Thinkpad p50 Sierra 10.12.6

[Topuz]

That's very good. I managed to drop my idle consumption to about ~1.3W by disabling unused devices (Intel WiFi and Card Reader via ACPI and some more through BIOS) and undervolting. My CpuFriend is set to "Balanced Power" mode. I could probably go under 1W with more tweaking but I have 4k with 2xSSD + HDD so I'm not sure if there's a point to push further.

 

[midi1996]

That's very good. I managed to drop my idle consumption to about ~1.3W by disabling unused devices (Intel WiFi and Card Reader via ACPI and some more through BIOS) and undervolting. My CpuFriend is set to "Balanced Power" mode. I could probably go under 1W with more tweaking but I have 4k with 2xSSD + HDD so I'm not sure if there's a point to push further.

that may also be the cause of the high consumption. my setup a Xeon E3-1505M v5 with 1080p screen and (for now) 1 nvme + 1 hdd. The wifi card is as you saw before, expresscard and sdcard are disabled through uefi, fp, camera, xrite color calib are disabled through USBMap, nvidia through a custom ssdt (not rehabman's, thanks to maemo for that, it's based off bumblebee way of disabling the dgpu).

 

[Topuz]

it's based off bumblebee way of disabling the dgpu

Can you elaborate? What are the differences and/or benefits compared to the usual method?

 

[midi1996]

Can you elaborate? What are the differences and/or benefits compared to the usual method?

ACPI for Developers

Bumblebee daemon and client rewritten in C. Contribute to Bumblebee-Project/Bumblebee development by creating an account on GitHub.

github.com

/*
* Disable discrete graphics for Optimus (credit) Bumblebee-Project
* >> https://github.com/Bumblebee-Project/Bumblebee/wiki/ACPI-for-Developers#acpi-and-the-nvidia-card
*
* --- --- --- --- ---
*
* Important:     Change all ACPI paths accordingly.
*                dGPU path is _SB_.PCI0.RP01.PEGP / \_SB.PCI0.RP01.PEGP in this example.
*                Optimus dGPUs contain NVOP / NV3D / WMMX method, inside one of the SSDTs or the DSDT.
*
* If successful, Display / nVidia disappears from Graphics/Displays << System Information
*/

DefinitionBlock ("", "SSDT", 2, "MAEMO", "NoHybGfx", 0x00000000)
{
    External (_SB_.PCI0.PEG0.PEGP._DSM, MethodObj)    // 4 Arguments (from opcode)
    External (_SB_.PCI0.PEG0.PEGP._PS3, MethodObj)    // 0 Arguments (from opcode)

    Device (NHG1)
    {
        Name (_HID, "NHG10000")  // _HID: Hardware ID
        Method (_INI, 0, NotSerialized)  // _INI: Initialize
        {
            // If methods exist
            If (LAnd (CondRefOf (\_SB.PCI0.PEG0.PEGP._DSM), CondRefOf (\_SB.PCI0.PEG0.PEGP._PS3)))
            {
                // Card off preparation, card off
                \_SB.PCI0.PEG0.PEGP._DSM (ToUUID ("a486d8f8-0bda-471b-a72b-6042a6b5bee0"), 0x0100, 0x1A, Buffer (0x04)
                    {
                         0x01, 0x00, 0x00, 0x03                         
                    })
                \_SB.PCI0.PEG0.PEGP._PS3 ()
            }
        }
    }
}

 

[midi1996]

Yes I fixed the LED issue, I'm writing the guide atm. Sorry I couldn't finish it before, because I had many real life stuff (you know what's going on now).

    Method (_WAK, 1, NotSerialized)  // _WAK: Wake
    {
        If (_OSI ("Darwin"))
        {
            If ((0x03 == Arg0))
            {
                \_SI._SST (One)
            }

            If ((Arg0 < 0x04))
            {
                If (((\RRBF & 0x02) || (B1B2 (\_SB.PCI0.LPC.EC.AC10, \_SB.PCI0.LPC.EC.AC11) & 0x02)))
                {
                    Local0 = (Arg0 << 0x08)
                    Local0 |= 0x2013
                    \_SB.PCI0.LPC.EC.HKEY.MHKQ (Local0)
                }
            }
        }

        Local0 = \XXAK (Arg0)
        Return (Local0)
    }

 

[Topuz]

That's great news, can't wait to see the guide. I'll probably try switching to OpenCore once you publish it.

