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

[IVX]

According my calculations that pair should produce about 120db SPL, it is fine to me.

 

[Nango]

According my calculations that pair should produce about 120db SPL, it is fine to me.

It is often not only a matter of delivering apropriate amount of voltage or current for a certain SPL, not only to move the diafragm but to accelerate, attenuate, or to sharply stop the diafragm.

 

[solderdude]

It is often not only a matter of delivering apropriate amount of voltage or current for a certain SPL, not only to move the diafragm but to accelerate, attenuate, or to sharply stop the diafragm.

Does it need to have special properties for that other than low output R and enough voltage swing ?

 

[Nango]

Yes, enough voltage to reach a certain SPL might not be the voltage/current needed to handle well the driving of a certain membran (I am thinking on an HD800) . Simple as it is. One is the amount needed to reach a certain SPL and different thing is to handle well (this is accelerate, attenuate, stop and accelerate again, and so on....) a membran which impedance swings from 300 to above 600 and back to 300 Ohm depending on which frecuency is being applied. Actually not a single frecuency but a whole bunch of frecuencies at the same time.

 

[Tks]

A lot of the cleverness in this device is due to being balanced. That produces a lot more output which makes it suitable for many headphones. Take that away, i.e. provide 3.5 mm connection, and there may not be much value to set it apart from other devices.

If the cable is removable on your headphone, see if you can find a cheap balanced cable or make it yourself.

Yeah, I'm not particularly sure why people would want this device unbalanced seeing as many like it exist. If measurements are the excuse, then it's just for the sake of spec-sheet satisfaction. I think the actual great value of this is precisely because it's balanced on top of great performance at such a size.

But as always the market demands are constantly something companies are always baffled about, and is no wonder why they decide to pander to oddities even though people who rationally think about things can't make sense of such demands' existence.

 

[mmicko]

Yes it works with Android. Whether or not it works with you're phone will depend on the (possible) power output limitations on some brands. It works like a charm with my Galaxy S8. Battery drain is quite high though (as should be expected from a device with these specs).

What cable you used for connection of 9038S to S8 ?

Based on your S8 battery drain comment, I guess that S8 is able to work directly, without Y cable and power bank ?

 

[Peter Pan]

What cable you used for connection of 9038S to S8 ?

Based on your S8 battery drain comment, I guess that S8 is able to work directly, without Y cable and power bank ?

I use a Shanling USB C to micro cable with an micro to C adapter. Not ideal but it works. I've ordered some other cables as none of my USB C to C cables works with the 9038S.

Yes, I'm using it without any external battery.

 

[solderdude]

Yes, enough voltage to reach a certain SPL might not be the voltage/current needed to handle well the driving of a certain membran (I am thinking on an HD800) . Simple as it is. One is the amount needed to reach a certain SPL and different thing is to handle well (this is accelerate, attenuate, stop and accelerate again, and so on....) a membran which impedance swings from 300 to above 600 and back to 300 Ohm depending on which frecuency is being applied. Actually not a single frecuency but a whole bunch of frecuencies at the same time.

Agreed about some people wanting more than 3.5 V .
The acceleration and stopping thing as well as impedance swing have nothing to do with it though. As long as the voltage swing is less than 3.5V (10Vpp) there is no difference between this device and a 2Watt headphone amp. There is no magic ingredient other than a higher voltage swing, unless one counts deliberately added distortion as such.

 

[IVX]

I testing the new version of the 9038D(same size but 3.5mm jack TRS) PCB and got some results. 32ohm THD=1% power is the same as S model 340mW@1kHz. THD(not THD+N!) at -3dbfs and 32ohm load .0005%(or .0003% if adding one more 1000uF to the -5V rail but no room!). S/N looks a bit better vs 9038S. 3rd PCB iteration requires to clean up rest current loops artifacts and so on.

 

[Johnb]

I testing the new version of the 9038D(same size but 3.5mm jack TRS) PCB and got some results. 32ohm THD=1% power is the same as S model 340mW@1kHz. THD(not THD+N!) at -3dbfs and 32ohm load .0005%(or .0003% if adding one more 1000uF to the -5V rail but no room!). S/N looks a bit better vs 9038S. 3rd PCB iteration requires to clean up rest current loops artifacts and so on.

You work fast!

 

[headwhacker]

I testing the new version of the 9038D(same size but 3.5mm jack TRS) PCB and got some results. 32ohm THD=1% power is the same as S model 340mW@1kHz. THD(not THD+N!) at -3dbfs and 32ohm load .0005%(or .0003% if adding one more 1000uF to the -5V rail but no room!). S/N looks a bit better vs 9038S. 3rd PCB iteration requires to clean up rest current loops artifacts and so on.

Awesome work

 

[letrungdat]

I testing the new version of the 9038D(same size but 3.5mm jack TRS) PCB and got some results. 32ohm THD=1% power is the same as S model 340mW@1kHz. THD(not THD+N!) at -3dbfs and 32ohm load .0005%(or .0003% if adding one more 1000uF to the -5V rail but no room!). S/N looks a bit better vs 9038S. 3rd PCB iteration requires to clean up rest current loops artifacts and so on.

