Thank you for your donation!


Cloudsmith graciously provides open-source package management and distribution for our project.


Schiit Modi 3 not recognized without powered usb hub
#1
HI,
New to this RPi audio stuff. Was tearing my hair out for days trying to figure why Moode wouldn't offer my Schiit Modi 3 as a selection in the MPD settings.  After many hours of web searches I finally got a wiff of a hint (from a Schiit forum post) and installed a powered USB hub between the RPI 4B and the Modi 3 and voila! So excited to actually get sound out.

So here are my questions: 
1. Anybody else experience this?
2. The RPi 4 is rated at 1200 mA output on the USB ports (combined, though nothing else plugged in.) Why should the the modi 3 draw anything close to that?
    - I haven't measured USB current draw so no clue where that stands.
    - I thought maybe the attached 7" touch screen was perhaps loading the device but disconnected it to no effect.  
3. Is my Rpi faulty?  (I am using the official RPi 4 power supply.)
4. Any suggestions?  I'd just as soon not have the powered hub in the chain.  There are too many wires hanging off the desk as it is. :-)

Notes:
- Brand new RPi
- Using the latest stock Moode image
- Using the included power supply with the Schiit Modi 3 has no effect.  
   - Schiit says the PS is not necessary for the USB input and is only required for the coax and optical inputs; what they don't say is that the PS serves no purpose for the USB input (at least as experienced in my test.)
Reply
#2
Try the command below to see if the Pi is experiencing an under-voltage event.
https://www.raspberrypi.org/documentatio...cgencmd.md

Code:
vcgencmd get_throttled
Enjoy the Music!
moodeaudio.org | Mastodon Feed | GitHub
Reply
#3
(02-17-2021, 03:36 AM)Tim Curtis Wrote: Try the command below to see if the Pi is experiencing an under-voltage event.
https://www.raspberrypi.org/documentatio...cgencmd.md

Code:
vcgencmd get_throttled

Under-voltage does not seem to be the issue; the return value is 0 (OxO)
Reply
#4
If under-voltage can be ruled out then its not obvious to me what might be happening. I have an Allo Revolution DAC that I've connected to a 3b+ and a 4B and no issues whatsoever. I also have an older JDS Labs ODAC and no issues with that.

Maybe try swapping in another USB DAC if you have one handy and see if the issue reoccurs.
Enjoy the Music!
moodeaudio.org | Mastodon Feed | GitHub
Reply
#5
So if not under-voltage (did you measure over all operations ?) then other connection criteria may apply...

Current draw...some instantaneous events from connected hardware can cause more current draw than is acceptable.(usually USB hard-drives but other devices can produce the same effects)

Noise/interference from poorly shielded devices..


The RPI4 official supply is 'Just Enough". Other instances of a better supply requirement are known... see 'Argon One' where a more powerful supply is specified.

A cable allowing a dedicated isolated supply may be of use and it is cheap to make...
https://www.tnt-audio.com/clinica/221_diy_usb_e.html
----------
bob
Reply
#6
@DonquayHoetay

Don't forget - USB is about signals as well as about power. Schiit has its own proprietary Unison USB interface. It has caused a minor Internet storm of supporters vs haters.

I have no opinion about either the necessity for or the implementation of this interface but what you have said so far doesn't rule out the possibility that the USB signalling between Modi 3 and RPi4 isn't up to snuff. Whether or not your powered hub is supplementing current at 5 volts to the Modi 3, it is actively processing the signals going back and forth. It could be that this signal processing is making the difference in your setup.

Regards,
Kent

PS - love your user name. I hope Rocinante is well.
Reply
#7
Thanks Tim, Drone7 and OldPresbyope,

Problem solved. I swapped out the Schiit supplied USB cable for something I had on hand, the power cable from the powered hub, and, again, voila! There's the Schiit.

Is it a faulty cable or did Schiit supply no-standard cable? Yet tbd.

I appreciate your ideas. Thank you, again. They are all spot on concerns to review for any system issue. Here's what I've been thinking:
I wouldn't expect PS capacity issues; RPi.org specs the bare board active current at at 600 mA. So plenty of headroom on the 3A PS especially in this case with nothing else connected to the board. The PS would be worthless if it didn't handle the small transient loads of the USB ports. And I'm guessing that any inductive load on the RPI is going to require an interface, driver and its own PS. That said, you can't offhand rule out faulty PS, or board.

EMC/RFI? mmmm, maybe but probably not in this case; these devices all have been working separately in the same space prior. Though if suspect it can easily be evaluated by just moving the device(s) across the table or across the room, moving the electrical connections, etc.

I considered the USB hub influence so I tore apart the hub to see what was inside. Didn't have time or inclination to evaluate the circuit or devices employed. Thinking about it though. if the hub signal processing was to provide a favorable interface to the Schiit but the RPi weren't that would be just a happy coincidence for the hub as this was just some random off the shelf device. And, as another data point, the Schiit was working with my laptop.

Rocinante remains the first and foremost hack

-DQ
Reply
#8
Update:
Of four other cables in the reserves, two work with the Schiit/Pi. One from an old Blackberry and one from an old Palm Treo. Both of these were intended for power and data. Could it be the non-functional cables are data only?
Reply
#9
(02-18-2021, 04:02 AM)DonquayHoetay Wrote: Update:
Of four other cables in the reserves, two work with the Schiit/Pi.  One from an old Blackberry and one from an old Palm Treo.  Both of these were intended for power and data.  Could it be the non-functional cables are data only?
USB cables are either power and data or power only but not data only. The older cables for the Palm and Blackberry would have been made with higher spec (more expensive) cable than to-days meagre offerings.
Or.. your other cables are power only and don't do the USB handshake to configure as they don't have data ? Your multimeter should show that.

I would be considering the conductor size/rating. I have today binned a USB cable that was the root cause of my Pi2B power light flickering on and off. (Under-volt condition)

The Pi worked and played but replacing the cable with one having a greater diameter and so heftier conductors fixed the power problem.
----------
bob
Reply


Forum Jump: