GLK12232-25 displays repeatedly break
Posted: Mon Jan 27, 2003 12:26 pm
We are experiencing some very odd behaviour on a large number of GLK12232-25 which we are using as status displays in our network equipment.
When powering up our unit, the GLK12232 displays the startup bitmap. When our firmware has been loaded, a "clear screen" command is sent to the display, followed by a number of "draw line" commands to draw a simple "background" for the status output.
Then, the firmware continously draws status info on the display using "set pixel" commands. The redraw rate is approximately 2 times per second.
This works excellent for most of the displays, but for 10% of the units (5 out of 50 so far), the drawing fails. In these occasions, there are two different (although similar) problems:
1. On some of the faulty displays, there is a rectangular area of the display (often the leftmost part) that is invalid, i.e. it is impossible to draw pixels in it. The rest of the display gets updated correctly, but the invalid area stays blank. The initial startup bitmap, however, is displayed correctly.
2. On some of the displays, some of the commands sent to the display simply don't work. In the process described above, the first "clear screen" command and the initial "draw lines" do not work, but the status updates (set pixel commands) work as expected. Also in this case, the initial startup bitmap is displayed correctly, but as the clear screen and draw line commands fail, the result is a startup bitmap being garbled with set pixel commands. Please note that it is the same commands that fail to work, even if the unit is restarted.
It could be worth notice that the devices could be powered off at "any time", meaning that there is a possibility that only partial command sequences are being sent to the display before losing power.
Disconnecting the power and have the display rest for a large number of hours can sometimes cause it start working as expected again. This method only work for some of the faulty displays however, and if the error occurs at customer site, we can not advise our customers to shut down their equipment.
We are very satisfied with the displays so far, but as this has happened to quite a large number of displays, we don't know if we see a software issue here or (worse) a display quality problem.
Please advise!
/John
When powering up our unit, the GLK12232 displays the startup bitmap. When our firmware has been loaded, a "clear screen" command is sent to the display, followed by a number of "draw line" commands to draw a simple "background" for the status output.
Then, the firmware continously draws status info on the display using "set pixel" commands. The redraw rate is approximately 2 times per second.
This works excellent for most of the displays, but for 10% of the units (5 out of 50 so far), the drawing fails. In these occasions, there are two different (although similar) problems:
1. On some of the faulty displays, there is a rectangular area of the display (often the leftmost part) that is invalid, i.e. it is impossible to draw pixels in it. The rest of the display gets updated correctly, but the invalid area stays blank. The initial startup bitmap, however, is displayed correctly.
2. On some of the displays, some of the commands sent to the display simply don't work. In the process described above, the first "clear screen" command and the initial "draw lines" do not work, but the status updates (set pixel commands) work as expected. Also in this case, the initial startup bitmap is displayed correctly, but as the clear screen and draw line commands fail, the result is a startup bitmap being garbled with set pixel commands. Please note that it is the same commands that fail to work, even if the unit is restarted.
It could be worth notice that the devices could be powered off at "any time", meaning that there is a possibility that only partial command sequences are being sent to the display before losing power.
Disconnecting the power and have the display rest for a large number of hours can sometimes cause it start working as expected again. This method only work for some of the faulty displays however, and if the error occurs at customer site, we can not advise our customers to shut down their equipment.
We are very satisfied with the displays so far, but as this has happened to quite a large number of displays, we don't know if we see a software issue here or (worse) a display quality problem.
Please advise!
/John
