Marmite in electrical circuits

25-09-2023

Prelude

Following material-replication fiascos that killed off the LK-99 buzz, one cannot help but wonder whether or not there are superconducting materials left out there right under our noses. Could Marmite be a suitable candidate? Not in anyone's right mind, but that did not stop me from checking it out after someone pointed out Marmite could actually work as a half-decent conductor.

Due to its salt content, Marmite is most likely going to perform sufficiently well, due to the presence of sodium and chlorine ions. I've set out to not only understand whether or not you can power a LED through Marmite, but understand some of its resistance characteristics. This is not a serious attempt at science, and this should not be taken into consideration when the thought of sectioning me arises.

Marmite as a conductor

As part of a preliminary test, I've used a standard through-hole 20 mA LED, powered from a 5V power supply, with no current limiting resistor, placed in such a way that the anode is connected to the power source through a marmite track (roughly 2 cm), whilst the cathode is directly connected to ground.

For completeness' sake, Marmite was spread on a slice of Hovis wholemeal seeded bread. There is no scientific evidence pointing to Tesco Deluxe or Lidl's own brand performing better. Although the LED seems bright, I've had to mess around with camera angles and brightness to get it to show up like that. It is much dimmer in person.

This, therefore, makes it clear that Marmite is a sufficiently adequate conductor for extremely low powered applications, given the tracks are short. As I moved the LED anode further away from the voltage source, it got progressively dimmer, until it was essentially undistinguishable from an unpowered LED, at 6 cm from +5V.

Resistance characteristics

As part of this mostly useless test, I've built a voltage divider utilising a 100 ohm resistor, and a Marmite track, in which I gradually increased the distance between +3.3V and ground.

To measure voltage, I've used a Raspberry Pico, which has a 12 bit ADC capable of giving me a resolution of 0.8 mV.

This resulted in the following measurements:

When plotted, this gives us the following curve:

Now, does this tell us anything interesting? Does this reveal any secrets? No, it just confirms our suspicion that Marmite, although "serviceable", is a remarkably poor conductor and should not be used in any serious applications. Unless you're in a pinch.

You may notice resistance scales up linearly, but not in the tidiest manner. This is not any sort of interesting behaviour, but rather a limitation of the tools I had available (a Pico and a ruler, really) and reproducibility issues (I did not go out and purchase a syringe to make consistent lines, nor did I account for bread saturation between measurements).

Conclusion

Marmite, as a matter of fact, works surprisingly well. I have not tested data transfer utilising Marmite tracks, nor do I plan to in the near future, however, I can imagine some "robust" differential protocols such as CAN might be able to work with yeast extract instead of copper.

As a matter of fact, CAN buses in one of my jobs kept working even after someone inverted polarity and forgot to account for ground on a circuit, so I'm not doubting it'll work.

Am I willing to test with other condiments, spreads and kitchen ingredients? Maybe, but I'm not keen on going through all new flavours of Marmite and figuring out which one works best, unless someone is willing to support me in my grant proposal efforts.