Modules 14 & 15 - RAM and Static Storage Elements, Dynamic Storage Elements

8.1.2: A 256-Mb Memory Chip

When a row is selected, all 127 bits are all accessed in parallel for a word. These are wordlines and the data in vertical columns are called bitlines. The column address can extract each specific bit if needed.

8.4: Sense Amplifiers

8.7:

8.3: Dynamic Memory Cells

Power applied to static memory is required to refresh their memory/data from losing charge. But smaller memory cells can also be built, known as dynamic memory.

8.3.1: The One-Transistor Cell

This is a 1-T cell:

Pasted image 20240315215400.png

We have a Cc cell capacitor that stores the data. Leakage currents exist in the drain-bulk and source-bulk junctions of the transistor, and in the transistor channel. Periodically, cell data is written back to keep the desired cell voltages. This is the refresh operation. Each storage cell in a DRAM usually must be refreshed every 2 to 10 ms.

8.3.2: Data Storage in the 1-T Cell

A 0 is represented by 0V and a 1 is represented by a high level on Cc. Writing occurs by putting the desired voltage level onto the bitline and turning on access to the MA transistor.

Storing a 0

In this case, the bitline is a 0, and the bitline terminal of the MOSFET acts as the source of the MOSFET. The gate is raised to VDD=3V. If the cell voltage is already zero, then ,the drain-source voltage is 0, so the current is 0. If the cell has a 1 with vc>0 then the MOSFET completely discharges Cc to give a 0.

Pasted image 20240315215812.png

Storing a 1

In this case, the bitline is a 1, and the bitline terminal acts as a drain. If the gate gets raised to VDD then:

Because VDS=VGS and since MA is an enhancement mode device, MA will operate in saturation.

Pasted image 20240315220155.png

Notice that:

Vc=VGVtn=VG(VTO+γ(Vc+2ϕF2ϕF))

Notice here body effect is good since it reduces Vc and hence reduces power draw of the 1-T.

8.3.3: Reading Data from the 1-T Cell

We first precharge the bitline to a known voltage, usually VDD or half of it. The access transistor G is turned on, and charge sharing occurs between CBL and Cc.

Pasted image 20240315220535.png

We get that:

QI=CBLVBL+CcVc

after closing, the current equalizes the voltage across the two capacitors (no current will eventually flow, so no voltage drop over the MOSFET). Hence:

QF=(CBL+Cc)VF

Thus:

VF=CBLVBL+CcVcCBL+Cc

The signal to be detected the change in the voltage on the bitline, so then:

ΔV=VFVBL=CcCBL+Cc(VCVBL)=(VCVBL)CBLCc+1

Typically we have CBL>>Cc, since we want a large enough ΔV to detect, implying that:

VFVBL

so when we read (closing the switch), we'll accidentally destroy the data we had on there. Hence, we must write-back any data we read. The sense amplifier helps read the voltage change ΔV.

8.4.2: A Sense Amplifier for the 1-T Cell

See the figure below for a simple sense-amplifier:

Pasted image 20240315223634.png

The stages are: