| Author |
Topic |
|
|
pacsciinst
15 Posts |
 Posted - 03 Aug 2005 : 17:49:40
|
Hello,
I noticed that in version 4.2.17 of the CE image for BitsyX, the LCD panel power and LCD backlight on/off and backlight pwm do not function the same as version 4.2.14. In 4.2.14, when entering sleep mode, the panel power is shut off to reduce sleep mode power consumption.
Also in version 4.2.14, the backlight on/off signal is activated automatically when setting the backlight timeout value in the system registry. However, in V4.2.17, it does not shut off but rather dims a little and flickers.
I am using the same registry (ADSLOAD.REG) and hardware (ADSLOAD.HWT) files in my system, therefore, I believe the problem lies in the OS or in an OS driver. I have not tried V 4.2.16 to see if the same anomaly exists.
Please check into this problem and advise.
Thanks.
|
Edited by - pacsciinst on 03 Aug 2005 17:50:48 |
|
|
akidder
1519 Posts |
Posted - 03 Aug 2005 : 18:45:43
|
As you might have noticed in the revision history, there were a lot of changes in the BitsyX CE builds between 4.20.14 and .17.
I don't see any notes about this change in the build history, but we'll take a look at the drivers to see what's going on and let you know what we find. |
 |
|
|
pacsciinst
15 Posts |
Posted - 08 Aug 2005 : 18:15:42
|
Hello:
Just checking to see if any discoveries have been made.
Thanks |
|
|
|
akidder
1519 Posts |
Posted - 09 Aug 2005 : 11:44:16
|
Thanks for the follow-up. We ran a test yesterday and didn't see this issue.
Can you remind us what backlight inverter you're using? Does it have different polarity from the LS520 of our dev systems? That might be why we're not seeing it here.
p.s. We emailed you this follow-up via the forums yesterday, but it looks like your email address is out of date. You might update your member profile with your current address to reduce delays in the future. |
|
|
|
pacsciinst
15 Posts |
Posted - 09 Aug 2005 : 12:24:35
|
| I guess I'm mainly interested in why my system works fine with 4.2.14 and not 4.2.17. The OS is the only variable in my system that causes the above-mentioned problem. The polarity of the inverter seems irrelevant in this case. |
|
|
|
akidder
1519 Posts |
Posted - 09 Aug 2005 : 15:31:44
|
Thanks. It may be that the new backlight driver is included in build 4.20.17 (see topic 1902 for its spec). I'll check on that.
The spec was created to be able to support different types of backlight inverters without recompiling the build or ADSmartIO code. It may be that the new build is not handling your backlight control correctly during sleep.
Without knowing what kind of backlight inverter you're using, it's not clear what next step we should take to help you resolve this issue. |
|
|
|
pacsciinst
15 Posts |
Posted - 09 Aug 2005 : 17:05:54
|
The reason I say the backlight is irrelevant is our backlight converter uses opposite logic for enable/disable than that of the ADS development systems, however, the "personality" card we developed inverts the signal so we can use the signal (/BACKLIGHT_ON) from the BitsyX directly. Our card uses a 10K pullup on the OC output of this pin.
We noticed the BACKLIGHT_PWM signal is inconsistent with the different OSs. The /BACKLIGHT_ON stays low (enabled) with V4.2.17. We also observed that the /PNL_PWR pin on the LCD interface stays at 3.3V during sleep mode. In V4.2.14, it does not. Also, I just confirmed that V4.2.16.13 also fails to control these signals too.
If you still want the specs on our backlight inverter, I can send you the datasheets. I updated my profile to include my latest email address.
Thanks |
|
|
|
akidder
1519 Posts |
Posted - 09 Aug 2005 : 17:16:46
|
Thanks. It does sound like the specific inverter isn't the issue.
