USB-C to 3.5mm adapters (DAC) - Let's find the best

Jan 17, 2022 at 12:12 PM Post #3,436 of 6,968
Power expressed in mW for different loads (impedance/resistance) actually effectively includes voltage capabilities of the source.
Not always understandable by users.
For high Z headphones the amplifier should ideally be a voltage amplifier to provide higher voltage gain and not a current amplifier which provides higher current.
The dongles are optimized as current amplifiers in lower impedance values.
They are calibrated to usually have the best performance at values between 32-50Ω.
 
Jan 17, 2022 at 1:46 PM Post #3,438 of 6,968
Hey good news methinks. I saw that AKM is restarting production using other fabs. Even better new product is coming out. I never thought of myself of a fan for any particular chip manufacturer before but now I realize that I like the AKM house sound.
That's good news indeed. On the subject of preference I must admit I am seeing it the same way as you do. Last year I was a staunch fan of ESS Sabre DACs, And yes tuning matters more as the same ES9038Q2M can sound differently depending on the implementation. But in the grander scheme of things, I have found that my taste seems to favor the native "old school" feel of AKM, a very subtle thing. ESS Sabres in native signature can be outright bright and edgy, AKM on the other hand are natively smoother. At least that's what I have observed
 
Jan 17, 2022 at 2:17 PM Post #3,439 of 6,968
Not always understandable by users.
For high Z headphones the amplifier should ideally be a voltage amplifier to provide higher voltage gain and not a current amplifier which provides higher current.
The dongles are optimized as current amplifiers in lower impedance values.
They are calibrated to usually have the best performance at values between 32-50Ω.
My point is that many things that you write may be less understandable for users.
For instance, "calibrated", do you mean "designed" or "engineered"?
Then, more importantly, I hope not "32-50 Ohm", I would not buy any of those, but 16 Ohm and higher, and ideally 8 Ohm and higher for IEMs. The latter sources are exactly designed to provide higher currents for those IEMs with the impedance below 16 Ohm that can be crazily source dependent, like Andromeda, because of the requirements of high currents and related fast modulations of high currents to reproduce fast transients.
 
Jan 17, 2022 at 3:00 PM Post #3,440 of 6,968
My point is that many things that you write may be less understandable for users.
For instance, "calibrated", do you mean "designed" or "engineered"?
Then, more importantly, I hope not "32-50 Ohm", I would not buy any of those, but 16 Ohm and higher, and ideally 8 Ohm and higher for IEMs. The latter sources are exactly designed to provide higher currents for those IEMs with the impedance below 16 Ohm that can be crazily source dependent, like Andromeda, because of the requirements of high currents and related fast modulations of high currents to reproduce fast transients.
I am not native English speaking especially using technical terminology so please excuse any phrasing confusion.
Well I mean engineered.
Now please keep in mind that NOT all USB DAC dongles are able to provide high current flow for loads lower than 32Ω.
A lot of them (even expensive stuff like the THX Onyx) when used with loads from 22Ω and lower will trigger internal protection and limit the output current.
 
Jan 18, 2022 at 7:20 AM Post #3,443 of 6,968
Yesterday I spent some time comparing my Meizu Hi-Fi + Yinyoo_ 3.5 mm cable against the TempoTec Sonata E44 + Yinyoo_ 4.4 mm cable. Both dongles have excellent measurements, same CS43131 DAC chip (1 chip versus 2 though), and same cable material (single ended versus balanced though).

The E44 brings noticeable more natural sound and better definition even with a mid-range IEM like the Moondrop Aria. Midrange more resolving, bass less cluttered. Not huge differences, but to me it is well worth the price. The only single reason I would choose the Meizu while on the go is because it is really small, like a cable extension, while the E44 is a dongle itself, not large or heavy, but still.
 
