SNES cartridge power consumption

Published 2019-01-13

So I made a device for measuring the power consumption of SNES cartridges. As one does.

I decided to take a look at the power consumption of a few games.

Here’s the setup:

The display can show current (A), voltage (V) or power (W). This game currently draws 0.03 W, or 30 mW.

On the lower right you can see a dummy cartridge, containing a 9.8 Ohms resistor, used to verify the current readings.

How it works

The power meter has a current sensing resistor in-line with the 5V rail, generating a voltage difference proportional to the current drawn. This voltage is then amplified, read by a microcontroller and then displayed.

The microcontroller also senses the actual voltage of the 5V rail, allowing us to make sure that we have a good voltage and also calculate the power consumption.

Results

Super Mario World

6 mA / 30 mW

This iconic game draws a measly 30 mW. The cartridge contains a read-only memory (ROM) and a few support chips.

Super Mario Kart

16 mA / 80 mW

Some of the power consumption of this game can be attributed to the DSP-I enhancement chip that allows the game to perform real-time math calculations that the SNES processor would not be capable of performing.

Yoshi’s Island

78 mA / 390 mW

This games draws a lot of power, due to its embedded Super FX2 enhancement chip. This chip allows the game to render advanced special effects for its time.

Gradius III

6 mA / 30 mW

I got hold of this game on a recent trip to Japan. To my surprise, it’s in English..!

Now, if you’re wondering, no, my console has not been hacked. Normally, a Japanese game shouldn’t run on a European console. They use different region lockout chips. However, I’ve included a selectable european lockout chip on the power meter, allowing me to test imported games!

Vortex

104 mA / 520 mW

This is the winner in terms of power consumption. It’s a 3D shooter game, powered by the Super FX chip.

It does consume more than Yoshi’s Island, even with the same family of enhancement chip. I would attribute the difference to a more intense usage of the chip. There could also be a difference in power consumption between the two versions, this one being the older one.

Bonus

I noticed something weird when using my import cart adapter. Take a look at this:

It’s barely visible on the photo, but power the meter shows a supply voltage of 4.29 V. And this is with no game inserted!

The import cartridge succeeds in dropping the supply voltage by 700 mV. Putting the power meter before the import cart adapter yields 5 V, indicating that the adapter does not stress the 5 V rail of the Super Nintendo.

With a game inserted, the supply voltage is at 4.27 V, about the same.

This could be some kind of diode protection. It would be interesting to do a teardown of this adapter to see what’s really going on!



Feedback

If you have any questions, a bit of feedback, or just want to leave a comment, please enter it here and it will be sent to me.

I read all feedback. Include an email if you would like me to respond.