Hi Hagen, hi list Cedric and I just talked about the Pearl River test wafer, and I had an inspirational flash: We totally should make it possible to hook up LEDs to the output pads of the Pearl River on a PCB, so that we can demonstrate that the pad cell drivers have the capability of driving an LED! At the SEG plaza in Shenzhen there are many vendors with demo circuits showing the color and brightness of the LEDs in action. We could also make such a demo board in order to demonstrate the multiple different capabilities of the LibreSilicon process. Would do you think?
Cheers David
Hello David.
Well, to connect a LED to a output pad *outside* the chip is quite easy. The pad should able to drive a current of round about 10..20 mA, with voltage higher than the LED threshold voltage.
I already thought about an Open Drain output driver, which is quite useful for driving LEDs even with lower supply voltages than the LED threshold. Another topic is a bipolar output driver (aka BiCMOS) https://en.wikipedia.org/wiki/BiCMOS with higher current capabilities.
The other think, we *can not* do, is implementing a LED *on chip*. The brightness of different colors of LEDs resulting from different materials, see https://en.wikipedia.org/wiki/Light-emitting_diode#Colors_and_materials
And yes, all this thoughts belongs to the documentation / wiki ;-)
Regards, Hagen.
On 07/18/2018 11:57 AM, David Lanzendörfer wrote:
Hi Hagen, hi list Cedric and I just talked about the Pearl River test wafer, and I had an inspirational flash: We totally should make it possible to hook up LEDs to the output pads of the Pearl River on a PCB, so that we can demonstrate that the pad cell drivers have the capability of driving an LED! At the SEG plaza in Shenzhen there are many vendors with demo circuits showing the color and brightness of the LEDs in action. We could also make such a demo board in order to demonstrate the multiple different capabilities of the LibreSilicon process. Would do you think?
Cheers David
Hello Hagen
Well, to connect a LED to a output pad *outside* the chip is quite easy. The pad should able to drive a current of round about 10..20 mA, with voltage higher than the LED threshold voltage.
Of course I ment outside! On a PCB! The idea would be that we flip chip mount our Pearl River onto a PCB and solder some LEDs onto the PCB, so that we can demonstrate that it blinks. Maybe a ring-counter or so?
I already thought about an Open Drain output driver, which is quite useful for driving LEDs even with lower supply voltages than the LED threshold. Another topic is a bipolar output driver (aka BiCMOS) https://en.wikipedia.org/wiki/BiCMOS with higher current capabilities.
The other think, we *can not* do, is implementing a LED *on chip*. The brightness of different colors of LEDs resulting from different materials, see https://en.wikipedia.org/wiki/Light-emitting_diode#Colors_and_materials
And yes, all this thoughts belongs to the documentation / wiki
Ok! I will put it there! I'm totally aware that we can't build LEDs just yet. For that we would have to drastically modify the recent process. Already implementing the flash cells will be challenge on its own, because it will require another nitride layer deposition step after the gate oxide deposition step in most simplistic approach case.
Cheers David
libresilicon-developers@list.libresilicon.com
-
David Lanzendörfer -
Hagen SANKOWSKI