Well, an update on my October 20 post!

 

Here was the setup, then:

 

 

 

As you can see from the pic, the brass cylinders that form the 'motor pods' go through large holes cut in the perspex base - so they don't touch the perspex (and so don't transfer motor vibrations to it).  When the perspex base was resting on the benchtop - the motor pods rested on that too.  But lifting the perspex base off the benchtop meant that I had to provide a support for the motor pods - so I cut up a bamboo cutting board to fit across the Qubelok frame.

 

I thought this would do a good job of:

• supporting the motor pods
• yet - being bamboo - would not pass motor vibrations into the Qubelok frame.

 

However, I decided use a stethoscope on the perspex base, to check this assumption - and, unfortunately, could hear a noise when the motors were switched on.  

 

So it would seem that:

• motor vibrations were exiting the motor pods, into the bamboo shelf
• and were then being passed into the Qubelok frame
• and - even though the perspex base wasn't sitting directly on the frame ... but was resting on 4x silicone O-rings - vibrations were getting into the perpex base.  

 

So I got a mate with a drop-saw to cut up 3 pieces of ply which I happened to have lying around and glued them together ... to make an alternative support for the motor pods.  This rests on the benchtop ... not the Qubelock frame.  See pics:

 

 

 

 

 

Result:  success!  No noise through the stethoscope, when the motors are turned on!  👍

 

This suggests to me that bamboo may not be the isolation material that many people think it is!

Great work Andy! I did notice when I was last at your place that heavy footsteps was enough to make your TT skip grooves. I trust that is no longer an issue? 

7 hours ago, Keith_W said:

Great work Andy! I did notice when I was last at your place that heavy footsteps was enough to make your TT skip grooves.

 

Yes, Keith.  Changing from springs on the TT to Audio Silente 'silicone mushrooms' ... improved SQ - but left the TT vulnerable to footfalls.  

 

7 hours ago, Keith_W said:

I trust that is no longer an issue? 

 

The isolation platform solved the footfall problem!  👍

 

The latest change has stopped motor vibrations from getting into the perspex TT base.  

 

great work and a great looking turntable setup
so god to see someone enjoying that their time and effort not to mention the expense is worth it in the long run👏

might be a heap of IKEA  bamboo chopping boards listed in the classified🤣s 

4 minutes ago, ray4410 said:

might be a heap of IKEA  bamboo chopping boards listed in the classifieds🤣 

 

May well be.  

 

On 6/11/2023 at 12:51 PM, andyr said:

So I got a mate with a drop-saw to cut up 3 pieces of ply which I happened to have lying around and glued them together ... to make an alternative support for the motor pods.  This rests on the benchtop ... not the Qubelock frame.  See pics:

Just to clarify, you are resting the Qublelock and the bamboo on the ply? The reflections in the photo confusing me a bit.😉

On 6/11/2023 at 12:51 PM, andyr said:

This suggests to me that bamboo may not be the isolation material that many people think it is!

My experience after trying numerous bamboo cutting of various layouts and thickness, is that they have different impacts on the sound. The Myers’s store lady who facilitated my trial was very interested in the outcome! 😂

Some I liked, more than others and some didn’t like. Must have different resonance characteristics rather than “isolation”. 
 

I found the IKEA Apptlig butcher block, with separate bamboo layers in the block, the “best”, but now use it as part of a 3 tiered arrangement platform for “isolation” of electronic equipment. 

37 minutes ago, vivianbl said:

Just to clarify, you are resting the Qublelock and the bamboo on the ply? The reflections in the photo confusing me a bit.😉

 

Yes, sorry ... the reflections are confoosing.    The Qubelock frame (same as the Dekstand - except there is a lip front & back, to support a shelf) sits on squash balls ... which rest on the long, birch-ply shelf which supports all my equipment.

 

The cut-down bamboo shelf which I used previously ... sat on the front & back lips of the Qubelock framing.