Jan 18, 2022 at 11:56 AM Post #3,444 of 6,968
I have a DT880 600 Ohm, one of the main criteria I extensively tested for #donglemadness is how capable they are driving stubborn magnetic planars and up to 600 Ohm dynamic drivers. Assuredly, CEntrance DACport HD (4.1 Vrms, 775 mW) and Apogee Groove (5 Vrms) being the most competent driving DT880 600 Ohm, at approx 90% fidelity as compared to when driven with iFi ZEN Stack of ZEN DAC V2 + ZEN Can (15.1 Vrms). Next in line are Cayin RU6, TempoTec Sonata E44 and Ovidius B1. These dongles has proven excellent driving power with that DT880. In fact all of the 5 Stars performers of #donglemadness has passed this test, you can check my article on them here:
https://andyaudiovault.com/donglemadness/
There's something I'm not understanding, in the Fiio KA3 review you say only power that's preventing it from getting 5 stars, but Sonata E44 is significantly less powerful under loads up to 300ohm (when they're equal pushing 4V). Already @68ohm the KA3 gets 4V while E44 is about 2.7V.
And then there's this statement about KA3 cons: "Weak for a 2 Vrms rated dongle", it's rated for 2.7V only @32ohm (while E44 is the one pushing 2V), it gets up to 4V under higher loads.
 
Jan 18, 2022 at 12:41 PM Post #3,445 of 6,968
There's something I'm not understanding, in the Fiio KA3 review you say only power that's preventing it from getting 5 stars, but Sonata E44 is significantly less powerful under loads up to 300ohm (when they're equal pushing 4V). Already @68ohm the KA3 gets 4V while E44 is about 2.7V.
And then there's this statement about KA3 cons: "Weak for a 2 Vrms rated dongle", it's rated for 2.7V only @32ohm (while E44 is the one pushing 2V), it gets up to 4V under higher loads.
Numbers don't tell all the story. E44 in real life tests has proven to be exceedingly powerful to drive very stubborn stuffs. And I am not just factoring in loudness levels, I factor in the richness of dynamics density and stability when subjected to tough loads. That is why I never judge anything purely on measurements. Take Luxury & Precision W2 for example, advertised at 250 mW and with dual discreet circuitry, but the fact is W2 wasn't as impressive as E44 when subjected to driving a stubborn magnetic planar like Fostex T40RP MK3. Don't just read the specs, do actual A/B comparisons and perhaps you can make sense of what I am saying
 
Jan 18, 2022 at 12:48 PM Post #3,446 of 6,968
Numbers don't tell all the story. E44 in real life tests has proven to be exceedingly powerful to drive very stubborn stuffs. And I am not just factoring in loudness levels, I factor in the richness of dynamics density and stability when subjected to tough loads. That is why I never judge anything purely on measurements. Take Luxury & Precision W2 for example, advertised at 250 mW and with dual discreet circuitry, but the fact is W2 wasn't as impressive as E44 when subjected to driving a stubborn magnetic planar like Fostex T40RP MK3. Don't just read the specs, do actual A/B comparisons and perhaps you can make sense of what I am saying
Power delivery isn't linear under lower impedances, there's current limits, the voltage can't be pushed without clipping. You're basing the score over a fringe case when more than 90% of use cases for dongles won't be 600ohm headphones. This weird criteria can compromise all the great work you did on subjective assessments.
Your reader searching for a dongle to power his Sundara will be misled into thinking the E44 is more powerful than KA3 when the opposite is true.
At least I hope you fix the wrong statement about KA3 voltage rating.
 
Jan 18, 2022 at 1:00 PM Post #3,447 of 6,968
Power delivery isn't linear under lower impedances, there's current limits, the voltage can't be pushed without clipping. You're basing the score over a fringe case when more than 90% of use cases for dongles won't be 600ohm headphones. This weird criteria can compromise all the great work you did on subjective assessments.
Your reader searching for a dongle to power his Sundara will be misled into thinking the E44 is more powerful than KA3 when the opposite is true.
At least I hope you fix the wrong statement about KA3 voltage rating.
The question that you have not answered if you have done actual tests like I did? Like doing lengthy A/B. From what I have heard from other users owning E44, it is consistent with my findings. So do you actually have an E44 in hand that you have compared extensively with KA3 side by side driving a 91db magnetic planar or 600 Ohm DT880? wrong statement? I never make assumptions. I only say what I heard and have experienced with
 
Jan 18, 2022 at 1:15 PM Post #3,448 of 6,968
The question that you have not answered if you have done actual tests like I did? Like doing lengthy A/B. From what I have heard from other users owning E44, it is consistent with my findings. So do you actually have an E44 in hand that you have compared extensively with KA3 side by side driving a 91db magnetic planar or 600 Ohm DT880? wrong statement? I never make assumptions. I only say what I heard and have experienced with
"2V rated dongle" is a objectively wrong statement in your KA3 review and should be corrected if you care about accuracy at all.
My opinion about power is the same as Crinacle: https://crinacle.com/2019/09/30/on-the-record-power/
So while I did like your subjective impressions I don't think your assessments about power have any value now knowing how you think.
 
