The John Westlake/Lakewest MDAC/FDAC, VFET and Detox

[iansr]

The problem being that state of the art is now a considerably higher bar, than it was when John first charged us a deposit for a new DAC.

It is also considerably more affordable.

Though state of the art is the endless 'what sounds best to me' debate.

That’s true, Chi Fi has come a very long way in recent times. However, I know John believes he can do a lot better than any ESS or AKM based DAC and I’ll be very disappointed if that doesn’t turn out to be true.

 

[Tw99]

I wouldn’t disagree - I believe in science. However it doesn’t address my point of view which is that I’m by no means convinced that the usual suite of measurements that are undertaken by Amir etc, whilst important, don’t tell the WHOLE story and that there are as yet unidentified parameters that also influence what sounds good to most people. This would explain why the engineers I mention all say that the usual sus specs are important, but they only get you so far . . .

By way of example, the Purifi drivers are widely recognised as being at the leading edge of loudspeaker design. Everyone seems to love how they perform and sound. One of the reasons appears to be that they identified a particular electrical artefact as being significant for SQ which has previously been ignored by loudspeaker manufacturers. They had to build a completely new type of test equipment to be able to monitor it. Understand that I mention this purely to illustrate my point.

I remain open minded about this stuff, but that’s where I am for the time being at least.

You also have to consider that all the engineers you mention need a USP to be able to sell product, so they will always find some particular aspect of their design to talk up as being extra important, even if it isn't really.

 

[misterdog]

That’s true, Chi Fi has come a very long way in recent times.

Though the RME DAC is German and the Okto Research is Czech.

The reason most of us bought the original MDAC was because it offered great value for money. State of the art at that time was a minimum of 3/4 times the price.

The concept behind MDAC2 was an affordable step up, though that step has long been surpassed.

 

[sq225917]

Yeh, hes been squeezed out by a 10db improvement in snr, thd plus noise is the past five years and a 15-30db improvement in jitter. Dacs are now really close to the mathematical limit of the data formats they handle.

 

[Hawtin]

This just in on the facebook side:
------------------------------------------------------------

• 
  John Westlake
  Sorry everyone for the lack of posts - as you know I'm working on the FPGA for the FDAC / MDACII and I mean to post once I have some results.
  I'm at the stage with the FPGA developments where I need to measure hardware - so this past week I'm working an FPGA interface board that will interface the output of the FDAC / MDACII Modulator PCB and directly interface with the Audio measurement system so that I can start to confirm functionality and optimse performance of the FPGA.
  So far I've been passing the output into a DAC and measuring the Analogue output - this limits the Dynamic range - I'm seeing the DAC + FPGA performance rather then the FPGA alone, and I need to separate these to confirm FPGA performance...
  With a second FPGA platform I having decimate the output of the Modulators into a "PCM" type format that the audio systems can measure... I also need to be able to generate 768KHz 32 bit signals to confirm the modulators performance...
  I run into speed issue with the FDAC FPGA which had me chasing my tail for acouple of weeks and I've since change to a different (faster) FPGA family - as I've only started with FPGA programming at Christmas, its a VERY steep learning curve but things are moving forward - As soon as I've completed the FPGA System for measurement - I should be able to post measurement results.
  The Modulator board for the FDAC / MDACII will be user upgradable (recall the Dacapo Digital module) - so I'm not planning to have the very ultimate upon launch - but it has to be better then anything else so far... early "limited" testing looks good...
  I think I'll need a one more month working on the FPGA - its hard to say as I'm so new to FPGA's... but I'm happy I've got so far in so little time..
  I should be able to post first results this week.... if I mange to solve the hardware interface issues...
  The FDAC / VFET chassis design is also progressing well - but I've really had my head down since Christmas working on the FPGA development...
  I'll get the results of the FPGA measurements ASAP...
  3
• 
  
• ·
  Reply
• · 2h

 

[Stunsworth]

The MDAC 2 has been in development longer than most products’s lives.

 

[Hawtin]

A new post by JohnW on Facebook:
----------------------------------------------

Lakewest Audio
1h ·
I’ve just spent about 20 minutes typing a long update when I received a “something went wrong” message

Sorry for this shorter message now:-
I’ve manage to get the Direct digital interface from the UPD to Modulator board working (32bits / 768KHz) – and current decimating the output from the Modulator back in to the UPD at 24bits.. (when I have the energy and time I’ll upgrade this to full 32bit input).
First direct digital path results (-100dB input) showing the EXPECTED truncation error – whats important is that I now have the direct means to measure within the digital domain and I can work on resolving the truncation issues.
Ignoring the terrible truncation errors, - looking good with the modulators Noise shapers “hump” at 20KHz being -175dB down – in practice (I ‘m currently limited by the FPGA interface to the UPD) this will be 6dB lower at -181dB at 20KHz in FDAC hardware…
The final Modulator board will accept a full 32bits input…
I’ll now resolve the truncation errors and the FPGA board will be ready to design into the FDAC / MDACII hardware…
I’m look forward to posting these results ASAP – hard to say how long this will take – I’d like to say a week but….

