04-20-2017 03:18 PM - edited 04-20-2017 03:25 PM
Thinkpad X1 Carbon 3rd Generation
i7 5600 CPU, 8GB RAM, 256GB SSD
Windows 10 Creators Update x64 (clean install, currently build 15063.138)
Issue also present after using rescue media to return to Windows 8.1
Steps to duplicate:
1. Unplug the AC adapter.
2. Put the machine to sleep.
3. Turn off the WiFi access point the machine normally connects to.
4. While still unplugged, wake the machine up.
5. Attempt to connect to a weak WiFi access point. Occasionally (not always) the screen will beging to flash and flicker erratically, with some occasional horizontal displacement. If left alone, the flashing will often degenerate into a totally black screen.
6. Moving the mouse or opening the START menu will sometimes stop the flashing while there is movement on the screen but the flashing resumes as soon as the screen once again becomes static.
7. The ONLY way to stop the flashing altogether once it starts is to plug the AC adapter back in. This ALWAYS stops it and (so far) it has NEVER misbehaved on AC power.... just on battery.
Alternate way to duplicate:
1. Generate a moderate RF field. In my case, I discovered this completely by accident. I made a brief transmission from a palm sized hand held 2W transciever on 443 MHz with the machine on my lap and on battery power. To my surprise it EXACTLY duplicated the flashing issue..... the screen began flashing erratically and continued after the transmission stopped. It eventually became completely black, the same as if I had been trying to connect to a weak WiFi access point. The field from a 2W transmitter up to three FEET from the laptop will sometimes cause it to begin flashing. As above, once the flashing started the only way to stop it was to plug in AC power.
Troubleshooting steps so far:
1. Set all BIOS and Power Manager parameters to minimize the difference between AC power and battery. NO CHANGE.
2. Make sure all drivers and chipset drivers are current and up to date. They are. NO CHANGE.
3. Revert to Windows 8.1 via Rescue Media. NO CHANGE.
4. Per a tip on a Dell forum (Dell and Asus machines with the same chipset also seem to have the same issue), reduced the WiFi transmit power from HIGHEST to MEDIUM (assumed 50%). ISSUE RESOLVED. However, not a totally acceptible work around since it reduces the effective range.
4. Replace the WiFi card (warranty replacement from Lenovo). NO CHANGE as long as the transmit power is at the default "HIGHEST" setting.
6. Opened a dialog with Intel. After much correspondence, Intel states that this appears to be an issue with the integration of their WiFi chipset and therefore not an Intel issue per se. I tend to agree. Intel suggested I continue to pursue the issue with Lenovo.
7. Opened a dialog with Lenovo warranty (the machine is currently under warranty). Lenovo warranty personnel initially wanted to replace the LCD and then the mainboard. I decided against that for now since there are posts elsewhere that these steps did NOT resolve the issue for others with the same symptoms. When the replacement WiFi card did not resolve it, they suggested that I post the issue here because "Lenovo specialists monitor the forum". Seems like a roundabout way to solve the issue and/or involve Lenovo (kinda like complaining on Twitter), and especially for a supposedly "business class" machine, but here I am.
The WiFi card is generating sufficient RFI to cause erratic behavior in an unknown internal device (possibly the HD5500 graphics adapter). As such, a "fix" would seem to be problematic?
OK, so I have done most of the heavy lifting. Does Lenovo have a solution? Am I mis-understanding the root cause? And is Lenovo interested in pursuing a solution? Unless I run at reduced WiFi transmit power, the machine occasionally becomes useless on battery power when trying to connect to some access points.
Solved! Go to Solution.
04-21-2017 03:41 AM - edited 04-21-2017 03:49 AM
This morning I tried an experiment. I disabled the internal WiFi adapter and connected a USB stick style 2.4GHz Linksys adapter on a USB extension cable. By observing the LED, I discovered that I can cause it to transmit repeatedly if I tell it to connect to a neighbor's AP but stop short of supplying the password. Left in that state, the LED flashes continuously and I assume the card is transmitting for each LED flash.
While on battery power, I can stimulate the screen flashing 100% of the time by simply holding the USB stick against the rear of the display and near a spot even with the bottom of the screen. I think the USB stick antenna is about 0.5" above the bottom edge of the screen and about 2" in from the left edge. Once started, the flashing continues (or sometimes the screen eventually goes completely black) until I reapply AC power.
So what say Lenovo? Looks like a confirmed case of internal RFI. Got a better solution than simply limiting the WiFi transmit power?
04-21-2017 03:13 PM - edited 04-21-2017 03:14 PM
And just when I thought I was converging on the root cause, I discovered that in my search for anything that changes state on battery power, I missed some "battery vs. AC power" settings in the Intel HD Graphics control panel.
When I set the Graphics Performance on battery to "maximum performance" and disabled "panel self refresh", "extended battery life for gaming" and "display power saving technology", I can no longer get the display to flash when attempting to connect to a weak AP, even with the transmit power set to MAXIMUM. That is real progress. HOWEVER, it appears as though the AC-7265 resets briefly while attempting to connect. The WiFi icon is replaced with the Ethernet icon and the AC-7265 momentarily disappears from device manager. A second or two later the WiFi icon reappears, the AC-7265 reappears in device manager and the machine reports it is unable to connect (understandable since the AP I am attempting to contact for this test is VERY weak and never connects reliably if at all).
