• 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!

JCALLY JM20 Headphone Dongle Review

Rate this headphone adapter:

  • 1. Poor (headless panther)

    Votes: 3 1.7%
  • 2. Not terrible (postman panther)

    Votes: 0 0.0%
  • 3. Fine (happy panther)

    Votes: 17 9.4%
  • 4. Great (golfing panther)

    Votes: 160 88.9%

  • Total voters
    180
Has anyone else been experiencing
connectivity issues with to the JM20? I have to unplug mine and plug it in again to get it to work (this is on a Pixel 8). I'll test it on my laptop later and see what dmesg reveals. -J
The problem is most likely OTG. Android turns off USB connections if not used for 10 minutes
 
Its quite old dongle actually, but nice to see it is good, I honestly didn't think it will measure even better than Tempotec Sonata HD Pro.
And I wonder if the recent Max version can be flashed with FiiO KA11 firmware just like JCAlly JM12 is flashable with FiiO JA11 firmware.

So far we know JM12 = JA11 (with lower noise floor I recall from the topic on ASR?), and also JCAlly AP7 = Sevenhertz 71 (same PCB), but with detachable cable and lower price. One of the AP series higher numbers also shares the case with some other brand's dongles.

I don't know if it will be of use for anyone, but here are all the JCAlly USB dongles I found written down in my notepad, the specs are taken from AliExpress seller's pages and can (likely are, including additional chips) incorrect at places, I skipped all on Realtek as didn't like my Avani.

JCAlly JM6 - CX31993, 1Vrms
JCAlly JM6 Pro - CX31993 (Max97220), 60mW@32Ohm 7.4mW@600Ohm
JCAlly JM7 - CX31993, 1Vrms
JCAlly JM08C - KT0210, 1Vrms
JCAlly JM09 - CS46L41, 25mW@32Ohm 8.6mW@100Ohm
JCAlly JM10 Pro - CS43131, 2Vrms
JCAlly JM11 - SPV4040, 10.5mW@32Ohm
JCAlly JM12 - KT02H20, 1Vrms //FiiO JA11 firmware compatible
JCAlly JM20 Pro - CS43131, 2Vrms
JCAlly JM25 - ES9318C (SA9312L), 35.15mW@16Ohm 1.2mW@600Ohm
JCAlly JM30 - CS43131 (SL9123L), 2Vrms, 31.25mW@32Ohm 6.7mW@600Ohm
JCAlly JM40 - ES9280C Pro, 2Vrms
JCAlly JM60 - ES9038Q2M (ES9603Q SA9317L RT1239Q), 2.2Vrms
JCAlly AP7 - AK4377, 1Vrms //Sevenhertz 71 PCB
JCAlly AP10 - 2x CS43131 1.8/3.8Vrms
JCAlly AP90 - 2x AK4493SEQ (CT601PR), 2/4Vrms, 125mW/240mW@32Ohm
JCAlly AP98 - 2x CS43198 (CT601PR), 2/4Vrms


I found this review online..

It has weaknesses would you believe:

I was going to say I recall Andy/donglemadness bashing JM20 so much, that even though I consider his reviews a SciFi, it made me look at JM10 Pro instead (I ended up with CS-Pro CS43131 as it was the cheapest). I kind of lost my marbles after he compared output power of VE Odo (KT02H20, under 1Vrms) and JCAlly JM6 Pro (CX31993 with Max97220, 1.6Vrms+ impedance dependent) using 32Ohm Sundara...
 
Conclusions
[Excellent] desktop class DAC performance...Really, what else would you be asking for?
Parametric EQ would be nice.
Its quite old dongle actually, but nice to see it is good, I honestly didn't think it will measure even better than Tempotec Sonata HD Pro.
And I wonder if the recent Max version can be flashed with FiiO KA11 firmware just like JCAlly JM12 is flashable with FiiO JA11 firmware.
Yep... I got the $6 JCAlly JM12 from Aliexpress last summer and can confirm that one sounds great and that flashing it to Fiio firmware provides parametric EQ.

How to details are easily available online. Flashing lets you program all of Amirm's headphone EQ curves (and lots of other curves) in the Fiio App.

All of this advanced digital technology in $10 dongles seems almost magical to an old man like me.
 
