• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Best cheap USB-C headphone dongles?

Thanks!! I found a website that lists the specifications for the FatFrequency Scarlet Mini IEM, including an impedance of 38 ohms, TechPowerUp Review. So, with an impedance of 38 ohms, wouldn't these dongles have more difficulty driving these headphones?
No, they will go like a charm, Apple dongle (us version) outputs 30 mW @32 ohm load, (very close to 38) that at 106dB/mw sensitìvity will much more than needed. JM20 goes up at almost double the power.
 
I just wanted to give a +1 for the Akliam PD5 with AKM AK4493SEQ along with RT6863 OP-Amps. I have a few similar devices and I like this one the most. Not only the slightly warm sound but also the klever creature comfort addition of a combined start/pauze and next/previous song button. Besides the volume buttons of course. It might seem silly but being able to change song from the device itself is very practical imho. Also, the machining of the case is excellent, with recessions for the metal buttons.
Besides the Akliam, i have a similar device from the Korean brand Kuang Pai with a CS43131 and ESS9601k Op-amps. No extra button. Just volume buttons. Sounds also quite good but has a little bit of a hiss on very sensitive iem's. The PD5 has a black background with the same IEM's. Not to say that the Kuang Pai sounds bad, it certainly doesn't and it will be a good choice for anyone looking for a budget 4.4mm portable device with more power then the usual 3.5mm CS43131 dongles.
I paid 62 Euro for the PD5 and 28 Euro for the Kuang. Both in discount with coupons etc on AliExpress.
 
No, they will go like a charm, Apple dongle (us version) outputs 30 mW @32 ohm load, (very close to 38) that at 106dB/mw sensitìvity will much more than needed. JM20 goes up at almost double the power.
For those who don't know, the IEMs in question have a ludicrous bass shelf upwards. One has to wonder what the impedance is at those low frequencies, for such a bass boost to be achieved. If the flat 38 ohm claim is true, or not.
 
Well yes, I kept it on the safe side, referring to ~85mW with opamp vs ~65mW no opamp @32 ohms. Just to say that no cheap dongle is going more than 100mW@32 in single ended (excluding the mentioned ka11)
How about jm20max? 195mW@32/ch (yes, I asked them, 195 per channel
 
How about jm20max? 195mW@32/ch (yes, I asked them, 195 per channel
It is indeed a "clone" of the Fiio KA11 with SGM8262 op-amp driver.
I heard it can output 2.5 Vrms at line level and being rated for 0.7 ohm output impedance it should output 2.445 Vrms @ 32 ohm load that converts into 187 mW @ 32, so seems legit.
 
I returned it after few days since it is seriously flawed in the control system, i found various firmware with different options about volumes buttons and steps but none gave a satisfactory and reliable user experience.
This is my review at aliexpress:

"Ok, it is cheap, but it has serious and potentially dangerous bugs, volume sometimes reset to max and button sometimes change track instead of control volume and are too small to be operated quickly, so you risk your ear damage. on technical side , i measured it and it can only deliver 1,5V, not 2V as advertised. Good side are decent power and no background hiss with sensitive iems. Not satisfied at all."

Don't know about power delivered into low ohms, it uses an ESS dedicated headphone driver (ES9603Q) that seems good but nearly never used in other products and didn't find any reference measurements about it, I was also disappointed measuring it at 1.5 V max out as line level but was a very quick test that I can't guarantee for.
ddHiFi tc35 pro M2 has cs43131 + ES9603Q, but the claimed power is 65@32/ch, which is strange bc 43131alone could do that...
 
Last edited:
ddHiFi tc35 pro M2 has cs43131 + ES9603Q, but the claimed power is 65@32/ch, which is strange bc 43131alone could do that...
We can clearly say nothing without measurements, but it seems surely a bit odd adding an additional op-amp when it is proven that the CS43131 alone can produce that power level @ 32 ohms.
Now, just for speculating, seeing how i got 1.5 Vrms at line level with that FAudio dongle with the same ESS op-amp, should this TC35 Pro do the same due to a hypothetical voltage limit of the op-amp, and assuming it has a low output impedance around 1 ohm, it would bring out 1..45 Vrms @ 32, that coincidentally is exactly 65 mW.
I repeat, it's just a fancy tale, but should this be the case then they would have made an unnecessary complication to achieve the same power than CS43131 alone at low ohms load but noticeably less power at high impedance load, where more voltage is needed, and with worse capability as a pure dac, and clearly it should be considered a failed implementation.
 
ddHiFi tc35 pro M2 has cs43131 + ES9603Q, but the claimed power is 65@32/ch, which is strange bc 43131alone could do that...
The CS43131 may be able to do that alone at 32 Ohm, but it appears it is current-limited at low impedance: the datasheet shows half of the power at 16 Ohm as compared to 32 Ohm. It may be the reason for the ES9603Q amplifier (speculating…).
 
The CS43131 may be able to do that alone at 32 Ohm, but it appears it is current-limited at low impedance: the datasheet shows half of the power at 16 Ohm as compared to 32 Ohm. It may be the reason for the ES9603Q amplifier (speculating…).
From L7audiolab measurements of mbquart mb21, basically a clone of Jcally JM20, we see that a single CS43131 alone can still drive 16 ohm load to 1.1 V -> 75 mW - 70 mA, so still more than 32 ohm

08 THDN-Ratio-vs-Measured-Level.jpg
 
From L7audiolab measurements of mbquart mb21, basically a clone of Jcally JM20, we see that a single CS43131 alone can still drive 16 ohm load to 1.1 V -> 75 mW - 70 mA, so still more than 32 ohm

View attachment 423383
So the 9603Q chip is more Marketing than anything…?

Unless it was originally designed around a DAC-only CS43198, and they switched over to a CS43131 due to part availability—other OEMs have done exactly that in the past.
 
So the 9603Q chip is more Marketing than anything…?

Unless it was originally designed around a DAC-only CS43198, and they switched over to a CS43131 due to part availability—other OEMs have done exactly that in the past.
Maybe it achieves that same levels without the pregressive increase in thd we se in the graph of cs43131 at lower ohm loads...Or maybe not... Who knows :rolleyes:
 
So the 9603Q chip is more Marketing than anything…?

Unless it was originally designed around a DAC-only CS43198, and they switched over to a CS43131 due to part availability—other OEMs have done exactly that in the past.
It's worse in all manners than integrated one in CS43131, the CS43198 doesn't have one nor do the ESS Q2M DAC's so they have to use external one.
 
CS43131 + DSP PEQ
This one was added to the TTGK/WalkPlay repository in Dec-2024. I suspect it’s using the same CB5100pro bridge as the JCally JM20-Pro.
What is interesting is that both the Tanchjim and WalkPlay app should work to set the PEQ.
 
I want to buy one of these 1$ usb dongle dacs from Aliexpress, between cx31993, kt02f20 and cb1200au. I have a bad experience with jcally jm6 pro, and testing some devices in this page shows very low snr at very low mw, indicating high noise, which is my experience, a lot of hissing.
Are there any tests with kt02f20 and cb1200au? I would like the one with less hissing and background noice. cb1200au would be the hi-max:

When it says eq tuning, I guess it doesn't mean it saves the eq inside the device, right?
 
I want to buy one of these 1$ usb dongle dacs from Aliexpress, between cx31993, kt02f20 and cb1200au. I have a bad experience with jcally jm6 pro, and testing some devices in this page shows very low snr at very low mw, indicating high noise, which is my experience, a lot of hissing.
Are there any tests with kt02f20 and cb1200au? I would like the one with less hissing and background noice. cb1200au would be the hi-max:

When it says eq tuning, I guess it doesn't mean it saves the eq inside the device, right?
A few of us helped you troubleshoot and from the information you gave there was no clear reason why you were experiencing abnormal hiss. Because of that there's no guaranteeing that future purchases will be different.
 
and testing some devices in this page shows very low snr at very low mw, indicating high noise, which is my experience, a lot of hissing.
For me, they fall into the “good enough” for portable use (IEMs, easy to drive HP), and I personally value the PEQ features of the CB1200AU a lot. But, if the JM6 Pro didn’t work for you, I’m not sure these will…

When it says eq tuning, I guess it doesn't mean it saves the eq inside the device, right?
Yes, the EQ settings from the ‘Walk Play’ Android app are saved onto the device—they “stick”, and carry over when plugging the dongle into something else.
 
For me, they fall into the “good enough” for portable use (IEMs, easy to drive HP), and I personally value the PEQ features of the CB1200AU a lot. But, if the JM6 Pro didn’t work for you, I’m not sure these will…


Yes, the EQ settings from the ‘Walk Play’ Android app are saved onto the device—they “stick”, and carry over when plugging the dongle into something else.

Well then, this settles it, I didn't expect this functionality in these really cheap dongles.
 
Back
Top Bottom