Hi E1, I am very interested in your PowerDAC, as I have been looking for a USB "dongle" DAC/Amp with a differential output. What is your plan for this product?
Also, just a gentle comment about your message formatting. It would be very useful for "readability" if you broke your messages into paragraphs.
Also, just a gentle comment about your message formatting. It would be very useful for "readability" if you broke your messages into paragraphs.
MrSlim, actually, PowerDAC isn't dongle DAC/AMP. I've checked your diyaudio.com activity, and I guess you interested in ES9038Q2M based model(I named that #9038S), right?
Well, currently #9038S housing isn't finished yet, I got 50m extruded aluminum profile, and gonna send that to CNC/sand_blast/white_anodize/gray_silk. Next, I'll arrange punch tooling for caps + black anodize, and I'll be ready to assemble few samples for youtube review. I hope all those steps will be done for 1 month, however, my timelines are always delayed because I'm just a single engineer, not a factory, not salesman, not a manager, and not a writer as you noticed.
Today I going to rent AP SYS2702(APx555 didn't found yet) to get more officially looking measurements of #9038S. To be honest, I bit afraid to run this model because it is not cheap enough to offer as E1DA no name. I could keep the retail price low if my PO would be 5-10KU but I don't think if it is a realistic prospect. BTW, I've some XL version of ES9038Q2M+HPA proto, so #9038X with XMOS USB bridge 32/384 and DSD 11MHz, with more optimal DAC's power quality etc. The PCBA could be fitted into the PowerDAC case(72x46x17mm) but in the USB dongle(48x21x9mm) no way, and too expensive for noname, IMO.
Well, currently #9038S housing isn't finished yet, I got 50m extruded aluminum profile, and gonna send that to CNC/sand_blast/white_anodize/gray_silk. Next, I'll arrange punch tooling for caps + black anodize, and I'll be ready to assemble few samples for youtube review. I hope all those steps will be done for 1 month, however, my timelines are always delayed because I'm just a single engineer, not a factory, not salesman, not a manager, and not a writer as you noticed.
Today I going to rent AP SYS2702(APx555 didn't found yet) to get more officially looking measurements of #9038S. To be honest, I bit afraid to run this model because it is not cheap enough to offer as E1DA no name. I could keep the retail price low if my PO would be 5-10KU but I don't think if it is a realistic prospect. BTW, I've some XL version of ES9038Q2M+HPA proto, so #9038X with XMOS USB bridge 32/384 and DSD 11MHz, with more optimal DAC's power quality etc. The PCBA could be fitted into the PowerDAC case(72x46x17mm) but in the USB dongle(48x21x9mm) no way, and too expensive for noname, IMO.
Thoughts about no-name audio stuff nowadays. I'm not sure if somebody will understand what I gonna say about but..
Let's imagine the past, about 20 years ago with big and serious audio names on the market. Sony, Philips etc with large scientific background, university programs, and olimpic games sponsoring.
Who was a no-name there? Correct - something tiny, low tech and cheap. Next 20 years the main business formula was dramatically changed, no science, no research, brands just involve outsourcers to design something or simply buying renamed products from chinse OEMs(what kinda farmers made that "design", you probably not able to imagine) and sometimes meet their own brand named stuff in their own booth on exhibitions just a day earlier than consumers(no joke!).
What about the performance of the products? Well, many USB dongle-like DACs even not specify the output power or THD, S/N etc )) Color, price and "true high-fidelity/UNBELIEVABLE SOUND QUALITY" that's it, seems enough to sell. It is not too easy but sometimes you can download a manual and find there S/N ratio -90db or 40mW@32ohm and THD .1% i.e. smartphone with 3.5 jack may kill that effortless. It is purely shameful because of DAC/HPA area is completely trivial technically, and to get promised by DAC's manufacturer specs can everyone who did implement datasheet without silly mistakes. How do they get awards with 90db dynamic range i.e. half of CD audio? In my opinion, audio industry is highly discredited today and I rather believe if some DIYer will get a great result than a manager with a colorful product and silly clickbait like "UNBELIEVABLE SOUND QUALITY". The end of self-promotion
Let's imagine the past, about 20 years ago with big and serious audio names on the market. Sony, Philips etc with large scientific background, university programs, and olimpic games sponsoring.
Who was a no-name there? Correct - something tiny, low tech and cheap. Next 20 years the main business formula was dramatically changed, no science, no research, brands just involve outsourcers to design something or simply buying renamed products from chinse OEMs(what kinda farmers made that "design", you probably not able to imagine) and sometimes meet their own brand named stuff in their own booth on exhibitions just a day earlier than consumers(no joke!).