 

Now ... the bamboo shelf has been removed and the ply block which I made up substituted - 'xept this sits on the equipment bench ... so doesn't make any contact with the Qubelock frame.

 

So the vibration path is different:

• before: motor pods sit on the bamboo shelf - which sits on the lips of the Qubelock frame.  The perspex TT base sits on the frame - so, vibrations travelled from the motor pods --> Qubelok frame --> perspex TT base.
• now: motor pods sit on the ply block - which sits on the benchtop.  The air suspension delivered by the squash balls prevents vibrations in the benchtop from getting into the perspex TT base.  👍

 

Andy

I have been going back through this thread and reading through your journey of curiosity and discovery, keeping an open mind, subjecting your creations in to your own listening and independent listener’s assessment/ judgement.
Super congratulations!

Today I was in communication with an audiophile friend re. video about the 

Linn LP12-50. That highlighted the importance of the plinth and the exacting process in constructing the compressed timber bedrock plinth. The Linn man said “ The plinth is one of most important components of an LP12”.!! Up go the prices of plinths!😱

 

I responded that maybe no plinth at all maybe better! Referred to your “skeletal Linn”!😀

12 minutes ago, vivianbl said:

The Linn man said “The plinth is one of most important components of an LP12”.!! Up go the prices of plinths!😱

 

I responded that maybe no plinth at all maybe better! Referred to your “skeletal Linn”!😀

 

Thanks, Vivian.  👍

 

Yes certainly, given that all the different woods that used to be turned into LP12 plinths by a master craftsman in the US (who unfortunately, I believe, no longer makes them   ) ... all make the LP12 sound different, it seems to me that removing the plinth (and similarly, the top-plate) - can only be a benefit!  

 

Well, today a major change was made to my system.  

 

Instead of having:

• Roon PC  ==>  miniDSP 'nanoDIGI'  (to provide XOs for my active spkrs: sub-to-woofers & woofers-to-tweeter)  ==>  3x Topping E30 DACs  ==>  6x amp channels

... I now have:

• Roon PC convolution providing:
  • the XOs  ==>  3x Topping E30 DACs  ==>  6x amp channels
  • and the time delay needed bcoz my subs are further away from my ears than the 'mains'.

So:

1. the miniDSP 'nanoDIGI' has been removed from the signal chain
2. the music signal is now upsampled to 384kHz - so the filters in the E30s are no longer in play
3. and the XO filters are all FIR - so there are now no phase changes across the XO regions.

 

Using something like 100,000 taps in total on the Roon PC, to create the XOs ... Roon is still only consuming about 1% of the CPU and 40% of the RAM.     What a powerful system, Roon is (if you know how to drive it, that is!   ).  The FIR filters IMO give a sonic delivery that feels effortless - and the lack of any phase changes across the sub/woofer crossover region ... fills out the bass.

 

Amazing result - thanks very much, Con!  👍

 

7 hours ago, andyr said:

Well, today a major change was made to my system.  

 

Instead of having:

• Roon PC  ==>  miniDSP 'nanoDIGI'  (to provide XOs for my active spkrs: sub-to-woofers & woofers-to-tweeter)  ==>  3x Topping E30 DACs  ==>  6x amp channels

... I now have:

• Roon PC convolution providing:
  • the XOs  ==>  3x Topping E30 DACs  ==>  6x amp channels
  • and the time delay needed bcoz my subs are further away from my ears than the 'mains'.

So:

1. the miniDSP 'nanoDIGI' has been removed from the signal chain
2. the music signal is now upsampled to 384kHz - so the filters in the E30s are no longer in play
3. and the XO filters are all FIR - so there are now no phase changes across the XO regions.

 

Using something like 100,000 taps in total on the Roon PC, to create the XOs ... Roon is still only consuming about 1% of the CPU and 40% of the RAM.     What a powerful system, Roon is (if you know how to drive it, that is!   ).  The FIR filters IMO give a sonic delivery that feels effortless - and the lack of any phase changes across the sub/woofer crossover region ... fills out the bass.