Its quite old dongle actually, but nice to see it is good, I honestly didn't think it will measure even better than Tempotec Sonata HD Pro.
And I wonder if the recent Max version can be flashed with FiiO KA11 firmware just like JCAlly JM12 is flashable with FiiO JA11 firmware.

So far we know JM12 = JA11 (with lower noise floor I recall from the topic on ASR?), and also JCAlly AP7 = Sevenhertz 71 (same PCB), but with detachable cable and lower price. One of the AP series higher numbers also shares the case with some other brand's dongles.

I don't know if it will be of use for anyone, but here are all the JCAlly USB dongles I found written down in my notepad, the specs are taken from AliExpress seller's pages and can (likely are, including additional chips) incorrect at places, I skipped all on Realtek as didn't like my Avani.

JCAlly JM6 - CX31993, 1Vrms
JCAlly JM6 Pro - CX31993 (Max97220), 60mW@32Ohm 7.4mW@600Ohm
JCAlly JM7 - CX31993, 1Vrms
JCAlly JM08C - KT0210, 1Vrms
JCAlly JM09 - CS46L41, 25mW@32Ohm 8.6mW@100Ohm
JCAlly JM10 Pro - CS43131, 2Vrms
JCAlly JM11 - SPV4040, 10.5mW@32Ohm
JCAlly JM12 - KT02H20, 1Vrms //FiiO JA11 firmware compatible
JCAlly JM20 Pro - CS43131, 2Vrms
JCAlly JM25 - ES9318C (SA9312L), 35.15mW@16Ohm 1.2mW@600Ohm
JCAlly JM30 - CS43131 (SL9123L), 2Vrms, 31.25mW@32Ohm 6.7mW@600Ohm
JCAlly JM40 - ES9280C Pro, 2Vrms
JCAlly JM60 - ES9038Q2M (ES9603Q SA9317L RT1239Q), 2.2Vrms
JCAlly AP7 - AK4377, 1Vrms //Sevenhertz 71 PCB
JCAlly AP10 - 2x CS43131 1.8/3.8Vrms
JCAlly AP90 - 2x AK4493SEQ (CT601PR), 2/4Vrms, 125mW/240mW@32Ohm
JCAlly AP98 - 2x CS43198 (CT601PR), 2/4Vrms




I was going to say I recall Andy/donglemadness bashing JM20 so much, that even though I consider his reviews a SciFi, it made me look at JM10 Pro instead (I ended up with CS-Pro CS43131 as it was the cheapest). I kind of lost my marbles after he compared output power of VE Odo (KT02H20, under 1Vrms) and JCAlly JM6 Pro (CX31993 with Max97220, 1.6Vrms+ impedance dependent) using 32Ohm Sundara...

I am willing to risk my JM20 Max to try this flashing thing. I already have a spare if it dies. I just have no idea how to go about doing such a thing. I do have a Windows PC and an eager mind for tinkering.
 
The problem is most likely OTG. Android turns off USB connections if not used for 10 minutes
It seems to happen only when first connecting so I am not sure that it could be to do with a 10 min timeout. I have had no problems with my Musiland MU1, an Envel branded DAC or another cheap no-name DAC. I tested with my laptop and it seems to connect perfectly every time and dmesg never shows anything failing. I'll test it on a few other phones to see if it is just the Pixel. Thanks, John
 
It seems to happen only when first connecting so I am not sure that it could be to do with a 10 min timeout. I have had no problems with my Musiland MU1, an Envel branded DAC or another cheap no-name DAC. I tested with my laptop and it seems to connect perfectly every time and dmesg never shows anything failing. I'll test it on a few other phones to see if it is just the Pixel. Thanks, John
There may be a compatibility issue between your smartphone and the dongle. I had a problem with the Apple dongle on my on my old Huawei. There was a crackling noise when playing music. This was not observed on other smartphones. Unfortunately, there are too many Android-based smartphones that have different components and levels of competence.
P.S. A manufacturing defect at Jcally cannot be ruled out either.
 
I would like to use the JM 20 as a source for stereo systems provided in the vacation apartments when traveling, so for the time being I have only measured with a high-impedance load. The class-H mode doesn't seem to be a disadvantage in this case. The THD vs. level plot certainly makes some other DACs envious.