What about the performance of the products? Well, many USB dongle-like DACs even not specify the output power or THD, S/N etc )) Color, price and "true high-fidelity/UNBELIEVABLE SOUND QUALITY" that's it, seems enough to sell. It is not too easy but sometimes you can download a manual and find there S/N ratio -90db or 40mW@32ohm and THD .1% i.e. smartphone with 3.5 jack may kill that effortless. It is purely shameful because of DAC/HPA area is completely trivial technically, and to get promised by DAC's manufacturer specs can everyone who did implement datasheet without silly mistakes. How do they get awards with 90db dynamic range i.e. half of CD audio? In my opinion, audio industry is highly discredited today and I rather believe if some DIYer will get a great result than a manager with a colorful product and silly clickbait like "UNBELIEVABLE SOUND QUALITY". The end of self-promotion
Today I started to measure #9038S with AP SYS2702 and found that S/N ratio is worse than I premeasured for 2.7-3db, so S/N =-122.2dbA. Yep, not so WOW as -125dbA but still cool! THD+N @1kHz -3dbfs is the same as premeasured: .0005% for Normal Mode, and .00043% for Performance Mode, and .25% for SE Mode(almost only 2nd harmonic presented). However, at lower levels, I see a big difference, and I gonna investigate the reason of that carefully.
Ok, so maybe I was being too "specific" about Dac/Amp Dongle. I'm not particularly concerned about how big it is really. And perhaps I shouldn't have called what I was interested in a DAC/Amp either. I am very interested in your Power DAC ideas, and you had mentioned that you had a few of the V1 version boards made up back in Sept. Did you do the same for the V2? The DSP/equalizer setup via BLE sounds very interesting as well.
MrSlim, sure, I have few samples but still waiting for correct housing parts. Last week I got steel caps but with the wrongly positioned hole for the LED(BTW, nice one, exactly like tube's cathode 1100K see attached). I over 12 years here and quickly gotcha the top mainland-chinese secret - they almost never do anything in time and correct, whatever it is punch tooling or car driving. So, PowerDAC V2 sample for youtube reviewer is delayed and I look stupid together with my little friends from the punch factory. BTW, they agree that it is their mistake, and ready to redo tooling "for free" but.. they anyway asked to pay me extra, kinda my order is too complex ))
Attachments
In post #73 the 3rd harmonic is only 75 dB down. Are you satisfied with this number?
And https://www.diyaudio.com/forums/att...92d1555763390-true-digital-3-32vrms_16ohm-png
show high numers... these are not commented really - why did you show these?
//
And https://www.diyaudio.com/forums/att...92d1555763390-true-digital-3-32vrms_16ohm-png
show high numers... these are not commented really - why did you show these?
//
1) your Fidelio L1 has no balance 4-wire interface, right? PowerDAC V2 has 2mm jack 4 poles for balance tied headphones
2) The only planars have flat impedance curve vs frequency, for dynamic one I can't say anything else, only try
3) if don't take into account 2 previous points, the THD should be quite low .006% the same for 1kHZ or 6.63kHz but low-frequency THD higher - need to pay for simplicity of power rail conditioning
2) The only planars have flat impedance curve vs frequency, for dynamic one I can't say anything else, only try
3) if don't take into account 2 previous points, the THD should be quite low .006% the same for 1kHZ or 6.63kHz but low-frequency THD higher - need to pay for simplicity of power rail conditioning
Last edited:
well, I only can notice that sound is like a razor detailed, and I noticed that most of listeners really prefer Harman response curve for EQ. The second model #9038S is crazy more precise in a formal specs (after individual tuning @1kHz test tone harmonics are under -120db level!) but I'm not sure if I hear the big difference there. PowerDAC has 1.5ohm output impedance @1kHz so damping factor is less than #9038S, and as suppose to be, PowerDAC bass a bit more punchy.
