MDAC first listen (part XXIV)

[JohnW]

Really John ?

I would expect the pcb for the new board to accommodate both through hole bulk foils as well as any decent cheaper alternative with its respective form factor, being SMD or through hole again. You aren't maintaining the floating rats wires on the new board, are you ?

No No No, the PCB will have positions for both the SMD footprint and Thou Hole Vishey Bulk foil...

To test and verify each board they will be first assembled with SMD parts - these will be removed to fit the Vishay Bulks.

Can't really see why you would make it difficult to have people postpone adding these expensive bulkfoil resistors at a later stage themselves.

Edit: missed the readjustment part (DC offset ?)

Yes, its an issue of DC offset calibration and testing. The design uses what we refer to as a "Crosscoupled" output stage design. As these Crosscoupled output stages are not driven via a zero ohm source (the DAC array has a 195Ohm output impedance) the stages crosscouple there DC offsets errors - making DC offset calibration very tricky - they affect each other.

The second issue is that the Match JFETS used in the frontend have a tightly Spec. VGS On, (within 20mV) but a very poorly Spec. VGS Off (I've seen differences of 500mV) - as the Jfets are used in a linear (non saturated mode) the Offset can be all over the shop. There as an "adjustment range" offered by the PCB's offset trimmers, but there range has to be limited to prevent the poor TempCo. of the trimmer pot itself causing too much DC offset drift as the unit warms up. The only solution is to add a parallel degeneration resistor on the relevant arm of the Diff pair to bring any offset error to within the adjustment range of the onboard trimmer pots... This is a manual process...

 

[JohnW]

Me too John. I've already said divert my £50 for the bypass to the PCB but for now ill take the Toy plus upgrade and you just hang on to the remaining resistors until the PCB is ready.

Iansr + EIffel, noted thank you

 

[JohnW]

Out of interest, on the new MDAC2 PCB, does it upset the output circuitry if its more towards 2.5v/5.0v balanced RMS output level?

The MDAC currently hits 3V / 6V when you use the +3dB setting, MDAC2 will be the same.

 

[JohnW]

1. I listen 90% through HP (Sennheiser HD800, Impedance @ 1kHz: 300 Ohms, Sensitivity: 102 dB).
My question is: Will the upgraded analogue output stage work in A-Class mode in the whole range (up to 0dB attenuation) when working into HD800?

With 300 ohm headset yes

2. Is there a chance to add a built-in galvanic isolation of the USB input to the MDAC2 project? If not, is there an external device that you can recommend? (E.g. out of electronicsshop.dk or lipoly.de? Both work for full/low speed only, as already mentioned by other posters here.)

Sadly no, the USB MCU is the main CPU on the design, there is no space or design time to isolate the all requires signals.

Using the MDAC2 with MTRN offers an isolated USB2.0 option.

3. Is there a chance to try to suppress the influence of the metalic case to the SQ of MDAC2? I mean something like adjusting the distance between the PCB and the case, or any other solution.

Not without a new chassis and that's not planned anytime soon! unfortunately you cannot adjust the internal height of the PCB / Top clearance of the current MDAC chassis as the PCB slides into the chassis between extruded rails + there is little internal height clearance to more the PCB higher if you could.

Producing a new chassis would cost almost as much as the new PCB - and I'm not getting involved with that ATM - otherwise I might was well just sell a complete new unit... Far more can be gained from new electronics

 

[JohnW]

Fusion case prototype.

you have been busy!!! I have some nicer vents somewhere (As used on the Peachtree designs), I'll send them to you once I come across them again

I side-step the whole issue by never having my units enclosed (Poor Renata) ... I'm always working on some new upgrade or modification...

 

[DANOFDANGER]

Not at all!!!, firstly I've only just sent out yesterday the first Fusion'ed unit to Fusion

For greatest "Punch" then Toy is the way to go, the Fusion upgrade tightens the Bass to a larger degree - if you have small box speakers with limited bottom end the fusion upgrade appears to lighten the Bass over the Toy upgrade as it removes any "overhang" to the lower end - its Bass is tight and "fast"...

That said what Fusion brings to the table can be so easily lost through the remaining signal chain....

With the toy upgrade on one of my tracks the system manages to excite and "Rattle" a support in the suspended celling above my head, while this does not happen with the Fusion unit. That said our feet are tapping away with the Fusioned unit and you are so brought into the recording - stunningly, its just so real...

If you listen and appreciate live recordings, or very well recorded studio recordings (such as the Waling Jennys) and have a high resolution system then Fusions your unit. The Fusion excels with Vocals, Stringed instruments, Piano, even the screaming crowds in the background, clearly reproducing the ambiance of the event....

If you mainly listen to commercial grade recordings or demand ceiling rattling bass from small speakers Toy's the way to go. If I'd not heard the Fusion upgrade I'd have been perfectly happy with a Toy unit as while I do miss the joist rattling Bass of Toy on my Tiny bookshelf speakers, Fusion does everything else so very right for me! I am after all an ESL speaker lover over a "Bass heavy" ported Box speaker...

