Flash memory explored

Imagine a mobile phone that lost all its numbers when you took out the battery. Or a digital camera that you had to keep charged, otherwise your holiday snaps vanished. Both would be pretty useless to most of us.

While we might expect a really old radio to lose the stations we’ve stored if it’s unplugged, we don’t expect it of a new one. Whether we remember to charge it up or not, we assume our MP3 player will have all our tunes, just as soon as it’s powered up.

For all these things, we rely on non-volatile memory, which carries on holding data when the power is removed. In most modern devices, that means Flash memory ­ compact, cheap, and found inside just about everything electronic. It’s even supplanting the hard disk, with products such as Apple’s Macbook Air available in an entirely Flash-based version, and the Asus Eee PC relying on it too.

So just what is Flash memory, how does it work, and is there really a difference between the brands and types that, increasingly, you can even pick up at the supermarket with your groceries? In this feature, we’ll explain all you need to know.

How Flash works
If all you want to know is what’s different between one brand of Flash and another, skip ahead to the section on buying Flash. But if you want to know the technical details, read on. Flash memory works by storing tiny electrical charges, representing individual binary bits. These charges can be retained for many years by the right type of materials.

Flash memory uses a specially configured transistor (technically it’s a metal-oxide-semiconductor field-effect-transistor or Mosfet) to store each of these charges. A ‘normal’ Mosfet has three connections, referred to as Source, Drain and Gate. The ‘channel’ between the source and the drain conducts current under control of the voltage present on the gate terminal. A Flash Mosfet has an extra gate layer underneath the main gate terminal. This is a ‘floating’ gate, so called because it’s surrounded by an insulating layer. When the floating gate becomes charged, it partially screens the effect of the main control gate.

The memory cell is read by placing a voltage on the control gate, and the presence or absence of a charge on the floating gate, affects the current flowing between the source and the drain, revealing the information stored.

The floating gate is charged by applying a higher voltage to the control gate, causing current to flow between the source and the drain, which induces some electrons to charge the floating gate via a process known as hot-electron injection.

Cells are erased by applying a large reverse polarity voltage pulse between the control gate and the drain, which pulls electrons off the floating gate in a process called quantum tunnelling.

On-chip charge pumps are used to generate the voltages required for programming and erasing the memory cells, enabling Flash memory chips to run off a single low-voltage supply rail.