#9038S log: formal specs measurement with Audio Precision SYS2702 analyzer(residual THD+N about .00015-.00018% in 10-22000Hz BW) which surely has not enough resolution to measure #9038S distortions. I want to show here THD vs power plots for different resistors load(as I remember I already showed that plot for actual HFM HE400i load), and ask you guys, if anyone saw before comparable product on the market? I quite new in that area, and #9038S is the second my model for HP world, so I'm not 100.00% sure about the specs is unique, rather 99%
1) 110ohm load i.e. LCD3 equivalent, 1kHz max power@THD+N=1% 104.6mW i.e. LCD3 with sensitivity 102db/mW will produce >122db SPL with #9038S output THD+N .00033% A-weighted or .0004% unweighted. The blue one curve is SE-mode with lonely 2nd harmonic. NSR -122dbA with PC USB or -123dbA with OTG USB Android/iPhone
2) 70ohm load i.e. LCD2 equivalent, 1kHz max power@THD+N=1% 164.3mW i.e. LCD2 with sensitivity 101db/mW will produce >123db SPL with #9038S output THD+N the same as LCD3
3) THD+N vs frequency plot is pretty much redundant because distortions are the same across the audio-band, but anyway, check this out the rare picture for that market
4) 32ohm load i.e. close to my HFM HE400i, 1kHz max power@THD+N=1% 343mW i.e. HFM HE400i with sensitivity 93db/mW will produce >118db SPL with #9038S output THD+N@half of max power again about same
5) 16ohm load, not sure which HP is close to 16ohm yet, let's take LCDX 20ohm as a close example, 1kHz max power@THD+N=1% 552.9mW i.e. LCDX(if accidentally it was -20% less impedance) with sensitivity 103db/mW will produce >130db SPL(DANGEROUS LOUD!!!) with #9038S output THD+N at 1/2 of max power .0009%
And another quite formal plot the frequency response, 7Hz-85kHz at -.2db or 8Hz-75kHz at -.1db
1) 110ohm load i.e. LCD3 equivalent, 1kHz max power@THD+N=1% 104.6mW i.e. LCD3 with sensitivity 102db/mW will produce >122db SPL with #9038S output THD+N .00033% A-weighted or .0004% unweighted. The blue one curve is SE-mode with lonely 2nd harmonic. NSR -122dbA with PC USB or -123dbA with OTG USB Android/iPhone
2) 70ohm load i.e. LCD2 equivalent, 1kHz max power@THD+N=1% 164.3mW i.e. LCD2 with sensitivity 101db/mW will produce >123db SPL with #9038S output THD+N the same as LCD3
3) THD+N vs frequency plot is pretty much redundant because distortions are the same across the audio-band, but anyway, check this out the rare picture for that market
4) 32ohm load i.e. close to my HFM HE400i, 1kHz max power@THD+N=1% 343mW i.e. HFM HE400i with sensitivity 93db/mW will produce >118db SPL with #9038S output THD+N@half of max power again about same
5) 16ohm load, not sure which HP is close to 16ohm yet, let's take LCDX 20ohm as a close example, 1kHz max power@THD+N=1% 552.9mW i.e. LCDX(if accidentally it was -20% less impedance) with sensitivity 103db/mW will produce >130db SPL(DANGEROUS LOUD!!!) with #9038S output THD+N at 1/2 of max power .0009%
And another quite formal plot the frequency response, 7Hz-85kHz at -.2db or 8Hz-75kHz at -.1db
Attachments
Oh, I had this morning some conversation with Amir, it was regarding the difference his old AP machine(pretty much the same which I borrowed) with new Apx555.
But you know, I can see harmonics below -120db level with my handcrafted ADC+notch filter, I tested that with AP 2702 generator which is way more precise than its analyzer, and AP guaranteed <-130db harmonics of the generator. I can see them as -125-127db, so I believe, I equipped enough to turn over the market Actually, even Amir's APx555 read the THD+N =.00013% when SNR =-131dbA and THD .00003% of that toy Okto Research Actually, after -90dbA SNR and 50mW(THD+N isn't specified at all) for competitors USB->DAC->HPA dongle, #9038S could be 10x times worse than now to be comparable ))
But you know, I can see harmonics below -120db level with my handcrafted ADC+notch filter, I tested that with AP 2702 generator which is way more precise than its analyzer, and AP guaranteed <-130db harmonics of the generator. I can see them as -125-127db, so I believe, I equipped enough to turn over the market
Actually, even Amir's APx555 read the THD+N =.00013% when SNR =-131dbA and THD .00003% of that toy Okto Research Actually, after -90dbA SNR and 50mW(THD+N isn't specified at all) for competitors USB->DAC->HPA dongle, #9038S could be 10x times worse than now to be comparable ))
The very first sample of PowerDAC V2 has fully assembled
laser print on polished stainless steel looks fantastic, like extraterrestrial tech on my hand!
However, it is expensive and still made not flawless, and I consider a fine-mate sandblast(like iphone's back) variant further. Tomorrow I'll send it to the USA for review.
laser print on polished stainless steel looks fantastic, like extraterrestrial tech on my hand! However, it is expensive and still made not flawless, and I consider a fine-mate sandblast(like iphone's back) variant further. Tomorrow I'll send it to the USA for review.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.