 

Amazing result - thanks very much, Con!  👍

 

Congrats…. Are you using Windows or Mac OS? 

if Windows, what software is in play to aggregate the DACs - and prevent clock drift?

Edited October 24, 2025Oct 24 by Grizaudio

8 hours ago, Grizaudio said:

Congrats…. Are you using Windows or Mac OS? 

if Windows, what software is in play to aggregate the DACs - and prevent clock drift?

 

Windows 10, Griz.

 

Roon itself enables the relevant DAC to be identified.

 

Currently, nothing is preventing clock drift.

 

2 hours ago, andyr said:

 

Windows 10, Griz.

 

Roon itself enables the relevant DAC to be identified.

 

Currently, nothing is preventing clock drift.

 

 

So your grouping/aggregating the DAC's via the multi zone grouping?

If this is the method Roon doesn't perfectly time align zones. There will be timing drift there too. 

Or are you using software like VB Matrix? 

 

Edited October 25, 2025Oct 25 by Grizaudio

Just now, Grizaudio said:

 

So you're grouping the DACs as rooms? 

 

 

Sorry, Griz - I have NFI!  

 

It's all magic to me - Con (@Ittaku) was the wizard.  

 

Can you expand on your "what software is in play to ... prevent clock drift?" question?  Why wouldn't Roon be doing this?

 

And it seems to me it has to be done in software (if it's necessary) - as otherwise:

• I would need 3x re-clockers - one on the USB cable to each DAC
• and how would you lock these 3 separate re-clockers together?

 

Testing whether there is clock drift or not would be really difficult because it only manifests itself after the clock drifts apart. If two DAC's have 25ppm clock drift in opposite directions, then the error is 50 samples per 1,000,000 multiplied by sampling rate. Since you are using 384kHz, then that's an error of 19.2 samples per second, or 0.05ms per second. After one minute of listening, that adds up to 3ms. After an hour, that's 180ms. 

 

To test that, you would need to set up REW to send a periodic impulse, along with an interface that has at least 2 analog inputs. I would probably use REW's Generator and Oscilloscope function and let it run for a few minutes to see if the two impulses drift apart. The difficulty is how you are going to send the signal from REW to Roon. Maybe record a wavelet burst like this and play it over 10 minutes and capture the DAC output with the interface and oscilloscope. 

 

Thanks, @Keith_W - I'll have a think about how to do it.  👍

 

Thinking about this on my way to and from the gym, this arvo ... can you tell me how the 'clocking' actually works, Keith.

 

Sure, I can understand that the clock in each DAC might 'drift' with time - but do they start from synced (ie. zero drift):

1.  when their power is turned on?

or

2.  when the USB data stream from Roon first hits the DACs?

 

If #1 then shirley an easy test would be to leave the DACs switched on for a week and then listen again?

 

As the DACs are arranged "horizontally" (L&R subs / L&R woofers / L&R tweeters), extrapolating your arithmetic, above, after a week, the DAC clocks could be 30 secs apart - so it should be easy to hear whether the tweeters were out of sync with the woofers!  

 

Hi Andy. As you know, USB DAC's can be synchronous (i.e. clock dictated by the source) or asynchronous (clocking dictated by the DAC). 

 

Now this is the point I am not sure about: whether the connection between Roon and your USB DAC is sync or async. If it's synchronous, which I highly doubt, then there will be no clock drift. If it's async, then you essentially have 3 independent clocks and they can potentially drift. 

 

The solution would be software capable of synchronous connection, e.g. VB-Matrix mentioned earlier. Or a DDC, like the RME Digiface USB. That's a bit expensive though, it's more than $1k the last time I looked. 

 

The clocks start when they receive data. 

