2018-08-20, 23:42
Hi Dr.Feelgood, I hope you had a pleasant weekend and thanks for the reply. Since it would appear that I have few options I may push ahead with the 1.5+1.6 bios updates to see if there is any improvement with the problem I documented. Perhaps I would feel more comfortable installing the updates if I knew why ASRock first changed the bios installation procedure to include only Instant Flash, to prevent the bricking, and then a few days later ASRock completely removed the updates from general distribution.
Since that day in March ASRock has released 40+ bios updates for other products, but has failed to release a revised update for the J4205. This makes me suspicious that other problems were discovered and remain unsolved. If I do decide to install these updates I hope they won't lead to new anomalies where I spend months trying to understand what caused them.
If it isn't apparent, I tend to be apprehensive when software and firmware updates are released so I will usually wait on the sidelines for a week or two until I know other people have successfully installed them. I will monitor support forums searching for reports of any problems caused by the updates, which is how I avoided bricking my J4205 with bios 1.5+1.6. I see you have the 1.5+1.6 battle scars, one of those I read about that had to resort to an external eeprom programmer to revive a bricked J4205. As an update pacifist I tip my hat to you for your gallant effort.
I did install LSPCON firmware 1.66 some time ago. Based on posts made to this thread, specifically page 117, I got the impression that it was best to stay with version 1.66 and not install 1.71. I read that some Kodi forum members had problems with the J4205 and LSPCON firmware 1.71, and had a desire to switch back to 1.66. Intel had removed the link to version 1.66 so member fritsch made a dropbox link available where 1.66 could be downloaded. That is how I was able to install LSPCON 1.66. I never bothered to install any other LSPCON update beyond 1.66 since it did nothing to solve the problem I was experiencing.
I would like to emphasize that I viewed the Intel LSPCON firmware updates as not being officially supported by ASRock for their Apollo Lake boards. Using these updates seemed to be merely a backchannel attempt by a few tech savvy Kodi users to compensate for the lack of official support from ASRock for the various HDMI problems some Kodi members were experiencing.
I know if I purchased a new car and discovered the digital gauges were blank each time I started the car, I would expect the manufacturer to acknowledge the problem and issue a recall to fix it. I would not expect that I would need to resort to an aftermarket solution provided by AutoZone. Now if the digital gauges were blank because I accidently dimmed the backlight down to nothing, or if a fuse blew, I hope that I would have the common sense to check for obvious causes before contacting the manufacturer, but I still believe the manufacturer should respond to the problem if I did report it. Just sayin'
"If your Scenario C works i don't see what is the problem anymore?"
As I mentioned in my original post, and as demonstrated using Procedures "A" and "B", the blank screen problem happens when Windows 10 restarts while the HDMI connected LCD display is powered-off. I built this fanless low power computer so I could leave it powered-on 24/7/365 to share files. It also functions as a general purpose computer that I use from time to time. As you may or may not be aware, Windows 10 has a tendency to automatically restart after installing patches or for other maintenance tasks. I have Windows 10 configured to perform these automatic restarts during the early morning hours, when the computer is idle, and the LCD display is powered-off, but I never know when these restarts will occur.
Many times I would power-on the LCD display only to discover a blank screen with a flashing mouse pointer. I never could determine why this was happening and, once it did happen, additional restarts would still leave me with a blank LCD display and a flashing mouse pointer. So I was left to ponder, was this a problem with Windows? Was it a problem with my LCD display or HDMI cable? Was it a problem with the motherboard hardware? Was it an Intel video driver problem perhaps?
Another frustrating new feature of Windows 10 is that the old reliable F8 key access to safe mode has been neutered making this mode of operation nearly impossible to access for the purpose of troubleshooting. From what I read, entering Windows 10 safe mode must now be done from within the Windows desktop environment or, the safe mode option will be presented as a start-up selection only after three Windows, hardware lock-up induced, restarts. Obviously a LCD blank screen prevents access to the Windows desktop, and apparently restarting from a blank screen caused by an Intel video driver "recover" is not considered a hardware lock-up by Windows, so accessing Windows 10 safe mode was quite difficult to do.
In my original post I mentioned that I eventually found the Windows event log had entries that showed there was an issue with the Intel video driver, but I wondered if the driver was truly the cause or just a symptom of some other problem. I tried installing different versions of Intel video drivers without any improvement. Several times I did fresh installations of Windows 10, as well as installing different versions of Windows 10, still not a solution. As ASRock released a few bios updates I would install those, but the problem persisted. I submitted the tech support request to ASRock where I described the problem, but I received no response. Eventually I found threads on the ASRock and Kodi forums where I noticed other people were experiencing problems with the J4205 HDMI port and having varying degrees of success at solving them.
So almost two years since I purchased the board, and after a lot of trial and error, I have finally discovered this unexplained quirk with the bios CSM setting that has a direct impact on my HDMI blank display problem. It would be nice to know if this problem could be fixed, or at least have some logical explanation why the CSM bios setting automatically changes from Disabled to Enabled during a Windows restart whenever the LCD display is powered-off.
Over the years I have probably built a dozen computers from the ground up. Some were gaming computers while others were just general purpose computers that had special requirements. I too have experienced the occasional unexplained anomaly that was often solved with a bios or driver update. I have also sent the occasional tech support request to other Taiwanese companies and always received a prompt and courteous response that either acknowledged or offered a solution to the problem. My experience with ASRock has certainly been the unfortunate outlier.
