Review and Measurements of E1DA 9038S BAL Portable DAC & Amp

[AudioJester]

Im in, still use my original E1DA.
2.5mm

 

[IVX]

So, 2.5mm is absolutely fine, and for the low distortions too. For instance, sorry for a little showoff, one of the preproduction samples gave me THD+N -126db@1k@0dbfs at no load, and -125.2db at 32ohm! I'm simply shocked..
I can't realize what is the reason for such low THD+N at the loaded condition. Perhaps, the newest AD8397 utilises another silicon version vs the old 8397, I have a lot of these.

 

[capslock]

I think there will be a market for a 9039S or D dongle. I bet you never believed you would sell as many Cosmos ADC. Enthusiasts tend to buy things they don't really need.

With the performance you just showed, a 9039S dongle would compete with the Cosmos DAC since you could use it as a source in measurements. I would still think people wanting to do measurements would prefer the cosmos DAC because it has a real, if small, case to put on your desk and a switcheable low pass filter.

Completely different question: 9038G3 with underlined 3 is the Susumu 2000 or 3000 edition? or is there another way to tell them apart?

 

[IVX]

3000 has 3 rectangles:

 

[capslock]

Ok, so if no rectangles at all but underlined 3, it is Susumu 2000?

The reason I am asking is that there is someone selling a used G3 as Susumu 3000, and it looks exactly like my Susumu 2000.

 

[IVX]

it is not S3000, indeed.
BTW, I didn't sell a lot of ADC yet, much more I sold portable DACs.

 

[capslock]

Then you should definitely launch a 9039S, there will be enough customers, I am sure.

 

[IVX]

Who knows.. in the case of portable DACs, there are two types of customers, no-tech background with simple demands - the normal type. And tech folk, DIYers, perfectionists etc, with the same vision of the problem as I do. The normal type has a large volume capacity this is a mass demand but requires formal signs of the quality, such as no POP noise or low power consumption. They would not forgive me for such things, which I have a habit of ignoring with a smile. And vice versa, I can't rate as a Hi-Fi the DAC with no pop noise, 5mA current draw but with a hidden noise reduction or, in other words, "dynamic range enhancement". I remember the fantastic episode, some factory people showed me the PCBA of a portable DAC, with "dual cs43131" and dual outputs, balanced and unbalanced. They used the second cs43131 in fact as an inverter to get a balanced output + a noticeable phase difference )) Furthermore, to commutate these outputs, was used one mosfet, for instance, the L-channel goes directly from cs43131 to the 3.5mm jack but R-channel thru the mosfet )) You know, it's like a stupid^stupid^stupid and the power of stupid again few times. I found in taobao a similar PCBA $40 as the illustration, not exactly the same product but s^s^s is still there(200mOhm mosfets switches the outputs).

 

[capslock]

Granted, your 9038 series sells to a limited audience. But those are the folks who may already have one and shell out another $90 if the now model has yet a few dB higher performance ( would, anyway, and I would likely also buy a Cosmos DAC or DAAD). Or those who may not have an 9038 but buy a 9039 if the see a review with excellent measurements.

 

[capslock]

On a different note, I am just now listening to 7Hz Salnotes Crinacle Zero:2, my first "cabled" in-ears apart from those that cheapo things that come with phones.

They came with a cable that is 3.5 mm TRS to those 2-pole fork-like things that plug directly into the earpieces. I have no idea what they are called. So I am listening with 9038D, and the combination is pure bliss. I would like to try them with my 9038SG3. I anyone aware of a cable for that? If not, I will have to order cheap headphones and sacrifice the cable to solder my own 2.5 TRRS in place of the 3.5 mm TRS.

 

[staticV3]

I would like to try them with my 9038SG3. I anyone aware of a cable for that?

Search for "2.5mm 2pin cable" on places like Amazon, AliExpress, eBay. There are hundreds of cables to choose from

 

[Jimbob54]

On a different note, I am just now listening to 7Hz Salnotes Crinacle Zero:2, my first "cabled" in-ears apart from those that cheapo things that come with phones.

They came with a cable that is 3.5 mm TRS to those 2-pole fork-like things that plug directly into the earpieces. I have no idea what they are called. So I am listening with 9038D, and the combination is pure bliss. I would like to try them with my 9038SG3. I anyone aware of a cable for that? If not, I will have to order cheap headphones and sacrifice the cable to solder my own 2.5 TRRS in place of the 3.5 mm TRS.

The Tripowin zonie (different plugs available) is pretty decent. Around £15 from amazon.

 

[CedarX]

"dual cs43131" and dual outputs, balanced and unbalanced. They used the second cs43131 in fact as an inverter to get a balanced output + a noticeable phase difference ))

I have read about these “out of phase” issues with the CS43131/CS43198. I think Wolf measured a Zishan device showing that very issue. I have also read that the fix is fairly simple (some registers?).
It’s obviously a problem when it affects L & R channels being out of phase, though it may be inaudible in practice—it could even be seen as “wider space presentation”
But is it really an issue in a balanced (differential) implementation if it affects, say L- and R- vs. L+ and R+? It may reduce the full swing slightly, but beyond that…?

 

[Thomas savage]

My 33 ohm load is unbalanced. Don't have time to build a balanced version right now given my workload.

I hope these kind of lame excuses are a thing of the past now , jump to it Amirm, our friend Bill has already saved the world 3 times in the time you took to chose your underpants this morning ..

 

[capslock]

Thanks. 2 pin is the search word I was looking for. It's not like there is much choice here, and most options are more expensive than the excellent 7Hz Zero:2. Ended up ordering Tripowin Zombur (who's making up those names?) at about 2/3 the cost of the Zero:2.

 

[capslock]

it is not S3000, indeed.

Well, I've seen a screenshot of the delivery note from Aliexpress (from 2020) and it clearly says Susumu 3000 3000 µF. Apparently, the original Susumu edition was a 3000, and did not have those markings. I ordered my Susumu 2000 nearly a year later.

 

[IVX]

Hahah! It seems I got pop-noise free Om/Off operation of 9039S proto It's not really zero pop, rather the same as 9038S.
Why does composite opamp tend to produce a pop during power on? The reason is the cap's tolerance, even C0G has 5%.
A huge loop gain of the composite opamp(165-170db@1kHz!), amplifies a tiny transient caused by cap's capacitance difference, and we have a loud
POP on the outputs.

 

[capslock]

Did the 9038S not employ a composite op amp? Can you say what types of opamp you were / are using now?

I just recently stumbled upon this thread:

OPA828 / OPA2828 vs OPA627

By now it has been a while since OPA828 (and OPA2828) were released. Any detailed comparison done (in the context of audio) to the opamp, OPA627, that they are supposed to replace? Thanks!

www.diyaudio.com

The OPA828 was not on my radar at all and seems to outperform both the venerable OPA627 as well as all the newish TI opamps that everybody is using these days and that seem to be good already for -140 dB distortion in practical DAC and ADC applications so it is hard to believe the older (though more expensive) OPA828 should be that much better yet.

I have not taken the time to understand exactly the test method used by mason_f8 but it seemed pretty straighforward: gain of +10 non-inverting circuit with a 2 K divider resistance. And then he seems to use a classic analog distortion meter and shows the residual.

 

[IVX]

Did the 9038S not employ a composite op amp? Can you say what types of opamp you were / are using now?

all details you can find on the previous page

 

[IVX]

The way I found to prevent POP noise requires 4 mosfets to short caps around the feedback. Another way doesn't need extra mosfets but about 10x times reduces the Aol of the composite opamp.. Intriguing choice.