Glitchy thunderbolt/firewire interface - but solved with USB peripheral... what's going on?!

[mankind-nil]

I have a new-to-me MOTU 828 mk3 firewire, plugged into my laptop via a thunderbolt to firewire adapter.
Basically it all connects and seems fine but the interface is glitchy when the laptop is plugged into a PSU (seems fine on battery). Doesn't matter what driver I use. Also the glitches don't affect analog inputs routed to outputs through the MOTU DSP mixer, it only affects the computer sound. Also, all the ins and outs sound clean (apart from the occasional glitches - there's no constant noise),
I'm assuming it's some interference issue.

I have unplugged everything else to isolate the problem, and have the MOTU and laptop running from the same wall socket, so I don't think it's a ground loop issue.

But the really interesting part is that if I plug in my Native instruments Maschine (USB) to the laptop, everything works fine, the glitches disappear. Note that Maschine is not connected to anything else like MIDI, power or anything, so it cannot be providing any additional grounding. Also it doesn't matter what side of the laptop I plug Maschine into, it always seems fine (i.e. not just the port next to the thunderbolt fixes the problem).

To be clear I will usually be using this unit with Maschine plugged in anyway, but I really want to understand the issue so I can resolve it (if possible) properly.

Would ferrite rings on the thunderbolt and/or PSU cables be likely to help? Has anyone else experienced this kind of issue?
Thanks in advance!

[James Perrett]

If it is really working perfectly on battery then it certainly looks like a power issue. The first thing that I would check is the laptop's power settings. What differences are there between battery and mains settings? Could these affect the Thunderbolt connection?

If these settings are OK then I would try earthing the MOTU to a decent ground. It could be that the poor ground through the computer is causing problems and a decent ground will solve this.

[mankind-nil]

Thanks very much for your reply.
I've had a look at the advanced power settings and can't find anything obvious there, and tried different power plans/settings to no avail.
In terms of grounding, I don't know how to 'improve' the grounding, but I moved it all to another mains socket nearer the consumer unit in the house, and it made no difference.
Interestingly there does in fact seem to be a difference between the driver selected - using the ASIO drivers direct is markedly better (but still not quite right) than the Windows Audio driver - although I find it really odd that the driver would affect susceptibility to interference, unless the ASIO driver includes improved error checking or something.

Next step is to use another laptop and see what difference that makes. I've got an old MacBook in the loft somewhere...

[ajay_m]

Is the PSU a genuine manufacturer supplied unit? Note that Amazon and eBay are awash with counterfeit PSUs and I now never purchase these via those sites. So if you ever replaced the PSU the unit you're using, despite being covered with compliance logos, may be emitting significant amounts of RFI. It could also of course be faulty so my first port of call here would be to try another PSU. I also assume the thunderbolt adapter you are using is a genuine Apple adaptor (but again... Amazon... Sigh) and it might be worth swapping out the FireWire cable.

[mankind-nil]

Thanks for the reply, I have tried two OEM PSU's (both USB-C) which give basically the same result. At one point, I swapped over the PSU's and it improved for a minute and I thought I'd found the issue, but it came back - this may have been a factor of laptop charge level (i.e. if the battery is charging vs fully charged) but it would need further investigation to be sure, but in any case it is still glitching when fully charged and running off the PSU which is the condition that I will be using it in most of the time.

Thunderbolt adapters are genuine Apple. I've tried swapping firewire cables and made no difference.

But in any case, what's really baffling me is why would plugging in Maschine solve the problem? Other USB devices don't seem to have an effect. I'll try out my Traktor unit later to see if Native Instruments have some kind of noise cancelling magic in their hardware!

[James Perrett]

In terms of grounding, I don't know how to 'improve' the grounding,

I've just noticed that the 828mk3 is mains powered with a 3 pin IEC connector so it has its own ground connection. When I made my comment I was assuming that it was bus powered.

[mankind-nil]

Ah yeah, it is grounded with an IEC plug, but it was worth a try
Having said that, I haven't tried another power cable so I'll look into that tonight too...

[ajay_m]

Hmmm. Fascinating. Now I have seen issues where usb3 cables and connectors caused issues with WiFi because the switching frequencies go up to the 2.4GHz band so my immediate thought is that we may have an RF antenna situation here which is allowing standing waves to build up under certain wiring topologies. The NI unit may have very effective power bypassing via capacitors and possibly inductors which could ameliorate the problem.
It might be worth trying those clipon ferrite interference suppressors on the cables, one at either end. They are cheap and probably worth a try.

[mankind-nil]

Ok, I spent a few more hours last night trying to get to the bottom of it and am no further along. The most promising moment was when I realised there was updated firmware available for the thunderbolt ports (I'm using a Lenovo X280) and got hopeful that this would solve it - but alas, it's the same

Sadly, even on battery it seems to be glitching now (maybe it was a fluke before, or could be somehow dependent on battery/charge level).

