Contribute
Register

[SUCCESS] Gigabyte Designare Z390 (Thunderbolt 3) + i7-9700K + AMD RX 580

djx8605 said:
I have both an RX 580 and RX 590, both Nitro+. They work fine on silent mode and regular mode; is your card getting enough power? Are you using both 6+2 and 6 pins connectors? Or are you only using the 6+2 connector?
I've used both of these cards on the Z390 Aorus Pro, Z390 Aorus Pro WIFI, Z390I Auros Pro WIFI and now the Z390 Aorus Designare. Both of these cards work great, with either BIOS switch setting. I did always install them on PCIE slot 1, except on the Z390I which of course only has 1 slot :)

Note that these cards only have the newer EFI video driver stack, not the older legacy driver, so some OS like Windows 7 will give you a black screen unless CSM is enabled, but this is not a failure with the card, just a driver issue. A modern OS like Windows 10 and I'd imagine Mojave/Catalina, doesn't need CSM enabled to work properly though.

In my cases, the fans stay completely off (0 stop) in either mode (silent/standard), as long as the temps are low. I think 50C is the actual temp where the fans kick in. With such a big heatsink, I'd imagine the card can stay cool for a few temps :)

EDIT: Another thought, are you using the default VBIOS for the card? Or did you modify/switch them out? I know allot of people do this for compatibility or other reasons. [FYI] Mine are running the stock VBIOS that they came with...
Click to expand...

About VBIOS: I actually got it used so I have to ask and find out about that. Sadly I don't have windows installed on this machine and I don't know if there's a way to find that info in MacOS...

By the way one of the two BIOS settings on the GPU hangs on BIOS with CSM off. The red light (VGA) on my Designare z390 stays lit and the Boot doesn't ever proceed, can't even get in the BIOS itself.

I noticed that on 50 degrees the fans DO start but they never go higher than 25%, even on 75 degrees. The performance is great but that fan curve concerns me a bit.
 
iMac 19,1 on OpenCore
Z390 DESIGNARE |  i9 9900K |  RX 580 Pulse 8GB
z4k said:
About VBIOS: I actually got it used so I have to ask and find out about that. Sadly I don't have windows installed on this machine and I don't know if there's a way to find that info in MacOS...

By the way one of the two BIOS settings on the GPU hangs on BIOS with CSM off. The red light (VGA) on my Designare z390 stays lit and the Boot doesn't ever proceed, can't even get in the BIOS itself.

I noticed that on 50 degrees the fans DO start but they never go higher than 25%, even on 75 degrees. The performance is great but that fan curve concerns me a bit.
Click to expand...

Hmmm, since you have it used, I'd verify the VBIOS are not modified, this could be causing the issues you are talking about, such as fan curves or a corrupt VBIOS that won't launch.
 
  • Like
Reactions: z4k
@manfriday , Thank you for the EFI, you do have a lot of stuff in that folder, I have now moved my install to a simple open core EFI, however unfortunately it has not resolved my Studio One issue, but thanks for your effort.
 
Hi @kwalter,

I have a question, I have a similar problem.

Did you solve the problem:
2019-08-08 09:21:17.292246+0100 0x71 Default 0x0 0 0 kernel: (AppleACPIPlatform) [\_TZ.TZ10._STA] (Node ffffff8051c06b10)
2019-08-08 09:21:17.292247+0100 0x71 Default 0x0 0 0 kernel: (AppleACPIPlatform) [\_TZ.TZ10._STA] (Node ffffff8051c06b10)

Thanks for the help.
 
djx8605 said:
Hmmm, since you have it used, I'd verify the VBIOS are not modified, this could be causing the issues you are talking about, such as fan curves or a corrupt VBIOS that won't launch.
Click to expand...

OK something quite strange is happening. The fan "problem" still exists as I mentioned it.

The "Quiet" BIOS on my RX580 which I mentioned earlier didn't work with CSM off, turns out that it works BUT The red light (VGA) on my Designare z390 stays lit and nothing is displayed on the screen until I get to the MacOS login. Not even BIOS screen, no macOS loading nothing. The VGA light on the MoBo remains lit all the way through. Any ideas what that might be? (Checked the clocks on both modes, they are as expected so chances of modified VBIOS are lower)

Fan still never goes higher than 25% (in my standard "Performance" Bios of the GPU) even at 75 degrees. :crazy:

Maybe it’s just the aggressive MacOS fan curve, waiting to ramp up at something like 80? Or is it not communicating correctly with the OS? How are your RX 580 nitro+ special Edition fans spinning?
 