Jan 18, 2022 at 2:53 PM Post #3,449 of 6,968
"2V rated dongle" is a objectively wrong statement in your KA3 review and should be corrected if you care about accuracy at all.
My opinion about power is the same as Crinacle: https://crinacle.com/2019/09/30/on-the-record-power/
So while I did like your subjective impressions I don't think your assessments about power have any value now knowing how you think.
At which point am I wrong with saying KA3 is a 2 Vrms dongle? FiiO own specifications clearly indicated it is 2 Vrms and up to 2.7 Vrms on BAL:
1642534969310.png

https://www.fiio.com/ka3_parameters
Yes it can go up to 2.7 Vrms and that is still a region of 2 no?

And have you checked Tempotec E44 specifications? clearly it is a 4 Vrms device which makes is double the powerful than KA3 even on paper
1642535178269.png


And just because in my summary I said KA3 is a weak 2 Vrms dongle and not specifically mention 2.7 Vrms, now the the accuracy is questionable?
Don't take my words for it, your opinion seems already set on what you want to believe, get yourself an actual E44 and KA3 side by side and compare, while at it get also Questyle M12 and Ovidius B1 - both of which prime examples of 2 Vrms dongles, test them, and see if you are hearing which ones has the better drive for magnetic planars, we can discuss all day long on assumptions and theories but nothing beats the actual experience.

It is not my interest to convince anyone, that is not why I am doing this. I did what I did because it is fun.
 
Jan 18, 2022 at 3:51 PM Post #3,450 of 6,968
At which point am I wrong with saying KA3 is a 2 Vrms dongle? FiiO own specifications clearly indicated it is 2 Vrms and up to 2.7 Vrms on BAL:

https://www.fiio.com/ka3_parameters
Yes it can go up to 2.7 Vrms and that is still a region of 2 no?

And have you checked Tempotec E44 specifications? clearly it is a 4 Vrms device which makes is double the powerful than KA3 even on paper


And just because in my summary I said KA3 is a weak 2 Vrms dongle and not specifically mention 2.7 Vrms, now the the accuracy is questionable?
Don't take my words for it, your opinion seems already set on what you want to believe, get yourself an actual E44 and KA3 side by side and compare, while at it get also Questyle M12 and Ovidius B1 - both of which prime examples of 2 Vrms dongles, test them, and see if you are hearing which ones has the better drive for magnetic planars, we can discuss all day long on assumptions and theories but nothing beats the actual experience.

It is not my interest to convince anyone, that is not why I am doing this. I did what I did because it is fun.
Now it's clear you lack basic understanding on how voltage and power works. These sources can't output the same voltage levels under different impedances. It doesn't say "up to 2.7V", it's 2.7V at 32ohm load, it goes up to 4V, starting from 68ohm. E44 also goes up to 4V, but only delivers that at much higher loads. Under any impedances up to 300ohm the KA3 delivers more voltage and consequently more power. It's not about convincing, it's simple math.
@32ohm - KA3 = 2.7V, E44 =2V
@68ohm - KA3 = 4V, E44 = 2.7V
@300ohm - KA3 = 4V, E44 = 4V
This matches both Fiio and Tempotec official specs as well as independent measurements.
Now I'm thinking you got confused about KA3 power and your impressions are based on placebo, it's weird to claim the weaker dac somehow drives "better". I don't care about some honest mistake, but it does bother me misleading people who trust your opinions like me.

Sources:
https://www.l7audiolab.com/f/fiio-ka3/
https://www.l7audiolab.com/f/tempotec-sonata-e44/
 
Last edited:

Users who are viewing this thread

Back
Top