Jump to content
IGNORED

AudioLinux and NUC Troubleshooting and Tuning


rickca
Message added by austinpop

Summary of useful findings and recommendations

 

This section will be a living repository of useful info from this thread. It's very similar to a wiki and will be maintained by a small group of thread moderators.

 

Before you get started please refer to the Audio-Linux website to ensure you have the latest info and the proper versions of the OS. Audio-Linux.com  

 

**** Updated for AL 1.30 menu 118 or later.

 

  "First Run" setup for headless.  

 

Setup your NUC with a keyboard, mouse, and monitor to the NUC BIOS settings.  From the menu note the IP address of the machine to SSH into.  From a MAC the macOS terminal program supports SSH:

324537708_ScreenShot2019-01-28at3_02_19PM.thumb.png.739dc7f9cdb05e04da806c7c66877332.png

 

Then it is simpler to cut and paste into the terminal session. After entering the password for the audiolinuxuser you will be presented with the AL headless menu:

 

1518375894_ScreenShot2019-01-28at3_04_18PM.thumb.png.a7b2867a163f8f014e56e52ff69f94b4.png

 

Option 8 takes you to the command line for the following basic setup.  You will need to be the Root user for this setup and the su command first:

 

su

 

Fix the time zone:  (this is my timezone - look in directory /usr/share/zoneinfo)

 

timedatectl set-timezone America/Chicago
 

Setup and Start NTP daemon (to keep the system time in sync)

 

*** the config file is now properly filled in.

 

Now Start the daemon

 

timedatectl set-ntp true

 

 

NOTE: Sometimes the system takes a little while to get synced up.
 

Set hostname  (this provides a unique name for the machine on your network.  Replace <NAME> with your chosen name)

 

hostnamectl set-hostname <NAME> 
 

Once the above items are set up your machine is ready to be configured for say a Roon bridge/endpoint. That is done using the AL menu.  To return to the menu do the following commands.

 

exit

menu

 

----------------------------------

 

For most of us, the following basic settings are key.

From the configuration menu:

6. START and enable Roonbridge

15. SET Realtime Priority to extreme

16. ENABLE ramroot (reboot after)

 

Return to the main menu and reboot the NUC using 

 

11 Reboot

 

------------------

Roon Server setup is a bit more complex and we will cover it completely a bit later.  The key is where you are booting from and where the Roon database is stored.  In general; say a 32gb OPTANE "SSD".

 

  • You have to partition the SSD into a boot drive and a storage drive.  
  • The transfer the USB stick install to the boot partition.  
  • Reboot from the boot partition.  
  • Do the basic setup. Timezone and name
  • Transfer the Roon Database to the storage drive
  • Start the Roon Server
  • .....

 

----------------- 

The machine will reboot and from the display attached to the NUC you can watch it boot up and load into RAM.  Once the AudioLinux menu is showing the endpoint should be available in Roon.    This completes the basic startup sequence.  The system is ready to start testing.    

 

 

Recommended Posts

33 minutes ago, austinpop said:

Ahh - looks like a heat issue. Not had that myself. Hopefully others can elaborate.

Unfortunately, NUC8i7BEH cooling is insufficient for sustained peak performance.  I clipped the following quote from

https://www.notebookcheck.net/Intel-NUC-Kit-NUC8i7BEH-i7-8559U-Mini-PC-Review.360356.0.html

 

We run synthetic loads to stress the system and identify for any potential throttling or stability issues. When subjected to Prime95 load, the CPU can be observed running as fast as 4.0 GHz before hitting a temperature ceiling of 88 C. Thereafter, clock rate immediately drops to a steady 2.9 GHz to maintain a cooler core temperature of 74 C as shown by our screenshot below. When considering that the base clock rate of the i7-8559U is 2.7 GHz, this means that the NUC is only able to sustain a Turbo Boost clock of just +200 MHz with only a temporary boost to 4.0 GHz for just the first initial few seconds. The results line up with our CineBench loop test above where performance is fastest during the start of the test before immediately dropping. Running both Prime95 and FurMark simultaneously will throttle both the CPU and GPU.

 

Also see the following thread

https://communities.intel.com/thread/145444

Pareto Audio AMD 7700 Server --> Berkeley Alpha USB --> Jeff Rowland Aeris --> Jeff Rowland 625 S2 --> Focal Utopia 3 Diablos with 2 x Focal Electra SW 1000 BE subs

 

i7-6700K/Windows 10  --> EVGA Nu Audio Card --> Focal CMS50's 

Link to comment
23 minutes ago, afrancois said:

what does the command "sensors" give you?

