If this erasure of bits is a truly random process, I would expect the number of 0’s to drop with a factor 4.37 every 25 minutes. I put the EPROM under the UV LED initially for 25 minutes. I counted the number of 0’s initally it was 7651 (the other 57885 bits were 1’s). A measure of the progress of erasing is the number of 0’s still left in the EPROM. I presume this is sort of a random process, a photon just has to hit the bit at the right angle to clear it or so.
Erasing an EPROM with a LEDĪs an EPROM is erased, all the bits are reset to 1’s. So, exposure time should be not 20 minutes, but 200 minutes or just a little over 3 hours. But what counts is intensity × exposure time, says the datasheet. That is ten times as low as the example in the EPROM datasheet. This means the wavelength is good and if I manage to get 25% of the light coming from the LED to fall on 1 cm ² of the EPROM, the intensity will be 1.25mw/cm ². LED (Optosupply OSV4YL5451B): Optical characteristics: Peak Wavelength: typ. 20 minutes using a UV lamp of 12mW/cm² power rating. UV intensity × exposure time) should be ≥ 15 Ws/cm². But then I checked the datasheets.ĮPROM: Erasure of the μPD2764 programmed data can be attained when exposed to light with wavelengths shorter than approximately 4000Å. At first I thought: it can’t be true that a tiny 5 mW LED can do what a giant 300W tanning lamp can’t. Then I recalled that I had once ordered a collection of LEDs in different colours including a few infrared and ultraviolet LEDs. So I continued outside of the house and left the EPROM 2 cm under the tanning lamp for 6 hours and checked the EPROM again. After four minutes there was an alarming smell, so I had to turn off the lamp.
The first attempt almost burnt down the house, as the tanning lamp generates lots of heat. I assumed it would be good for erasing EPROMs as well. I used it back in the day to develop the photosensitive layer of pcb’s. On the attic I found a 300 watt Philips Ultraviolet tanning lamp, that my mother gave to me thirty years ago. The second problem: the 2764 was not empty and needed to be UV erased. I imagined I could do something similar with the 2716/2764, although my EPROM has 4 times less memory than required, but it could still contain 1 ASCII character set.īut then my friendly colleague Bart came in and said: are you looking for this? And he handed me a second-hand 2764. In 1981 I changed a few pcb traces on the Junior and it worked as a treat. I remember the Junior required a 2708, but by that time the 2708 was already obsolete and the 2716 was more or less pin compatible. I bought this in 1981 or so for the Elektuur Junior computer. I did have a 2716 (16 kilobit), new old stock. The EPROM required was a 2764, this is a 64 kilobit (8kiB) eraseable programmable read only memory, which is eraseable by ultraviolet light. The EPROm should contain a character font that goes into the LCD so it can show characters. For a project (revitalizing a large 80’s liquid crystal display) I needed to program an EPROM.
0 kommentar(er)