Last edited:
iMac 19,1 on OpenCore
Z390 DESIGNARE |  i9 9900K |  RX 580 Pulse 8GB
CaseySJ said:
@xanderevo,

Let's see if the community can help. It seems you've already tried some of the suggestions that have worked in the past:
  • Disable XMP or reduce memory speed to 2666 MHz or 2933 MHz
  • Disable on-board audio
Some questions so others can also help:
  • If you've turned off on-board audio, through which device are you playing audio?
    • HDMI or DisplayPort audio through your monitor?
    • Sound card? If so:
      • Which make/model?
      • How are speakers connected to the sound card?
        • RCA, XLR, ?
  • If you remove the sound card and use on-board audio, do the pops/clicks still occur?
Click to expand...

Yeah, do all of that and more:

It sounds like a sample rate change. Like when you open a DAW that has a locked sample rate at 44.1 KHz and you accidentally open another DAW that was last locked at 48 KHz, etc.

Are you using that Clover driver that loads the "startup sound" -IDK what bitrate that file is, nor its sample rate. I'd remove any startup sounds for a DAW machine. Also, disable any and all user feedback sounds, then select "Tink" as the alert sound, then yank the volume slider to zero. FWIW, I never sleep my machines. It's just trouble for audio devices.
 
Daystrom M5 Multitronic Unit
Z390 DESIGNARE |  i9 9900K |  RX 6600 Pulse 8GB
The Guardian of Forever
Z370 AORUS Gaming 7 |  i7 8700K |  RX 580 Pulse 8GB
kejocu said:
Hmm, I re-checked that just now and it looks like both parameters are set as needed. Could it be that F9b vs F8 is causing the issue? Maybe with some new option toggled somewhere... Here is a picture of the Thunderbolt BIOS page. View attachment 448815
Click to expand...
@CaseySJ just in case my reply got buried. I bet it must be a miscongfigured setting on my part (though which, I'm not sure), but it is worth raising in case something deeper/more interesting is going on.
 
CaseySJ said:
Agreed! And only a hard reboot makes this happen. A proper reboot (Apple Menu --> Reboot) is not sufficient...

Correct -- ThunderboltConfig is being used by macOS drivers.

Not sure I understand. Which bytes should be 01 and 02?
<00 02 1c 00 02 00 05 03 01 00 04 00 05 03 02 00 03 00 05 03 01 00 00 00 03 03 02 00 03 00 04 00>

I spent most of the time trying to understand why CRMW was failing -- and it is still failing! :) Nevertheless, I attempted to hot plug bot TB3 and USB-C, but unfortunately the system hanged as before with more or less the same ACPI Debug Log messages.

I believe there is something quite wrong in CRMW. Here is a partial log from a cold boot just a few minutes ago:

2020-02-08 03:30:50. ACPIDebug: "TBT & USBC already on."
2020-02-08 03:30:50. ACPIDebug: "UGIO Exit"
2020-02-08 03:30:50. ACPIDebug: "PCED UPSB- restored flag, THUNDERBOLT_PCI_LINK_MGMT_DEVICE.PRSR"
2020-02-08 03:30:50. ACPIDebug: 0x0
2020-02-08 03:30:50. ACPIDebug: "PCED UPSB- Wait for config space..."
2020-02-08 03:30:50. ACPIDebug: { "PCED UPSB- Read VID/DID =", 0x15ea8086, }
2020-02-08 03:30:50. ACPIDebug: "UPSB PCED Exit"
2020-02-08 03:30:50. ACPIDebug: "CRMW Enter"
2020-02-08 03:30:50. ACPIDebug: "CRMW Read Value1"
2020-02-08 03:30:50. ACPIDebug: 0x34021
2020-02-08 03:30:50. ACPIDebug: "CRMW Write Value1"
2020-02-08 03:30:50. ACPIDebug: 0x4034021

...9 second gap...and CRMW does not return...There should be 2 calls to CRMW from UPSB._PS0()...

