Agreed about bandwidth specification. What I was trying to get at is that the AKM graph showed AKM and ESS noise was about the same up to 50kHz. At that point ESS noise rises and at 200kHz or a little higher AKM noise starts to rise a little. They cut off the graph at 300kHz so we can' see anything after that. Does it matter? I think it might. The concern would be that HF junk that we didn't see when looking A-weighted might cause problems in other equipment such as power amps that may not adequately protected against it.
To try putting the foregoing another way, in regard to graph you linked to of AKM and ESS dacs, the noise was shown different primarily from 50kHz to 300kHz. Still don't know if it works the same way for DSD. Also, don't know if A-weighted measurements can show much in the HF band that AKM says they are proud of. We don't know what AKM DSD may be doing up there if we measure A-weighted, for example. In a similar vein, if we only care about A-weighting, then the fact the ESS gets noisy only starting at 50kHz may be completely irrelevant since it is above where A-weighting cares about.
Sorry, but for some reason I seem to be having trouble organizing what is nagging at me about what we have been discussing. Hopefully, I am starting to get more clear about it.
To try putting the foregoing another way, in regard to graph you linked to of AKM and ESS dacs, the noise was shown different primarily from 50kHz to 300kHz. Still don't know if it works the same way for DSD. Also, don't know if A-weighted measurements can show much in the HF band that AKM says they are proud of. We don't know what AKM DSD may be doing up there if we measure A-weighted, for example. In a similar vein, if we only care about A-weighting, then the fact the ESS gets noisy only starting at 50kHz may be completely irrelevant since it is above where A-weighting cares about.
Sorry, but for some reason I seem to be having trouble organizing what is nagging at me about what we have been discussing. Hopefully, I am starting to get more clear about it.
Last edited:
Right, I did make an unnecessary assumption about the blue graph, but it is kind of irrelevant to the issue about A-weighting vs at what frequencies is noise actually important. If noise is important well above A-weighted frequencies, then what bandwidth do we consider important and what noise level are we willing to say is okay? If the blue line dac has noise the same as AKM up to 50kHz, who cares above that, nobody could hear anything bad from it (if we are using A-weighting as a standard or proxy for frequency bandwidth that can cause audible problems).
In the meantime, I do take your point re trying to measure the ES9038Q2M dac noise. However, don't think that will be easy at this point. The dac is down in a shielded enclosure, and the dac outputs are held at a fixed voltage by the I/V opamps. Even trying to measure HF/RF at the I/V outputs would probably be too difficult without removing the dac from the enclosure which I am not ready to do.
In the meantime, I do take your point re trying to measure the ES9038Q2M dac noise. However, don't think that will be easy at this point. The dac is down in a shielded enclosure, and the dac outputs are held at a fixed voltage by the I/V opamps. Even trying to measure HF/RF at the I/V outputs would probably be too difficult without removing the dac from the enclosure which I am not ready to do.
what do you mean with "AKM mobile"? in real life it was quite an effort for TEAC to get SNR down to -110db for their 503 DAC as they say. according to AK specs 4490/4493 figs are -120/-123 dB
one of the best "kind of DIY" implementation of AK4490 Lynx d78 managed to get bellow -115dB within 50kHz for 0dB 24bit PCM with quite a bit filtering.
IMHO there is no question whether LPF or not, it is rather the features of it. perhaps 3rd order Bessel might be sonically the best
this one is mobile https://www.akm.com/akm/en/file/datasheet/AK4376AECB.pdf They claim -125dbA but actually, there is 116dbA. TEAC got THD+N at -110 or .0003%, it is probably not too easy for 4490 but S/N -110dbA is a trivial task even for DIYers.
Last edited:
Have to agree with eziitis. Lots of cheap dacs, including USB powered models on ebay and Aliexpress. Not much to be said in favor of competing with cheap. At a somewhat higher level, a pretty okay USB dac is expected out fairly soon that will likely be hard to compete with in terms of being good value for the sound quality.
What remains not so well understood is making very high performance general purpose dacs, which tend to remain expensive.
What remains not so well understood is making very high performance general purpose dacs, which tend to remain expensive.
