Totally fine with me - just put it in it's own thread 🙂 Just keep in mind - there is a very good reason I did not already take the approach of using ordinary op-amps.
SQ wise the 9038 with mercury should be slightly better than 9028. In practice - it will be very close.
The Mecury easily supports either DAC just be changing the I/V gain resistors.
For MONO the 9028/38 is far easier to work with than the 9018. I am still working on the mono firmware (not a normal use case for me) and I will provide more detail later. DSD mono should be more straight forward than the 9018.
Cheers!
Russ
It would require custom firmware - I will create a repo for that and describe how to do it.
Nice work Russ and David. I'm planning for dual mono 9038 implementation, do you recommend BIII SE Pro 9038 over BIV for dual mono?
Regarding I/V stage, I see OPA1632 are extensively use since IVY I/V and will bring along to Mercury. I'm currently using Sonic Imagery Labs 992 discrete opamps for my 9018 (not Buffalo) and by so far it is the best sounding opamp I have used. I'm not saying anything bad with OPA1632 as it is a full differential design opamp and I'm not a pro on circuitry. My question is, is there any possibility to swap OPA6132 with any conventional opamp like the 992 with circuit modification?
Personally if I were doing dual mono right now I would likely choose the 9028. The current for a mono 9038 is extreme. Mercury does support it - but with the 9028 you could actually easily use existing IVY-III or Legato - while still having the option for the Mercury. You just have more options.
As for the opamp question. I am a very careful designer - and there is an excellent reason for using fully-symmetrical parts. You simply can't get the same kind of CMRR with conventional opamps. With ESS DACs there is a lot of CM signal you want to R. 🙂
Cheers!
Russ
Basically in stereo 9028 is a 3.2V swing across 200R while the 9038 is 50R
For mono it is half that resistance respectively.
Hi Russ, which combination will be more practical: one IVY-III for a pair of mono 9028s or a pair of IVY-IIIs dedicated to each mono 9028, respectively?
Regards,
My goodness ... That means that the 9028 outputs 16 mAs pp and the 9038 outputs 64 mAs pp in stereo mode ... ??
Cheers,
Jesper
Hello Russ, a follow-up question regarding powering the 9028 board if I may.
I run my current Buffalo 3SE in full sync mode, feeding the master clock for the board from the reclocker; that means I have disabled the B3SE's onboard oscillator by omitting the Trident that powers it. If I replicate that arrangement how much will I shave off the 750mA power supply requirement stated for the new 9028 board? Just exploring whether, with some modification to increase it's current capability, I can reuse my exisiting power supply.
Thanks ever so.
I run my current Buffalo 3SE in full sync mode, feeding the master clock for the board from the reclocker; that means I have disabled the B3SE's onboard oscillator by omitting the Trident that powers it. If I replicate that arrangement how much will I shave off the 750mA power supply requirement stated for the new 9028 board? Just exploring whether, with some modification to increase it's current capability, I can reuse my exisiting power supply.
Thanks ever so.
Not a lot - the clock actually only uses something like 20-30ma. VDD and AVCC on the 9028/38 consume the most power.
It boils down to this:
AVCC * .924 Volts out point to point.
bipolar zero output is AVCC / 2
We generally run AVCC at 3.6V
I would use a single IVY-III to start - but you could indeed use two. Why not try both and see what you like best? 🙂
Guess we have to see what the power requirements are for the Mercury IV and for the multichannel board. I am interested in the multichannel one also.
So for an 8 channel Buffalo III Pro, you would need 4 of the Mercurys for 800ma. Would that still be just one Placid HDBP Supply? Can you stack the Mercurys? Just trying to determine a rough case size to fit the DAC and all it's associated power boards and transformers.
That should be doable - I would probably use two Placid HD-BP if I had space. Also you would need a fair bit of shunt current to handle transients I would shunt something like 200-250ma per pair of Mercs.
