Microchip 24LH08X


This Microchip 24LH08X memory chip was on the Meridian handset and dates from around 1993.

 

 

 

I was unable to find a datasheet or any reference to a 24LH08X. I am pretty sure it is an 8kbit serial EEPROM. Normally looking at memory chips with an optical microscope is a waste of time as they are so dense.  However I thought a 1993 chip might be old enough that it was worth a look.

.

The die marking confirms it is Microchip and designed in 1990. The S7122 is no help in confirming the parts function.

 

 

Die Photo

Here is the (not very good die) photo of the 4.78 x 2.16 mm (10.3mm2) dieMicrochip 24LH08X

It is not very good as the array has misaligned tiles.  To get high resolution die photos I take multiple images with the microscope and stitch them together.  To do the stitching I use a program called Hugin, which although a bit clunky, and can take some time, generally does a good job.  However when you have a very repeating array it struggles to find unique features to align images, so making stitched die images of memory arrays is very difficult.  I spent way too long trying to get a good die photo, in the end I decided to accept this as the best I will do.

Looking at the die photo, the first thing that struck me was how the memory array only takes up just over one third of the die area.  That seems really small.  Packaged in a 14 pin DIP yet  the part only has 8 bond pads.

The read/write logic area on the left is pretty low density, even for a 1990 design. The part looks like it is using 3μm minimum geometries, with single metal and of course two poly layers to form the floating gate structures.

Memory Array

Looking at the array corner with 20x objective.

Zoomed in with the 80x objective (Focus Stacked image.) Even with this geometry the memory array is so dense not a lot of structure can be picked out with the optical microscope.Here is a nice image of the row decoder area leading to the logic cells.

One mystery on the device is the two rows of poly-poly capacitors at the right of the array.  I assumed they were for a charge pump for device programming/erasure. To program an EEPROM (or Flash memory) a high voltage is required. A circuit called a charge pump that basically stacks capacitor charge to boost the input voltage 2X or 3X is used.  What is the mystery is that this is a 1990 design part, and I don’t think charge pumps were being used in EEPROM until mid 90’s.

 

 

This entry was posted in Die Photo and tagged , , . Bookmark the permalink.

2 Responses to Microchip 24LH08X

  1. Jeremy says:

    Might be related to 24LC08? The logic on the left might be to handle I2C.

    http://www.microchip.com/wwwproducts/en/24LC08B

Comments are closed.