I checked with our team, and I think the change in control is due to inclusion of the v 1.0 backlight driver spec in the build (again, see topic 1902). I'm looking into the build into which it was first added, and I'll update the revision history accordingly. Our team will also take a look at why it's not behaving as expected. |
|
|
|
pacsciinst
15 Posts |
Posted - 09 Aug 2005 : 18:04:57
|
| I took a look at topic 1902 - I'll experiment with some of those registry entries. |
|
|
|
pacsciinst
15 Posts |
Posted - 15 Aug 2005 : 16:11:45
|
Hello,
Any news on this topic? Were you able to find when the new driver was implemented? Any ideas on how to resolve my issue?
Thanks!
|
|
|
|
hchin
30 Posts |
Posted - 26 Aug 2005 : 16:22:42
|
An update on this investigation.
We are still in the process of looking into the differences. Looking deeper into this, even though the display "turns off" during sleep and timeouts, we have confirmed that the PNL_PWR stays ON (5v or 3.3v) in CE 4.20.17.
You said that in CE 4.20.17 the backlight does not turn off during timeout but rather dims a little and flicker, if you use the new keys found in the document at: ADS backlight driver spec, topic 1902, I think this will resolve your issue.
We added the following keys to our adsload.reg. Since you say the polarity of your backlight is irrelevant (inverted on connector board) then this should work for you as well:
;---------------------------------------------------------------------
; Power Management: Backlight Settings
; Spec version 1.0 document 100050-10013
;---------------------------------------------------------------------
[HKEY_LOCAL_MACHINE\Platform\Backlight]
"StateForOn"=dword:0
"BacklightRunLevel"=dword:FF
"BacklightIdleLevel"=dword:0
|
Let us know how this works for you. |
|
|
|
pacsciinst
15 Posts |
Posted - 26 Aug 2005 : 18:37:46
|
Thanks for the response. I entered the above entry into my .reg file and updated my development platform. Unfortunately, the settings did not make a difference for the backlight operation. I used regedit to verify the registry keys were added and updated. The legacy entries (BLightLevel4Off...) are still in the \Backlight key - would this make a difference?
Thanks. |
|
|
|
hchin
30 Posts |
Posted - 29 Aug 2005 : 16:21:20
|
The legacy keys should not make a difference, however, have you tried removing them to see if it makes a difference? Also, could it be possible that the PWM input on your inverter is the opposite polarity to our standard inverter (LS520)?
Is it possible to try one of our standard panel and inverter (Sharp LQ64D343 and Xentek LS520)? |
|
|
|
pacsciinst
15 Posts |
Posted - 29 Aug 2005 : 16:58:55
|
I believe I have already eliminated our inverter/display as the problem in this scenario. The problem I see is the inconsistency in the OS's ability to control the backlight enable line (/BACKLIGHT_ON) and the LCD panel power (/PNL_PWR.)
To restate the problem: Our entire system works correctly with version 4.2.14. LCD panel power and the backlight enable lines function correctly. However, with 4.2.16/17, the /BACKLIGHT_ON stays low (enabled), and the /PNL_PWR pin on the LCD interface stays at 3.3V during sleep mode. Have you been able to confirm that the /BACKLIGHT_ON pin on your system functions correctly? On our system, it does not.
|
|
|
|
hchin
30 Posts |
Posted - 29 Aug 2005 : 18:25:47
|
Hello,
As I have stated above, with the CE 4.20.17 build, I have verified that the PNL_PWR stays ON (high) during both sleep and timeouts. As for the /BACKLIGHT_ON signal, in CE 4.20.17, we have confirmed that it is disabled (high) during sleeps but stays enabled (low) during timeouts. However our backlight turns off since the PWM signal goes high (5V).
So we do acknowledge the differences between the builds. It is in the queue to address these issues.
Thanks. |
|
|
| |
Topic |
|
All Forums
Microsoft Windows CE
BitsyX Panel Power and Backlight PWM in CE 4.20.17
Forum Locked
Send Topic to a Friend
Printer Friendly