ackoDAC based on ES9018

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

Attachments

  • AKD SYS-BUILD.JPG
    AKD SYS-BUILD.JPG
    79.3 KB · Views: 742
no, not possible and other than adapters will not ever be possible, firewire is dead in the water and will be phased out now that thunderbolt is available. do you have an old computer or something? i cant think of many computers that have firewire and not usb, any solution would end up costing as much as buying a used mac mini
 
I plan to build a 192/24 dac driven by async I2S usb interface soon. It seems as the es9012 ic is one of the best out there. As I want to use a tube I/V I fancily build I'm wondering what's the I/V resistor value range that the 9012 likes to see. Now I use 39R with the vintage Philips TDA1540 and TDA1541.
 
Trevor, yes i agree, i think he also needs a new naming convention, when the project started there was only a few different items so it wasnt so bad, but now its a bit of a mess of similar names. shoot him an email i'm sure he'll be glad to help. alternatively i'm pretty familiar with the line up, so shoot me a pm if you have a simple technical query.

@crazyfrog: sure mate, but be aware that you will not be getting even close to the performance the dac is capable of. i've used sumR tx on the output of buffalo 2 with reasonable results, but the dac performs in voltage mode and not under ideal conditions for voltage ode either. discrete solid state IV works best imo. in the d1 thread owen did comprehensive tests and passive iv was by far the worst for measurements. of course numbers arent everything, but imo they sure are something. especially when going to the trouble of using all these power supplies and tweaks on the ackodac, best to try to eek more than under 100db out of it
 
Last edited:
Trevor, yes i agree, i think he also needs a new naming convention, when the project started there was only a few different items so it wasnt so bad, but now its a bit of a mess of similar names. shoot him an email i'm sure he'll be glad to help. alternatively i'm pretty familiar with the line up, so shoot me a pm if you have a simple technical query.

@crazyfrog: sure mate, but be aware that you will not be getting even close to the performance the dac is capable of. i've used sumR tx on the output of buffalo 2 with reasonable results, but the dac performs in voltage mode and not under ideal conditions for voltage ode either. discrete solid state IV works best imo. in the d1 thread owen did comprehensive tests and passive iv was by far the worst for measurements. of course numbers arent everything, but imo they sure are something. especially when going to the trouble of using all these power supplies and tweaks on the ackodac, best to try to eek more than under 100db out of it

I do not know how Lundahl would perform with Sabre, but on AK4396 works like a charm. It doesn't measure like an opamp attached to the DAC, but it sounds much better than any opamp. Do you know what is the minimum impedance Sabre DAC allows?
Lundahl plus jFet buffer works even better. After I made buffers after Lundahl transformers, I got much better performance.
 
@crazyfrog: sure mate, but be aware that you will not be getting even close to the performance the dac is capable of. i've used sumR tx on the output of buffalo 2 with reasonable results, but the dac performs in voltage mode and not under ideal conditions for voltage ode either. discrete solid state IV works best imo. in the d1 thread owen did comprehensive tests and passive iv was by far the worst for measurements. of course numbers arent everything, but imo they sure are something. especially when going to the trouble of using all these power supplies and tweaks on the ackodac, best to try to eek more than under 100db out of it

Most people try to use the wrong transformers, like these Lundahls.
I understand, that they don't want to buy expensive transformers, not knowing if they succeed. Shunt regs from Salas seem to be an excellent choice as well and they cost nearly nothing.
If you have suitable trannies, then the numbers may still be worse, but you will not argue any more, you will listen.
 
I do not know how Lundahl would perform with Sabre, but on AK4396 works like a charm. It doesn't measure like an opamp attached to the DAC, but it sounds much better than any opamp. Do you know what is the minimum impedance Sabre DAC allows?
Lundahl plus jFet buffer works even better. After I made buffers after Lundahl transformers, I got much better performance.

exactly what i' saying, you should use something before the trannies to do iv, but not resistor, a simple differential jfet pair or something that can float its input above ground and present a low impedance. i believe the tx should be last, after the jfets and act as a buffer itself. impedance of the following stage should be as low as possible to maintain current mode, but if you accept that you will use voltage mode it should be higher than 200R or so per leg

Most people try to use the wrong transformers, like these Lundahls.
I understand, that they don't want to buy expensive transformers, not knowing if they succeed. Shunt regs from Salas seem to be an excellent choice as well and they cost nearly nothing.
If you have suitable trannies, then the numbers may still be worse, but you will not argue any more, you will listen.

i have used budP's onetics (not sumR as i wrote above, richard makes my power transformers, doesnt make signal tx) and they are very well suited, but no matter what type of transformer, it still is about the worst thing the dac could see at its output, inductive, high impedance. sound is quite good, easy on the ear, but lacking some dynamics and low level detail. i also have salas shunts, but only really have room for one for vosc in my build and its gone now due to already having plenty of heat sources in my dac. I'm using a lifepo4 battery instead, even cheaper/better. with heatsinks taken into account and the larger transformers, salas shunts are not as cheap as you make out, also with the amount of different supply voltages needed for this dac; all shunts will take up a lot of room

I subjectively prefer transformer outputs, there's an ease to the sound that's very difficult to match going the discrete route. Best is to build and listen for what you prefer - the beauty of the Sabre is that it can do both (whether it measures ideal or not).

i have done that, i have had an ackodac for more than 18 months, recently upgrading to the teflon version and buffalo for nearly 3 years now. ive tried it all, discrete solid state in the form of opc's d1 mosfet iv takes all comers for my taste, puts out nearly 45w of heat plus power supply though, thus the heat

anyway we should probably get back on topic. i agree build it and see, but i do not recommend skipping trying something other than tx if you can
 
Last edited:
Mosfets probably are interesting on the DAC output, as the gate capacitance probably provides a "free" bit of hf filtering...lol I've been toying with th Sabre and Opus for acouple of years now, too.

I might check out the teflon version, but I'm waiting to see what others offer in the forum of USB to I2S first. I like to run my sub off it's own DAC (and hence output stage), so am weighing up all options before I commit to a new build.
 
Last edited:
no losses to these ears, transconductance is so high with these particular parts that owen has measured real world 115db thd+n fro a single discrete stage with sound to match. a transformer man talking about hf loss hmmm.... haha ok lets just be happy each other is happy, i may even try out the semisouth R100 as a drop in tweak for fun.

i'm going with one teflon dac for each pair of drivers in 2 ways (already have both dacs) and then probably going with the aku-16/24 for bass when i get around to building my subs. alephjx for midbass, f5x for highs and chip/fet for sub bass.

i have titan on the way soon, but figure acko will come up with something for multichannel i2s also, hopefully with crossover
 
Last edited:
no losses to these ears, transconductance is so high with these particular parts that owen has measured real world 115db thd+n fro a single discrete stage with sound to match. a transformer man talking about hf loss hmmm.... haha ok lets just be happy each other is happy, i may even try out the semisouth R100 as a drop in tweak for fun.

i'm going with one teflon dac for each pair of drivers in 2 ways (already have both dacs) and then probably going with the aku-16/24 for bass when i get around to building my subs. alephjx for midbass, f5x for highs and chip/fet for sub bass.

i have titan on the way soon, but figure acko will come up with something for multichannel i2s also, hopefully with crossover


Chuffed to bits for you sir, that sounds like quite a complex system you have there. Using amorphous OPTs here which need some roll-off - hence why I made the comment about "free filtering" with gate capacitance ..lol
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.