My CPU is at 72 °C

 

Mine show:

 

Package id 0: 64 °C

Core 0: 67°C

Core 1: 51°C

Core 2: 51°C

Core 3: 70°C

Intel NUC NUC8i7BEH Roon Server running Audio Linux in RAM -> Sonore UltraRendu (Roon Endpoint) -> Uptone ISO Regen -> Singxer SU-1 KTE -> Holo Audio Spring Level 3 DAC -> Nord One UP Monoblocks -> Spendor LS3/5as | Music controlled via iPad (Power Conditioning: Audience adeptResponse aR12).  Twitter: @hirezaudio

Link to comment
14 minutes ago, rickca said:

Unfortunately, NUC8i7BEH cooling is insufficient for sustained peak performance.  I clipped the following quote from

https://www.notebookcheck.net/Intel-NUC-Kit-NUC8i7BEH-i7-8559U-Mini-PC-Review.360356.0.html

 

We run synthetic loads to stress the system and identify for any potential throttling or stability issues. When subjected to Prime95 load, the CPU can be observed running as fast as 4.0 GHz before hitting a temperature ceiling of 88 C. Thereafter, clock rate immediately drops to a steady 2.9 GHz to maintain a cooler core temperature of 74 C as shown by our screenshot below. When considering that the base clock rate of the i7-8559U is 2.7 GHz, this means that the NUC is only able to sustain a Turbo Boost clock of just +200 MHz with only a temporary boost to 4.0 GHz for just the first initial few seconds. The results line up with our CineBench loop test above where performance is fastest during the start of the test before immediately dropping. Running both Prime95 and FurMark simultaneously will throttle both the CPU and GPU.

 

Also see the following thread

https://communities.intel.com/thread/145444

 

Interesting read! Thanks for posting the links. I'll follow the community thread to see if Intel comes up with any solutions.

Intel NUC NUC8i7BEH Roon Server running Audio Linux in RAM -> Sonore UltraRendu (Roon Endpoint) -> Uptone ISO Regen -> Singxer SU-1 KTE -> Holo Audio Spring Level 3 DAC -> Nord One UP Monoblocks -> Spendor LS3/5as | Music controlled via iPad (Power Conditioning: Audience adeptResponse aR12).  Twitter: @hirezaudio

Link to comment
1 hour ago, gsquared said:

 

No it's the stock case until Akasa makes a case that will work with this model.

 

@Dev have you experienced this with yours?

 

Yes, when there is compute intensive work, the CPU throttles easily. Currently I am running this as an endpoint with half-extreme mode which seems to do fine but running with full extreme will start throttling again. Either the cooling seems pretty insufficient with the stock or its bug reading the temps off the die but it does sound pretty good.

 

I noted this sometime earlier:

 

 

Link to comment
29 minutes ago, greenleo said:

After knowing the sensors command, I have left my stock 7CYJH headless at standard mode for 8 hours.  The temperature stayed at less than 42C.  At this temp, leaving it on 7x24 is not a problem.

 

If you have a fanless Akasa case, it should be able to do AL extreme with the same temperature. I leave it on 24x7.

Link to comment
3 minutes ago, Dev said:

 

If you have a fanless Akasa case, it should be able to do AL extreme with the same temperature. I leave it on 24x7.

Thank you.  I have the JC, just haven't switched the case yet.  I always switch the headless back to standard mode when not playing music for "green" reason?

Link to comment

Thinking of jumping in with  NUC7CJYH when it becomes available at BH or some other places. As Rajiv  has stated, large part of the SQ depends upon which LPSU is being used. Since the best SQ was achieved with SR-7, are there any other alternatives since SR-7 is almost impossible to get in a short term? (I'm aware of LPS 1.2 and  sbooster). Are there any other choices? If I choose sbooster, shall I opt for 18v to go with NUC7CJYH?