If this behavior remains consistent, this may be a more acceptable work around than reducing WiFi transmit power. But it begs the question.... why does the WiFi adapter seemingly reset? Is that normal behavior?
My original hypothesis may still be correct that this is an RFI issue. Changing the display adapter settings may simply make it more immune to getting into a continuous / unrecoverable flashing state. More testing is needed.
Still waiting to hear from Lenovo.
04-27-2017 05:49 AM - edited 04-27-2017 06:29 AM
OK, I think I am zeroing in on a mostly acceptable solution. No thanks to Lenovo or Intel.
Since I can stimulate the issue at will with an external RF field from a 2W 433MHz transmitter, it makes it easy to explore solutions and be fairly confident that they are in fact 100% effective. Remember, this external RF field EXACTLY duplicates the flashing screen I sometimes get when trying to connect to a weak WiFi access point on battery power.
And what I have found is:
1. On AC power, the external field DOES make the screen flash. But just once. It then recovers and normal operation continues.
2. On battery power, the external RF field makes the screen flash once, same as on battery power. But it then continues to flash erratically forever even with the external RF field removed.
3. Setting the option "Panel Self Refresh" to DISABLED when on battery power in the Intel HD graphics application prevents the flashing from developing into continuous flashing. It flashes once and then recovers, same as on AC power.
Disabling the option "Panel Self Refresh" would seem to prevent some power savings on battery power. How much is unknown. However, the benefit would seem to outweigh the disadvantage.
Again, this still looks to me to be an RFI issue between the RF field generated by the WiFi card (possibly more intense during an attempted connect to a weak AP) and the graphics adapter or possibly the display itself. If that is true, there is probably NO hardware fix short of redesign since it is caused by an RF sensitivity in the display or graphics adapter in combination with the placement and orientation of the WiFi antennas (in the display itself). It might explain why Lenovo thinks it can be fixed by replacing the LCD and/or the system board. However, if successful those "fixes" would be extremely tentative and the issue would possibly reappear at some later date. Unless there IS a redesigned part they aren't telling me about?
I think I am at the end of the road here. I have a work around with minimal negative impact. But unless I am the ONLY one who has discovered what appears to be the root cause, I am very disappointed in the lack of response by Lenovo. I understand if Lenovo does not actually have a fix. But at least say so? The way it has been left makes it appear as though they A) know what the issue is and B) are ignoring it in the hopes I and others with the same issue will just give up and go away.
Lenovo.... still "over to you".
Shortly after I posted this, I got an Email from Lenovo. They are suggesting that replacing the LCD and cable will fix this. We're in negotiations..... I have seen several posts saying that replacing the LCD does NOT solve it and I am trying to determine if this is a guess or a known issue / fix. If it is a known issue, it is a well kept secret.
05-03-2017 11:34 AM - edited 05-03-2017 11:35 AM
OK, follow up time. I sent the machine in for service and it came back with a new mainboard, LCD cable (but NOT the display itself) and (I think) WiFi card. In my initial tests, I can no longer duplicate the issue with WiFi connect attempts OR an externally applied RF field. So it would appear that somewhere in there, the issue was addressed.
The only fly in the ointment was that since they replaced the mainboard, Windows deactivated and my first couple attempts to reactivate failed. The reason the reactivation failed is unclear.... it should have worked. After several attempts (and in spite of a statement on the activation page that the machine "could not contact the activtion server"), I was asked to sign into my MS account and switch to a Microsoft account (I normally use a local account). When I did that, it reactivated and I switched back to a local account.
So..... all seems to be well.
05-06-2017 09:39 AM - edited 05-06-2017 08:09 PM
OK, one last (I hope) follow up.....
I am spending the weekend away from my home WiFi and in places the (open) WiFi here is VERY weak or even not QUITE able to connect. And I am running 100% on battery. The exact scenario that used to cause screen flashing. And I am happy to report there is NO flashing or other odd behavior with all WiFi and display power saving options to their defaults. I can finally call this one solved.
12-17-2017 08:52 AM - edited 12-17-2017 08:59 AM
It has now been about eight months since I went through all that and sent the machine back to Lenovo for service. I am happy to report that the issue is still 100% resolved. I cannot even stimulate it with a nearby VHF transmitter.
I was initially skeptical but whatever Lenovo did to mine (I think they replaced the display cable and not the entire display) did indeed resolve it. My advice would be to sent it in for service if it is still under warranty. Make sure you include a DETAILED description of the issue and maybe even a reference to this thread since they might try to claim they cannot replicate the issue.
If you cannot send the machine in, reducing the WiFi power does solve it for the most art.
10-27-2018 06:08 PM
I purchased a Thinkpad X1 Carbon 3rd Generation
i7 5600u CPU, 8GB RAM, 500GB SSD on ebay, in the ad it said that the equipment was 100% but when I turned it on and got it back from the factory I saw the problem described here.
Not getting any help from the seller I chose to repair the equipment by myself, with the help of pgoelz1 I confirmed that the problem is interference.
I solved the problem just by putting a sticker of aluminum in the zones where the cord of the wireless passed next to the cable of lcd as you can see in the images.
Thanks pgoelz1 for the tip
10-27-2018 06:58 PM
Interesting fix. I wonder if you did the same thing with the foil that a replacement cable accomplished in mine? I always assumed the cable that they replaced was defective. Perhaps it was simply re-designed. Too bad I had to go through all the angst with mine... I still have a roll of copper foil tape from my engineering days and I would not have been afraid to use it