the eMMC Booting section for more detail.

Often it seems once there has been a problem, the bits in the memory corrupt and force AutoRCM. The symptoms of this are usually a straight jump to 200-240mA and no further and no screen.

Hack the Switch or use Tegra and RCM inject Hekate, then go to Tools > AutoRCM and disable it if auto-enabled.