I havn't said it would be easy on the way there. 😀
Completely different? You've been solving more complex challenges by now. 😉
And I havn't said you just go and switch the DAC chips. Obviously you'd have to switch the DAC HATs. What people might be able to reuse is the outputstage HAT. Or the main PS HAT.
If you'd need e.g. another 1.2V for thr DAC you could simply arrange that on the DAC HAT.
Will take it under consideration (all other stages can stay the same)
Meanwhile we are back to testing. THD+N is -102db. We want better.
I keep the fingers crossed that the next version of the isolator HAT comes with 5V supply voltage. 😉
As soon as you'll have the DC/DC supply challenge under control, applying that to the isolator shouldn't mean much of challenge anymore. 😀
As soon as you'll have the DC/DC supply challenge under control, applying that to the isolator shouldn't mean much of challenge anymore. 😀
If you use the iFi ipower smps ground the outer barrel of the dc supply to remove noise generated by leakage current. Would be interesting to see if you observe any difference with and without the earthing connection.
We can clearly see the leakage of IFI on our tests. Will check in the future for earth connection on DC
I replaced the base NDK quartz NZ2520SD on NZ2520SDA. The difference is amazing! I'm not sure that Crystek 957 will be better. 🙂
Was it complicated to replace? I'm hoping to do it soon, as have a pair of NZ2520SDA. Which Mhz parts did you use?
Wow.
1db down per day. That's what I'd call progress
For how many more days are you gonna keep tweaking?? 😀
BTW.
For sure you've seen the images of IanCanada and his transformer
outputstage HAT that'll go on top of his 9028Q2M HAT.
Looks nice to me. 😉 That'll perfectly fit my earlier expressed idea
of seeing modularity on the HAT side.
1db down per day. That's what I'd call progress
For how many more days are you gonna keep tweaking?? 😀
BTW.
For sure you've seen the images of IanCanada and his transformer
outputstage HAT that'll go on top of his 9028Q2M HAT.
Looks nice to me. 😉 That'll perfectly fit my earlier expressed idea
of seeing modularity on the HAT side.
I think that Ians took a different route with his design. In our design there is much less to tweak , everything is optimized by us. Also we are concentrating on the ess9038q2m .At last , like I said , our board is in master not slave or ASRC with latest NDK clocks.
As far as tweaking , we want to make sure that it sounds great. Thats our main focus.We just changed the capacitor in Vref to film and found a 1db decrease in 2 and 3 harmonics .
Also , we have added an "active filter" in the power supply input. 0-20Khz noise is down to 3 micro volts.
DC offset is hand trimmed (and will be same for all finished products) at about 600uV
DC/DC convertor is finished..15V is very clean.
Its coming , but we wont rush it. Every single rail has to be optimized. 75% are done.
Our AP machine seems to be out of whack on the N. EVM board of ess sabre (with a THD+N of 115db tested by ESS sabre) and Katana both test at -102db
We have been in contact with AP people.
Seems that THD is working...shows 107 and 109 on THD