My testing procedure was to launch NI Traktor DJ software and play a test song, running with the MOTU ASIO driver at 44.1kHz.

So the things I've tried now include:

• Different laptop USB-C chargers, grounded and ungrounded
• MOTU interface power grounded/ungrounded
• Different MOTU IEC power cables
• Different firewire cables
• Enabling/disabling flight mode
• Enabling/disabling automatic time/date check (I read this affected some people)
• Unplugging all audio connections, including headphones (required detecting glitches by watching for stuttering in the transport)
• Disabling Windows virus realtime checking
• Updating thunderbolt port firmware
• Updating all other system drivers
• Every conceivable combination of uninstalling, reinstalling MOTU drivers and restarting. At one point the MOTU driver wouldn't load until I manually went into device manager and selected 'update driver', which was weird. But I don't feel that it's a driver issue.
• Plugging in other USB devices, including NI Traktor Kontrol S2
• Testing when everything is cold, or when everything has been up and running for a while
• Opened up the interface to check for any obvious broken bits/grounding issues, but looks ok
• Checked electrical continuity of interface to ground

And still, plugging in Maschine ALWAYS fixes it!

I've just fired up my old 2011 macbook and can confirm that it all seems to be working fine there (it has a proper firewire port). So I don't feel like there is a problem with the interface itself.
I will also test another Windows laptop later this evening. Gut feeling at the moment is that it is an issue with interference at the thunderbolt port of this laptop, which is probably unfixable.
I'll see how I get on and may add some ferrite rings if I haven't lost heart by then!

[James Perrett]

Have you tried a different audio interface (or just the computer's built in sound system)? I'm wondering if it isn't an issue with your interface but the computer is actually spending too much time looking for Maschine when it isn't there?

[mankind-nil]

Good thought, but yes it's fine with inbuilt audio and through other USB audio interfaces.

[Wonks]

Glitching can often be down to clocking issues. But quite how Maschine plugged in sorts this I have no idea, unless it forces the PC to work in a way it's not doing with just the MOTU and provides a system clock signal.

The 828 isn't set to use an external clock or anything is it?

[Wonks]

Have you done a factory reset on it?

[mankind-nil]

Thanks for your thoughts, more results are in, but no difference sadly:

• External/internal clock
• Factory reset
• Enabling/disabling optical ports
• Works fine on a Macbook with native firewire

I also tried to test it on another Windows laptop, but it wasn't even recognising the Thunderbolt to Firewire connector, so this was a non-starter!

For the moment, I've ordered some ferrite rings as a last-ditch attempt, and failing that will have to move onto a USB audio interface, which will be a lot more money than the £100 I spent on this (thinking it a bargain at the time! £50 for the unit on eBay, and then about £50 on the Apple adapters). Mind you, it still works well as a digital mixer

[James Perrett]

Mind you, it still works well as a digital mixer

Probably the best way to use an old Firewire interface is to make good use of its standalone features. I'm now using my old Saffire Pro26io for its A/D and D/A convertors in conjunction with an RME Digiface USB.

[mankind-nil]

The ferrous rings arrived and I tried adding them to all connections, but they made no difference sadly
Thanks very much for everybody's help/suggestions, but it looks like this is a dead-end for this interface (with this laptop anyway). I feel a bit disappointed that I've spent so much time trying to fix the issue - of course, if it never worked at all then I probably wouldn't have tried so hard!

I'll be looking to replace it with the USB version (or more likely the hybrid Ultralite) in due course.

Cheers and all the best

[Wonks]

I’m just wondering if it’s down to the power supply for the Thunderbolt ports being too lightly loaded with just the adapter plugged in, resulting in noise from the PSU getting through and causing the glitching.

I’ve got something similar with a DIY treble booster pedal that’s fine on battery but there’s noise with a PSU plugged in. I tried two PSUs and one is definitely noisier than the other. Both PSUs came supplied with pedals - one MXR, one EHX. But occasionally I get no noise with the quieter PSU.

Switched mode PSUs need a minimum load to operate as designed and my single transistor pedal draws very little.

So I’m guessing (I certainly don’t know) that the voltage regulation on the Thunderbolt circuit might not be stretched enough to filter out PSU noise, or noise from within the computer, with just the adapter, but once you power Machine over Thunderbolt, there’s enough draw to get the regulation working properly

I’ve probably got it very wrong, but there might just be something in it. I don’t know if you’ve got anything else that draws power and that’s physically small you can plug into a Thunderbolt port to see if that cures the glitching as well.

[mankind-nil]

An interesting thought, thanks, but Maschine isn't plugged in through the Thunderbolt connection, it's on its own USB connection directly into the laptop. And it doesn't matter which USB port either, even when plugged into the USB port on the opposite side of the laptop it seems to fix the problem!

[Wonks]

The USB and Thunderbolt power supply lines are probably common.