As you implied, maybe this is the end of the road where I either choose to live with this problem or explore other motherboard options. Of course the Windows 10 license hardware lock-in doesn't make the switch to a new board an easy solution to accept, but that is a story best left for another lenghty post.
Since that day in March ASRock has released 40+ bios updates for other products, but has failed to release a revised update for the J4205. This makes me suspicious that other problems were discovered and remain unsolved. If I do decide to install these updates I hope they won't lead to new anomalies where I spend months trying to understand what caused them.
If it isn't apparent, I tend to be apprehensive when software and firmware updates are released so I will usually wait on the sidelines for a week or two until I know other people have successfully installed them. I will monitor support forums searching for reports of any problems caused by the updates, which is how I avoided bricking my J4205 with bios 1.5+1.6. I see you have the 1.5+1.6 battle scars, one of those I read about that had to resort to an external eeprom programmer to revive a bricked J4205. As an update pacifist I tip my hat to you for your gallant effort.
I did install LSPCON firmware 1.66 some time ago. Based on posts made to this thread, specifically page 117, I got the impression that it was best to stay with version 1.66 and not install 1.71. I read that some Kodi forum members had problems with the J4205 and LSPCON firmware 1.71, and had a desire to switch back to 1.66. Intel had removed the link to version 1.66 so member fritsch made a dropbox link available where 1.66 could be downloaded. That is how I was able to install LSPCON 1.66. I never bothered to install any other LSPCON update beyond 1.66 since it did nothing to solve the problem I was experiencing.
I would like to emphasize that I viewed the Intel LSPCON firmware updates as not being officially supported by ASRock for their Apollo Lake boards. Using these updates seemed to be merely a backchannel attempt by a few tech savvy Kodi users to compensate for the lack of official support from ASRock for the various HDMI problems some Kodi members were experiencing.
I know if I purchased a new car and discovered the digital gauges were blank each time I started the car, I would expect the manufacturer to acknowledge the problem and issue a recall to fix it. I would not expect that I would need to resort to an aftermarket solution provided by AutoZone. Now if the digital gauges were blank because I accidently dimmed the backlight down to nothing, or if a fuse blew, I hope that I would have the common sense to check for obvious causes before contacting the manufacturer, but I still believe the manufacturer should respond to the problem if I did report it. Just sayin'
"If your Scenario C works i don't see what is the problem anymore?"
As I mentioned in my original post, and as demonstrated using Procedures "A" and "B", the blank screen problem happens when Windows 10 restarts while the HDMI connected LCD display is powered-off. I built this fanless low power computer so I could leave it powered-on 24/7/365 to share files. It also functions as a general purpose computer that I use from time to time. As you may or may not be aware, Windows 10 has a tendency to automatically restart after installing patches or for other maintenance tasks. I have Windows 10 configured to perform these automatic restarts during the early morning hours, when the computer is idle, and the LCD display is powered-off, but I never know when these restarts will occur.
Many times I would power-on the LCD display only to discover a blank screen with a flashing mouse pointer. I never could determine why this was happening and, once it did happen, additional restarts would still leave me with a blank LCD display and a flashing mouse pointer. So I was left to ponder, was this a problem with Windows? Was it a problem with my LCD display or HDMI cable? Was it a problem with the motherboard hardware? Was it an Intel video driver problem perhaps?
Another frustrating new feature of Windows 10 is that the old reliable F8 key access to safe mode has been neutered making this mode of operation nearly impossible to access for the purpose of troubleshooting. From what I read, entering Windows 10 safe mode must now be done from within the Windows desktop environment or, the safe mode option will be presented as a start-up selection only after three Windows, hardware lock-up induced, restarts. Obviously a LCD blank screen prevents access to the Windows desktop, and apparently restarting from a blank screen caused by an Intel video driver "recover" is not considered a hardware lock-up by Windows, so accessing Windows 10 safe mode was quite difficult to do.
In my original post I mentioned that I eventually found the Windows event log had entries that showed there was an issue with the Intel video driver, but I wondered if the driver was truly the cause or just a symptom of some other problem. I tried installing different versions of Intel video drivers without any improvement. Several times I did fresh installations of Windows 10, as well as installing different versions of Windows 10, still not a solution. As ASRock released a few bios updates I would install those, but the problem persisted. I submitted the tech support request to ASRock where I described the problem, but I received no response. Eventually I found threads on the ASRock and Kodi forums where I noticed other people were experiencing problems with the J4205 HDMI port and having varying degrees of success at solving them.
So almost two years since I purchased the board, and after a lot of trial and error, I have finally discovered this unexplained quirk with the bios CSM setting that has a direct impact on my HDMI blank display problem. It would be nice to know if this problem could be fixed, or at least have some logical explanation why the CSM bios setting automatically changes from Disabled to Enabled during a Windows restart whenever the LCD display is powered-off.
Over the years I have probably built a dozen computers from the ground up. Some were gaming computers while others were just general purpose computers that had special requirements. I too have experienced the occasional unexplained anomaly that was often solved with a bios or driver update. I have also sent the occasional tech support request to other Taiwanese companies and always received a prompt and courteous response that either acknowledged or offered a solution to the problem. My experience with ASRock has certainly been the unfortunate outlier.
As you implied, maybe this is the end of the road where I either choose to live with this problem or explore other motherboard options. Of course the Windows 10 license hardware lock-in doesn't make the switch to a new board an easy solution to accept, but that is a story best left for another lenghty post.