I'd like to believe that the combination of a Fusion'ed unit and full range speaker systems (not the tiny bookshelf speakers I use) the Bass will be "correct"... with no overhang... Even on my speakers you can hear its tight and fast... and your feet are tapping away... then you just have to remember to breath as you have held your breath as you are brought into the very midst of the "event"...

Thanks for this, I think I know what I will choose now.

 

[dtd]

The MDAC currently hits 3V / 6V when you use the +3dB setting, MDAC2 will be the same.

IIRC, does the output quality suffers when setting output levels to +3db / 6v?

The reason I ask is that I have a pair of NCores amplifiers with gain set to approx. 13db, so I may need to adjust the gain via resistor on the amplifiers themselves to complement the MDAC's most effective 'volume range' better.

 

[JohnW]

IIRC, does the output quality suffers when setting output levels to +3db / 6v?

The reason I ask is that I have a pair of NCores amplifiers with gain set to approx. 13db, so I may need to adjust the gain via resistor on the amplifiers themselves to complement the MDAC's most effective 'volume range' better.

You would never hit the peak level of 3V / 6V output, you would be blown into the next field!

At +3dB THD peaks at 0.0025%.... I'd be worried about your hearing / speaker drivers!

 

[WAD62]

Thanks for this, I think I know what I will choose now.

...all I can add is that the 'Toy' sounds superb with floorstanders, I'm able to pick out bass lines that have been rather indistinct before, very tight, plenty of weight and not boomy at all...and as for the kick drum!!!

 

[Sober]

You would never hit the peak level of 3V / 6V output, you would be blown into the next field!

At +3dB THD peaks at 0.0025%.... I'd be worried about your hearing / speaker drivers!

Anyway, maybe it is a good idea to bring the output down to industry standard 2V/4V? Most people are having trouble with too much gain in the system nowadays rather than too little...

More great news! Yes John that will make more sense, make it "fail unit" for me if this is viable, in fact I don't mind not a new unit as long as function perfectly. Added this option to my name in wiki.

Yes, sounds like a really good way to go. Either a fail unit or a new box itself bought from Audiolab. Let us know if it is possible. If yes, I would definitely go with this project. For now, doubling the cost of the original MDAC gets a bit too expensive for me.

 

[sq225917]

John given that no one has ever complained about lack of level is there any scope for reducing the output of the DAC electrically rather than digitally. The Weiss offers selectable maximum output voltages from half a volt upwards to the industry standard.

 

[sq225917]

Yup, think about it too much level Is the largest and most frequently complaint from mdac owners.

Most of us use 29db gain amps and 89-91db/w/m, we all have level to burn.

We should list our standard listening levels then you can figure out a better potential range. Unless you're utterly constrained by the out putt stage design.

-17db is peak level for me.

 

[fusion5]

At -17db my CA 840W would blow my head off. I rarely get below -30db (speakers: 85,5 dB 1w/1m).

 

[JohnW]

Anyway, maybe it is a good idea to bring the output down to industry standard 2V/4V? Most people are having trouble with too much gain in the system nowadays rather than too little...

The gain IS 2V at 0dB (the industry standard)

Insentive headphone user crank the level up to very high levels..

We have to meet the Industry standard of 2Vrms at 0dB, the +3dB is an over-range.

 

[sam_cat]

Yup, think about it too much level Is the largest and most frequently complaint from mdac owners.

Most of us use 29db gain amps and 89-91db/w/m, we all have level to burn.

We should list our standard listening levels then you can figure out a better potential range. Unless you're utterly constrained by the out putt stage design.

-17db is peak level for me.

-15db here (Mdac Soverign dc+ -> HackerNAP amp -> Ergo IX speakers)

 

[JohnW]

John given that no one has ever complained about lack of level is there any scope for reducing the output of the DAC electrically rather than digitally. The Weiss offers selectable maximum output voltages from half a volt upwards to the industry standard.

You should never throw away hard fought Gain - the MIMP will internally scale its Feedback Gain.. this is the best use of Gain.

 

[WAD62]

Anywhere between -30dB & -15dB (detached house), depending on the time, mood, album and wine consumption...

Edit; That's roughly +3 dB over pre-upgrade MDAC, perhaps still 'burning in'

MDAC L2 Toy, 8000P & 8000SX, Mission 753 (modified, probably about 90dB)

 

[MatiasN]

JohnW,
when do you estimate the MIMP will be available for purchase?

 

[Achim]

You should never throw away hard fought Gain - the MIMP will internally scale its Feedback Gain.. this is the best use of Gain.

I am actually very often creeping around -53 to -41dBs when over speakers in a block of flats...

Couldn't you include gain scaling in the output stage of MDAC as well, rather than using the ESS volume control?

Once you start to unuse chip functions further down the MK2/TDAC road you will probably need independent volume control anyway.

 

[Sober]

Scalling amplifier feedback gain and volume control are two different things I'm afraid.

Hmm... the output stage of MDAC also probably uses feedback, so may be also possible?