Link to comment
3 hours ago, Dutch said:

@hifi25nl,

 

Some questions after reading: http://www.audio-linux.com/html/realtime.html

 

I’m running just Roon Bridge on my AL headless (0.6) NUC (it’s the endpoint) and I’m seeing these related processes :

 

root       544     1  0 17:41 ?        00:00:00 /bin/sh /opt/RoonBridge/start.sh

root       549   544  0 17:41 ?        00:00:01 RoonBridge --debug --gc=sgen --server RoonBridge.exe

root       590   549  0 17:41 ?        00:00:10 RoonBridgeHelper --debug --gc=sgen --server RoonBridgeHelper.exe

root       595   549  0 17:41 ?        00:00:00 /opt/RoonBridge/Bridge/processreaper 590

root       601   549  0 17:41 ?        00:00:39 RAATServer --debug --gc=sgen --server RAATServer.exe

 

In the rtapp.conf file (below) I see RoonBridge is listed but not RoonBridgeHelper or RAATServer. Since these have used more CPU time, would it be beneficial if I/we added those processes to the applications list?

 

Also I see —debug in the process list? Is that necessary or better said would it be ‘leaner and meaner’ to not use those switches?

 

Thanks in advance. If you’d like me to email you instead please tell me. It’s just that I though this might be beneficial to other users as well. :)

audiolinux@audiolinux ~]$ more /etc/rtapp/rtapp.conf

APPLICATIONS="jackd mpd hqplayer hqplayerd RoonAppliance RoonBridge mediacenter24 networkaudiod deadbeef a2jmidid ardour-5.12.0 rosegarden audacity"

 

MAX_PRIORITY="93"

 

MODE="autodec"

 

For closed source application with multiple threads is sometime difficult to understand what thread must have higher priority of another thread.

 

With "autodec" mode a running  application will have an higher priority than the one following it in the list. 

You can experiment with this, I am curious if you can find differences. The configuration file is there to be tweaked for the better sound.

 

Remember that MAX_PRIORITY must be lower than dac irq priority in /etc/rtirq.conf

AudioLinux --> https://www.audio-linux.com

developer of AudioLinux realtime OS

Link to comment
16 hours ago, greenleo said:

I don't use the HQPE, E for embedded, correct?

 

Also, I don't use filters at all.  Sorry, can't help.

Noted with thanks. Yes, E stands for Embedded. The default SDM (1X) filter for the HQP Desktop and HQPE pre-installed on AL is “poly-sinc-lp-2s” which does not have the mentioned problem. But if you change it to non-2s filter, the problem bound to occur for HQPE.

Link to comment
4 hours ago, mourip said:

 

HDPlex now uses Linear Technology 3045 regulators. Their 200w version has 4 outputs and two are adjustable. At about $500 shipped they are a very good value and readily available. I have found their customer support to be excellent.

 

There have been a lot of posts about using the LPS1.2 but I'm personally struggling to get the power draw low enough for my LPS1.2 to be happy with the NUC. Will try again this morning but I also have an SPS-500 which should be capable of higher power output than the LPS1.2.

 

The Paul Hynes SR4 is also highly regarded but I think there is still a wait of a couple months. Relatively much faster than the SR7 but not ideal for impatient audiophiles.

 

PS I'm going to try the SPS-500 because I have one to hand but given the consensus was that the LPS-1 was better than the SPS-500, and given how much better the LPS1.2 was to the LPS-1 I'm not sure I'd advocate buying the higher priced SPS-500 if persevering and getting the LPS1.2 to work is an option.

 

Cheers,

Alan

Synergistic Research Powercell UEF SE > Sonore OpticalModule (LPS-1.2 & DXP-1A5DSC) > EtherRegen (SR4T & DXP-1A5DSC) > (Sablon 2020 LAN) Innuos PhoenixNet > Muon Streaming System > Grimm Audio MU1 server > (Sablon AES) Mola Mola Tambaqui DAC > PS Audio M1200 monoblocks > Salk Sound Supercharged Songtowers

Link to comment
9 hours ago, sam1000 said:

Thinking of jumping in with  NUC7CJYH when it becomes available at BH or some other places. As Rajiv  has stated, large part of the SQ depends upon which LPSU is being used. Since the best SQ was achieved with SR-7, are there any other alternatives since SR-7 is almost impossible to get in a short term? (I'm aware of LPS 1.2 and  sbooster). Are there any other choices? If I choose sbooster, shall I opt for 18v to go with NUC7CJYH?

auricgoldfinger posted his impression of the LPS Ref-1( I hope that my memory works correctly) from TLS and compared with his memory recall of the SR7.  This LPS sounds promising.  Please search his post.

Link to comment
10 hours ago, hifi25nl said:

 

For closed source application with multiple threads is sometime difficult to understand what thread must have higher priority of another thread.

 

With "autodec" mode a running  application will have an higher priority than the one following it in the list. 

You can experiment with this, I am curious if you can find differences. The configuration file is there to be tweaked for the better sound.