 

[s1h1]

Thanks for the update @Hawtin

 

[Cav]

Latest from Lakewest FB page,
"@ Uwe Lorenz
I'm taken my first steps in working on a "get me going" UI for the FDAC. I'm using a very basic UI engine where I'm basically hardcoding the the UI (drawing lines, Boxes, rectangle fill etc). What would really help is some very basic UI "layout" software - basically where I can define the LCD pixel resolution - and then edit and work with each pixel.. this would give me the pixel coordinates / 565RGB colour so I can then hardcode the graphics...
Everything I've seen is based for much more advanced HTML type UI design - I need a VERY basic Pixel by Pixel UI "editor", Lines, Boxes, rectangle Fill etc....
At the moment I'm using my PCB CAD package as a basic UI layout editor (working with its Grid as pixel location, and the line drawings) ... I'm hoping theirs a better solution!

- Any ideas that will work with Windows 8 (Adobe now only supports Win10)."

 

[sq225917]

I better tell my old dell to stop running photoshop 2016 on Windows 8 now.....

Typical, ignoring the well understood widely used and supported interface tools to plough his own pixel by pixel, no one can help me approach...

Why not take some of that hundred grand and hire MIT grads to create the ui ....

 

[Cav]

And this:

Recent family situation has impacted progress over the past couple of weeks - but I've got to the stage where I need a basic "contorl" UI to allow optimization of the FDAC / MDAC FPGA modulator development.
I put off writing a graphical UI as I've zero experience in Driving a high resolution LCD display - and dont have the luxury of a powerful Linux based SoC environment (not that I want the baggage of a LInux OS operating the DAC systems).
For initial testing and development (the Graphics engine will be the same for any large LCD panel size) I'm using a 1024 x 600 panel with Capacitive Touch) - the FDAC will use a 1280x120 LCD with a custom touch panel - the FDAC front panel will basically be a full width LCD touch panel.
The pictures attached you can see the basic UI which I'm using to aid the FPGA development, and also the LCD that I plan to use on the FDAC.
That that I have a working direction for the Graphics engine, I will design a Custom graphics engine PCB for the FDAC when I design the FDAC / MDACII modulator board.
Now that I have a working "control" UI I can continue the FPGA development - I want to complete the FPGA development ASAP so I can get back to completing the hardware design.
Creating a custom UI engine with smooth animations has been a big step forward to finalising the FDAC without having to depend on others who have let me down badly.

 

[Jason_PFM]

So 8 years later, this is where we are ?

JW needing us to tell him to use MS Paint to create a user interface on his obsolete WinXP OS.

Bravo, what an accomplishment

 

[nostromo]

I'm don't really know what an FPGA is other than it's a 'CPU' type device that needs to be programmed to do tasks on a PCB.

John needs a way to programme it to control his Advanced Clock PCB design and some clever decorrelating of signal noise or some such thing.

So John's found a basic way to control and program the FPGA and give it some simple instructions so he can see what the Advanced Clock circuit can do.

So it seems he's having to do programming work which is outside his expertise. I agree that he should seek help here...and I see Uwe is trying to help out. I guess John likes to work very closely, almost iteratively with such a programmer and tinker constantly.

Personally, I rather like the UI design he's come up with, I just hope he can use it to program the FPGA to his requirements.

If anybody here has the time, perseverance and personality match to John's and can help with this please contact John on FB. I'd like to see John get back to the PCB designing.

The look and feel of the front LCD panel, I'm sure can be improved much much later after a working DAC is in hand.

 

[sq225917]

He could use something like Figma to create all the ui layouts and elements and then wire them up however he likes. That'll allow his to create the entire schema map and all the page layouts with interactions just like you would if you flowcharted them on paper.

If he had the ui mapped already it'd be trivial.

 

[Stuart Frazer]

He could use something like Figma to create all the ui layouts and elements and then wire them up however he likes. That'll allow his to create the entire schema map and all the page layouts with interactions just like you would if you flowcharted them on paper.

If he had the ui mapped already it'd be trivial.

I would have thought that mapping the user interface out on paper would be the starting point for it then to be transferred to software? I'm no expert though. To me, these are all hurdles and much of them seem to be fabricated and very unnecessary with good design, project management and support.

 

[paulski]

In the words of Papa Lazarou ‘It’s a saga now…’.

 

[sq225917]

@sf, yeh, nothing that couldn't be sorted in minutes.

 

[adamdea]

Sounds like “As usual I’m dicking around with some stuff that sounds as though it could have something to do with whatever it is I’m supposed to be working on, but is probably unconnected to anything practical. And if it is, it will be sold to someone else.”
Happy new year.

 

[moosemeat]

Sounds like “As usual I’m dicking around with some stuff that sounds as though it could have something to do with whatever it is I’m supposed to be working on, but is probably unconnected to anything practical. And if it is, it will be sold to someone else.”
Happy new year.

Thanks Adam, this gave me a good chuckle!

 

[s1h1]

They’re (you’re?) not customers, they’re (you’re?) investors or more probably speculators. Customers are people who buy an actual thing, not the promise of a thing that doesn’t actually exist.

I'm a customer - I brought a MiniStreamer. I'm still waiting on the front panel, and have been removed from the Facebook group, and had emails ignored.