Last edited:
Hi Markw4, you know, my case a bit special, as I said, I have an idea to design something like The Cyrus soundKey USB DAC transforms your mobile music but less compromising(the sounkey looks like an extremely cheap solution for profit-maximizing, £100 retail. See the link picture, even 3.5 jack isn't gold-plated, it is few cents difference!) with premium DAC and opamps, with fully balanced outputs and 4-5x times more power. In fact I gonna use only the same a dongle form factor that's it. BTW, PCM5102 used in soundkey is the 10x times cheaper than SE9038q2m but the same as AK4490 has low out of band noise and its LPF simplified up to 1st order RC. http://www.ti.com/lit/ds/symlink/pcm5102a.pdf
Less compromising isn't compromise-free at all, life is life.
Less compromising isn't compromise-free at all, life is life.
IIRC, the torrid for that has 18v secondaries. Given peak rectification of 18v RMS, it is enough. I did readjust the pots on the power supply board to give the second stage regulators a little more headroom, and also checked the adjustments for the first stage regulators to make sure they were always in regulation.
EDIT: Just double checked, but 20v RMS transformer secondaries should be safe for the 1st stage regulators. The 2nd stage are LDOs and can tolerate much less input voltage. So, if you decide you want to tweak the pots I would suggest looking at the data sheets for the regulators and making sure you understand how things need to be adjusted. If anything turns out to be unclear, you can always ask here.
EDIT: Just double checked, but 20v RMS transformer secondaries should be safe for the 1st stage regulators. The 2nd stage are LDOs and can tolerate much less input voltage. So, if you decide you want to tweak the pots I would suggest looking at the data sheets for the regulators and making sure you understand how things need to be adjusted. If anything turns out to be unclear, you can always ask here.
Last edited:
Hello to all!
Markw4 @, I want to consult with you.
One of these days I am planning to order a power filter for moding a dac. But I can not decide what is best suited without laying out large sums. I looked at your link about your filter. What do you advise from this? if it doesn't fit, say so:
1. Lusya 5 28V STUDER900 Regulator Power supply board Can assembled into double power board DIY kit A3 006-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
2. Lusya HiFi High Speed Power Supply Output Ultra Low Noise Linear Regulator Power Core Power Supply T0158-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
3. Here is an interesting option for me - Double op amp OPA2604 Ultra low noise High Precision ON MUR460 Regulated power board for DAC decoding + 5V 3.3V low voltage-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
4. LM317 LM337 + TL431 High Precision Low Noise Linear Regulated Power Supply Module Board For Amplifier -in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group a2g0s.13010208.99999999.305.adf53c00u0tMi1
Is it possible to power two TPS7A4700, each to its AVCC, from one output AC power unit ~ 9 in., After straightening? Or should each of the TPS7A4700 have a separate output from the power supply?
Markw4 @, I want to consult with you.
One of these days I am planning to order a power filter for moding a dac. But I can not decide what is best suited without laying out large sums. I looked at your link about your filter. What do you advise from this? if it doesn't fit, say so:
1. Lusya 5 28V STUDER900 Regulator Power supply board Can assembled into double power board DIY kit A3 006-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
2. Lusya HiFi High Speed Power Supply Output Ultra Low Noise Linear Regulator Power Core Power Supply T0158-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
3. Here is an interesting option for me - Double op amp OPA2604 Ultra low noise High Precision ON MUR460 Regulated power board for DAC decoding + 5V 3.3V low voltage-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
4. LM317 LM337 + TL431 High Precision Low Noise Linear Regulated Power Supply Module Board For Amplifier -in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group a2g0s.13010208.99999999.305.adf53c00u0tMi1
Is it possible to power two TPS7A4700, each to its AVCC, from one output AC power unit ~ 9 in., After straightening? Or should each of the TPS7A4700 have a separate output from the power supply?
I will add the previous post.
position number 4 - I still do not understand whether this filter can bring power + -12 / 15. Or only to + -5. If this option is really good (since the Chinese have put the scheme). That may need some kind of small modding to make + -12-15 in. at the exit??? Can you help figure it out?))
position number 4 - I still do not understand whether this filter can bring power + -12 / 15. Or only to + -5. If this option is really good (since the Chinese have put the scheme). That may need some kind of small modding to make + -12-15 in. at the exit??? Can you help figure it out?))