2020-02-08 03:30:59. ACPIDebug: { "RTPC NHI0 called with arg: ", 0x0, }
2020-02-08 03:30:59. ACPIDebug: "_PS3 NHI0, Nothing to do!"
2020-02-08 03:30:59. ACPIDebug: "_PS3 DSB0"
2020-02-08 03:30:59. ACPIDebug: "DSB0 PCDA"

As shown in spoiler below, I added a Mutex and extra logging to CRMW.
Code: 
            Method (CRMW, 5, Serialized)
            {
                 Acquire (CRMX, 0xFFFF)
//                 Sleep (200) // sleep 200ms to wait for ACPI Debug to catch up (or something like that!)
                Local1 = One
                DBG1 ("CRMW Enter")
                If (((GGOV (FPG0) == FP0L) && (\_SB.PCI0.RP05.UPSB.AVND != 0xFFFFFFFF)))
//                If ((\_SB.PCI0.RP05.UPSB.AVND != 0xFFFFFFFF))
                {
                    Local3 = Zero
                    While ((Local3 <= 0x04))
                    {
                        Local2 = CIOR (Arg0, Arg1, Arg2)
                        If ((DerefOf (Local2 [Zero]) == Zero))
                        {
                            Local2 = DerefOf (Local2 [One])
                            DBG1 ("CRMW Read Value1")
                            DBG1 (Local2)
                            Local2 &= ~Arg4
                            Local2 |= Arg3
                            DBG1 ("CRMW Write Value1")
                            DBG1 (Local2)
                            Local2 = CIOW (Arg0, Arg1, Arg2, Local2)
                            If ((Local2 == Zero))
                            {
                                Local2 = CIOR (Arg0, Arg1, Arg2)
                                If ((DerefOf (Local2 [Zero]) == Zero))
                                {
                                    Local2 = DerefOf (Local2 [One])
                                    DBG1 ("CRMW Read Value2")
                                    DBG1 (Local2)
                                    Local2 &= Arg4
                                    If ((Local2 == Arg3))
                                    {
                                        DBG1 ("CRMW Success")
                                        Local1 = Zero
                                        Break
                                    }
                                    Else
                                    {
                                        DBG1 ("CRMW Mismatch with Arg3")
                                    }
                                }
                                Else
                                {
                                    DBG1 ("DerefOf second CIOR != Zero")
                                }
                            }
                        }
                        Else
                        {
                            DBG1 ("DerefOf first CIOR != Zero")
                        }

                        Local3++
                        Sleep (0x64)
                    }
                }
                Else
                {
                    DBG1 ("CRMW - UPSB.AVND == 0xFFFFFFFF")
                }
                DBG1 ("CRMW Exit")
                 Release (CRMX)
                Return (Local1)
            }

Thanks! Will do this now.
Click to expand...
Hi @CaseySJ
You should also try make a Reset when booting with no devices connected and see If TNode/ TBus appears.

About ThunderboltConfig trying patch, this is the last following bytes :
replace : <00 02 1c 00 02 00 05 03 01 00 04 00 05 03 02 00 03 00 05 03 01 00 00 00 03 03 02 00 03 00 04 00>
by : <00 02 1c 00 02 00 05 03 01 00 04 00 05 03 02 00 03 00 05 03 01 00 00 00 03 03 02 00 01 00 02 00>
​
I don't think that Mutex is needed or something like Sleep () for ACPIDebug, you should only modify included PLIST file on ACPIDebug by changing PollingInterval property value from 100 to 10. As default, I might be at 100ms polling time. Changing to 10ms might improve Debug log.
 
NorthAmTransAm said:
Wow thank you so much. I'm excited to play along and understand more as you all work. Just to recap.

Replace _L6F in DSDT with XL6F. (Found it by searching "Method (_L6F")

Replace the GPE method content on the SSDT with your new content above.

Change these:
FPGO to FPGN
FPLV to FP0L

Correct?
Click to expand...
Hi @NorthAmTransAm
half right half wrong !
Change these:
FPGO to FPGN >> Right
FPLV to FP0L >> Wrong​
rather FP0L to FPLV
There is two variable couple for Tbt Force power, depending on ACPI Bios firmware:
either FPG0/FP0L (this a zero, not O)
or FPGN/FPLV​
 
You must log in or register to reply here.
Copyright © 2010 - 2024 tonymacx86 LLC
Back
Top