Awesome. I have so much headphone with unblanced cable, you save me!

 

[Labjr]

No DSD decoding in specs?

 

[Veri]

No DSD decoding in specs?

It has a CMedia receiver not supporting that AFAIK

 

[MC_RME]

I testing the new version of the 9038D(same size but 3.5mm jack TRS) PCB and got some results. 32ohm THD=1% power is the same as S model 340mW@1kHz. THD(not THD+N!) at -3dbfs and 32ohm load .0005%(or .0003% if adding one more 1000uF to the -5V rail but no room!). S/N looks a bit better vs 9038S. 3rd PCB iteration requires to clean up rest current loops artifacts and so on.

I was looking for something like this for years for my iPhones. The EU iPhones are limited to a max output voltage of around -3.5 dBu, nearly 6 dB lower than US. These 6 dB are missing when you watch movies with a constant low volume, or if you use EQ for raising bass, because to prevent clipping one has to reduce the overall volume by the amount of gain - and volume is too low again.

So for me personally the 3.5 mm TRS version with such tech specs (indicating +12 dBu output level at 32 Ohm without audible noise playing silence) is a dream come true! The only question that remains is if my iPhone's lightning port is able to power your external amp reliably. Did you test that?

 

[Veri]

@IVX pretty cool when RME is praising your device
I hope the 3.5mm will deliver!

 

[IVX]

Oh, I tested the soft-start feature right now and even my Android starts normally only one time of 10 tries. I using P-mosfet 50mOhm with 10ohm in parallel i.e. no way for inrush >500mA is possible. Anyway, I see only about 1v at VB rail of USB in 90% of tries. As all of us know iOS USB currents even lower, I can't read USB flash with my iPad without Y-splitter, which Apple calls "Camera adapter".

 

[IVX]

No DSD for a few reasons:

1) I don't like the idea to add 2 ugly buttons for volume control

2) XU208 which I tried in my proto is quite costly for my price range, CM6642 5x times cheaper, 5x6mm and needs just 3 surrounding 0402 parts.

3) Almost nobody needs DSD seriously but only like a feature in the feature list. Seems our little friends marketing-guy realized that and offered lossy-BS MQA for the same role. I already got a few requests for future about MQA hardware supporting, and I ask too - do you need mp3, AAC, APTx HW-decoder as well? ))

4) If I got level#9 in the Harman's "How to listen" and hear no difference DSD vs DSD2PCM, who can objectively hear it, cats, aliens?
Who loves music quite happy with PCM 24/192, and can't hear the difference vs 16/44.1.

 

[bidn]

No DSD for a few reasons:

1) I don't like the idea to add 2 ugly buttons for volume control

2) XU208 which I tried in my proto is quite costly for my price range, CM6642 5x times cheaper, 5x6mm and needs just 3 surrounding 0402 parts.

3) Almost nobody needs DSD seriously but only like a feature in the feature list. Seems our little friends marketing-guy realized that and offered lossy-BS MQA for the same role. I already got a few requests for future about MQA hardware supporting, and I ask too - do you need mp3, AAC, APTx HW-decoder as well? ))

4) If I got level#9 in the Harman's "How to listen" and hear no difference DSD vs DSD2PCM, who can objectively hear it, cats, aliens?
Who loves music quite happy with PCM 24/192, and can't hear the difference vs 16/44.1.

Hi IVX,

thank you for your thoughtful and detailed answers.

I too don't believe in DSD (I don't see i t as bringing any benefits and is indeed less convenient re. volume control)
and I am very strongly against MQA ( I see it no: it is, contrary to the usual marketing buzz, actually lossy ⇒ quite nefarious backwards trend vs lossless formats; proprietary ⇒ again quite nefarious backwards trend vs free, open formats ; and last terrible thing for me: introduces in disguise DRM..., making it loved though among content-distribution businesses...).

However I would like, and can imagine also many potential customers who want to directly connect IEMs or headphones to your devices (so not though a separate amp), a wheel or at least buttons to control the volume.

All the best with your technically excellent projects,
bidn

 

[AudioSceptic]

I was looking for something like this for years for my iPhones. The EU iPhones are limited to a max output voltage of around -3.5 dBu, nearly 6 dB lower than US. These 6 dB are missing when you watch movies with a constant low volume, or if you use EQ for raising bass, because to prevent clipping one has to reduce the overall volume by the amount of gain - and volume is too low again.

So for me personally the 3.5 mm TRS version with such tech specs (indicating +12 dBu output level at 32 Ohm without audible noise playing silence) is a dream come true! The only question that remains is if my iPhone's lightning port is able to power your external amp reliably. Did you test that?

This applies to me too, but I'm wondering about the connections. Is there a reliable Lightning-to-USB-C adapter/cable, or will we need to use the rather pricey Apple Camera Adapter and then a USB-to-USB-C adapter?