By the way, my DW1820A arrived couple of days ago. I'm posting my experience so other people don't need to waste couple of days on troubleshooting as I did. Anyways, it's a DW1820A labeled as CN-096JNT. Got it for $17 on eBay. The initial awkward moment is that it has BCM94356ZEPA50DX_2 written on the back instead of the usual BCM94350ZAE_2. It identifies itself however as 14e4:43a3 and 1028:0023. This doesn't match any of the data I managed to find online. Anyways, I had a hard time getting it to work under Catalina and strangely couldn't get it to work under Windows at all (drivers would install fine, but WiFi would be disabled and I couldn't turn it on). Eventually I managed to get it running under Catalina with only AirPortBrcmFixup and pci-aspm-default 0 injected through Devices-Properties (not sure if any of it was necessary, read on to see why). Since I rely on dual boot and can't have a non-functional Wifi under Windows, I was about to give up because no driver that I managed to find allowed me to turn on the WiFi. Further more, if I booted Windows and then rebooted to Mac OS, WiFi would get disabled in Mac OS as well. Bonkers. As a last measure, I resorted to pin masking as I was ready to give up on it and tried it only because I had my keyboard already removed. And behold, once I masked 5 pins (2+3) everything started working as smooth as butter under both Windows and Mac OS. What baffled me really (and I tried this just for fun) is that I removed both AirPortBrcmFixup and pci-aspm-default 0 (so completely vanilla WiFi with no additional kexts or injections whatsoever, only masked pins) and WiFi still works perfectly fine (I can even reboot back and forth from Windows to Mac OS and vice versa without any issues). I'm at a loss how this is possible since everyone says that pretty much you must inject pci-aspm-default to avoid lock ups and freezes on any variation that's not CN-0VW3T3. Maybe it's a mislabeled CN-0VW3T3, who knows... On the Bluetooth side, I didn't really bother much since I don't use it, but I installed the Catalina DW1820A Bluetooth bundle from Daliansky's blog and I was able to transfer a file to my phone via BT, so I guess that's fine as well.

 

[midi1996]

That's great news, can't wait to see the guide. I'll probably try switching to OpenCore once you publish it.

By the way, my DW1820A arrived couple of days ago. I'm posting my experience so other people don't need to waste couple of days on troubleshooting as I did. Anyways, it's a DW1820A labeled as CN-096JNT. Got it for $17 on eBay. The initial awkward moment is that it has BCM94356ZEPA50DX_2 written on the back instead of the usual BCM94350ZAE_2. It identifies itself however as 14e4:43a3 and 1028:0023. This doesn't match any of the data I managed to find online. Anyways, I had a hard time getting it to work under Catalina and strangely couldn't get it to work under Windows at all (drivers would install fine, but WiFi would be disabled and I couldn't turn it on). Eventually I managed to get it running under Catalina with only AirPortBrcmFixup and pci-aspm-default 0 injected through Devices-Properties (not sure if any of it was necessary, read on to see why). Since I rely on dual boot and can't have a non-functional Wifi under Windows, I was about to give up because no driver that I managed to find allowed me to turn on the WiFi. Further more, if I booted Windows and then rebooted to Mac OS, WiFi would get disabled in Mac OS as well. Bonkers. As a last measure, I resorted to pin masking as I was ready to give up on it and tried it only because I had my keyboard already removed. And behold, once I masked 5 pins (2+3) everything started working as smooth as butter under both Windows and Mac OS. What baffled me really (and I tried this just for fun) is that I removed both AirPortBrcmFixup and pci-aspm-default 0 (so completely vanilla WiFi with no additional kexts or injections whatsoever, only masked pins) and WiFi still works perfectly fine (I can even reboot back and forth from Windows to Mac OS and vice versa without any issues). I'm at a loss how this is possible since everyone says that pretty much you must inject pci-aspm-default to avoid lock ups and freezes on any variation that's not CN-0VW3T3. Maybe it's a mislabeled CN-0VW3T3, who knows... On the Bluetooth side, I didn't really bother much since I don't use it, but I installed the Catalina DW1820A Bluetooth bundle from Daliansky's blog and I was able to transfer a file to my phone via BT, so I guess that's fine as well.

LMAO ok, so my DW1820A is actually a Lenovo part that the shady seller restickered it as a Dell one (smh, but i got it for $9 so it was fine). Tbh it doesn't matter which model you have. Just make sure you update to the latest BIOS version as it got the whitelisting disabled (fng finally). For the Windows/macOS issue, i never had that, maybe because i have the latest bios version maybe? I literally just plugged the card, boot up windows, installed some 2018 broadcom wifi drivers (will link them in the guide) and some slightly older bt drivers, they work. And for macOS, if i boot with the wifi card (untapped) it would slow down macOS and all the other symptoms, but using that aspm flag helped a lot. Rebooting between windows/macOS/linux was never an issue on my part. Bluetooth is really what concerns me, and i did the acidanthera rampatch but it didnt work, so i didnt investigate anymore, maybe ill test some other people's modded drivers but i dont trust binaries too much unless there is the sauce. Oh well, it appears that my neighborhood is experiencing a power outage atm, so ill postpone writing the guide for a later time, but i got all the files and everything ready for posting.