 

I dunno, I guess the easiest way to test is to listen to music and ask yourself if it sounds weird after a really long track. I said 25ppm clock drift, that's for an average DAC. Some are better, some are worse. In the worst case scenario, you could have up to 400ms drift over an hour. That should be easily audible. So maybe turn gapless playback on, play some music for a couple of hours, then come back and listen. 

21 minutes ago, Keith_W said:

So maybe turn gapless playback on, play some music for a couple of hours, then come back and listen. 

Or a long opera on youtube? if the lip synch is correct at the start then a 500ms drift will be easily visible if voice and video go out of synch?

1 hour ago, Steff said:

Or a long opera on youtube? if the lip synch is correct at the start then a 500ms drift will be easily visible if voice and video go out of synch?

 

Great idea, Steff - but, unfortunately, for me might be difficult to do?

 

Keith:

 

2 hours ago, Keith_W said:

The clocks start when they receive data. 

 

So you are saying when the USB output from the Roon PC initially hits the 3x DACs - they are all in sync ... they will just gradually all drift apart, the longer the data stream lasts?

 

And:

 

2 hours ago, Keith_W said:

Hi Andy. As you know, USB DACs can be synchronous (i.e. clock dictated by the source) or asynchronous (clocking dictated by the DAC). 

 

Actually, I didn't know this ... but, presumably, I should be able to find out from Topping whether the E30s are synchronous ... or asynchronous?

 

2 hours ago, Keith_W said:

So maybe turn gapless playback on, play some music for a couple of hours, then come back and listen. 

 

Easy to do - this I can try!  👍

 

7 minutes ago, andyr said:

So you are saying when the USB output from the Roon PC initially hits the 3x DACs - they are all in sync ... they will just gradually all drift apart, the longer the data stream lasts?

 

That's right. 

 

 

7 minutes ago, andyr said:

Actually, I didn't know this ... but, presumably, I should be able to find out from Topping whether the E30s are synchronous ... or asynchronous?

 

I think it's a software thing, so it's going to be Roon that dictates that. 

 

Ok, right now Andy hasn’t confirmed what the actual signal path... so... 

 

If Roon is being used to link zones with the goal of running a permanent multiway speaker system, I wouldn’t recommend that approach.

No disrespect intended, but it’s really not a practical or reliable way to do it in the long run.

 

When I was looking into a basic multiway setup myself, there were a few posts on this topic on the Roon forum. 

From memory, and its a while back, Roon even said at the time "that while zones can be synced, timing can still drift due to network or processing factors.

So they are talking about RAAT/Processing/Network drift here' but I'm not sure if that’s changed over the years."

 

In my experience (as someone running a multiway system), the best and simplest solution is to use a single multichannel DAC.

It eliminates a lot of the complexity and keeps everything properly aligned.

Brands like RME, Focusrite, MOTU, and Okto all make solid options.

 

I use the Motu Ultralite MK5... Its solid. Sell your topping's and buy one. 

If you want any ideas, or to discuss aspects of what I've said - please feel free to PM direct. 

 

 

Edited October 25, 2025Oct 25 by Grizaudio

Not sure if Windows has an equivalent but on a Mac I think it's pretty simple to create an "aggregate device" which can group multiple DACs and slave them to a single clock 

Re. your first comment, in your last post, Keith - I've just switched on the system, with my DAB+ tuner playing through the Roon "Live Radio" feature (making use of 'IceCast' and 'BUTT').  Amps are off - but the digital stream is going into all 3 DACs (which are switched on).

 

I'll switch on the amps in a week ... and see how it sounds.

 

Re. your second point - Roon connects to the DACs asynchronously.

 

But if the following is how it works:

 

16 hours ago, Keith_W said:

The clocks start when they receive data. 

 

If I do find after a week that the drift is hearable - and with further experimentation, even a couple of hours is a problem ... I can at any time disable the Input to the Roon PC, to stop the data stream.  Then enable the Input again, to get the clocks synced.

 

 

 

 

 

Edited October 26, 2025Oct 26 by andyr