 

Remember that MAX_PRIORITY must be lower than dac irq priority in /etc/rtirq.conf

 

Thank you Piero! I will experiment with the apps list. Austinpop already tried and couldn’t say he heard an improvment in SQ. As for the MAX_PRIORITY I’ll check but used the ‘out of the box’ settings and believe the IRQ for the USB port my DDC is connected to was set to a higher value.

 

Do you think we could somehow disable the —debug switches Roon starts with? If they’re not required of course. I mentioned them in my original post you responded to.

Link to comment
3 hours ago, BigAlMc said:

 

There have been a lot of posts about using the LPS1.2 but I'm personally struggling to get the power draw low enough for my LPS1.2 to be happy with the NUC. Will try again this morning but I also have an SPS-500 which should be capable of higher power output than the LPS1.2.

 

The Paul Hynes SR4 is also highly regarded but I think there is still a wait of a couple months. Relatively much faster than the SR7 but not ideal for impatient audiophiles.

 

PS I'm going to try the SPS-500 because I have one to hand but given the consensus was that the LPS-1 was better than the SPS-500, and given how much better the LPS1.2 was to the LPS-1 I'm not sure I'd advocate buying the higher priced SPS-500 if persevering and getting the LPS1.2 to work is an option.

 

Cheers,

Alan

I have both SPS500 and LPS 1.2. The SPS500 is a beast for quality current but it can't compete with LPS 1.2 device resolution when powering sensitive source electronics.

Regards,

Dave

 

Audio system

Link to comment
7 minutes ago, davide256 said:

I've reached the end of my playing around with NUC7PJYH as endpoint . The  way to resolve the issues with it was to use ISO Regen/LPS 1.2 at which point no more glare

or mid range suck out. Using ISO Regen/LPS 1.2 with  RPI3 remained better than NUC7PJYH.

Hi Dave,

 

Shame you live in Virginia and not London! Interesting findings. I have a couple of RPi3s lying around, I may well try one in place of my NUC7PJYH; although I must say it is doing sterling work for me.

 

ATB,

 

M

Link to comment
14 minutes ago, davide256 said:

I have both SPS500 and LPS 1.2. The SPS500 is a beast for quality current but it can't compete with LPS 1.2 device resolution when powering sensitive source electronics.

 

Two topics and not sure if they are related.

 

Working on my new i7DNBE based Roon endpoint in the Plato X7 fanless case. Running 4GB RAM, no wifi or bluetooth as the board doesnt support them. Using AL headless in Extreme mode with Ram root enabled.

 

A. Everytime I restarted the NUC endpoint I need to reselect and enable it in Roon. Any ideas why it's not automatically recognising it?

 

B. I can't seem to get the power down low enough for either the LPS1.2 or my SPS-500 to work with it. The LPS1.2 flashes red as soon as I connect it. The SPS-500 (on 19V and I thought it handled like 3amps) doesn't provide enough juice for the NUC on button to do anything.

 

I've put the power settings down to 10w and burst of 12w. I've disabled all the USB slots apart from the two being used for the USB flash and connecting to the DAC. I've even switched off 3 or the cores so that only 1 core is running and switched off hyperthreading and performance boost.

 

The LPS1.2 I can understand but I thought the SPS-500 was a decent plan B as it has more power but my i7DNBE thinks otherwise. 

 

Unsure what else to try disabling in order to get it working on a cleaner PSU.

 

Any thoughts?

 

Also, as per the picture below. Should I be disabling all these HDMI endpoints that are showing up on the NUC?

 

Cheers,

Alan

 

image.thumb.png.a6191251f8ff2b4abcc7cf818ff1ead5.png

 

 

 

 

 

 

Synergistic Research Powercell UEF SE > Sonore OpticalModule (LPS-1.2 & DXP-1A5DSC) > EtherRegen (SR4T & DXP-1A5DSC) > (Sablon 2020 LAN) Innuos PhoenixNet > Muon Streaming System > Grimm Audio MU1 server > (Sablon AES) Mola Mola Tambaqui DAC > PS Audio M1200 monoblocks > Salk Sound Supercharged Songtowers

Link to comment
47 minutes ago, MrUnderhill said:

Hi Dave,

 

Shame you live in Virginia and not London! Interesting findings. I have a couple of RPi3s lying around, I may well try one in place of my NUC7PJYH; although I must say it is doing sterling work for me.

 

If you ever do, that would be interesting to have second opinion. Mildish weather in London today..?

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...