Please note, that the THD+N curve will be pretty much dominated by the mains harmonics in my setup (no USB isolator used).

Left channel was measured only since the 1kHz spectrum just looked "identical" on the right channel.

Setup THD (vs level):
JCally JM20 @ 48kS, both channels driven
E1DA APU @ 1kHz (10kHz), 26dB gain
E1DA ADCiso @ 192kS, mono mode
Software: Multitone; appropriate calibration file for APU loaded
1kHz: BW 22.4kHz; 10kHz: BW 48kHz

Multitone measurements:
REW @ 48kS
JM20 directly connected to the ADCiso, so loading was ca. 650 Ohms
Armir's multitone measurement close to FS looks even better. I guess the pre-filters in the AP-analyzer do their job, while the ADCiso is struggling with the noise-shaping spectral products of the CS43131.

--> It's almost unbelievable that measurement-wise “a cable” declassifies most high-end DACs from 10 years ago ... With up to ca. 50dB attenuation the performance is on par with CD quality. EDIT: Probably a bist optimistic - I have to rethink this comment

The picture might look a bit different when the dongle is loaded with e.g. 32R due to the class-H mode that some sources claim is being used in the implementation of the JM20 firmware.
 

Attachments

  • JM20_L_0dB_48kS_THD-vs-Level-1kHz_openLoad-APU-1kHz-26dB_ADCiso-mono-192kS.png
    JM20_L_0dB_48kS_THD-vs-Level-1kHz_openLoad-APU-1kHz-26dB_ADCiso-mono-192kS.png
    53.3 KB · Views: 49
  • JM20_L_0dB_48kS_THD-vs-Level-10kHz_openLoad-APU-10kHz-26dB_ADCiso-mono-192kS.png
    JM20_L_0dB_48kS_THD-vs-Level-10kHz_openLoad-APU-10kHz-26dB_ADCiso-mono-192kS.png
    49.3 KB · Views: 51
  • JM20_L_0dB_48kS_THD-1kHz_openLoad-APU-1kHz-26dB_ADCiso-mono-192kS.png
    JM20_L_0dB_48kS_THD-1kHz_openLoad-APU-1kHz-26dB_ADCiso-mono-192kS.png
    62.3 KB · Views: 58
  • JM20_Multitone32_1.3dBheadroom_48kS_ADCiso-mono-48kS.png
    JM20_Multitone32_1.3dBheadroom_48kS_ADCiso-mono-48kS.png
    105.6 KB · Views: 55
  • JM20_Multitone32_21.3dBheadroom_48kS_ADCiso-mono-48kS.png
    JM20_Multitone32_21.3dBheadroom_48kS_ADCiso-mono-48kS.png
    100.5 KB · Views: 47
  • JM20_Multitone32_41.3dBheadroom_48kS_ADCiso-mono-48kS.png
    JM20_Multitone32_41.3dBheadroom_48kS_ADCiso-mono-48kS.png
    93.6 KB · Views: 48
  • JM20_Multitone32_61.3dBheadroom_48kS_ADCiso-mono-48kS.png
    JM20_Multitone32_61.3dBheadroom_48kS_ADCiso-mono-48kS.png
    88.7 KB · Views: 49
Last edited:
Only thing I found under 'JCALLY JM20' looks completely different with aluminum finished metal. I dont want to get stuck with a lookalike or fake item, so, where are people sourcing these? Amazon desnt have it. Aliexpress has a standard version, a 'high end' one with silver plated wire, and one with dual chips, none quite matching the unit depicted,, but all listed as JM20 something or other, Any guidance here?
 
Last edited:
Only thing I found under 'JCALLY JM20' looks completely different with aluminum finished metal. I dont want to get stuck with a lookalike or fake item, so, where are people sourcing these? Amazon desnt have it. Aliexpress has a standard version, a 'high end' one with silver plated wire, and one with dual chips, none quite matching the unit depicted,, but all listed as JM20 something or other, Any guidance here?

https://it.aliexpress.com/item/1005006807697568.html

this is identical to the one i got almost 1 year ago.
 
The one I measured I have bought at HiFiGo

Beware: The JM20 doesn't support the microphone of headsets. The JM20pro has an additional ADC and supports headsets.
 
It is considerably more expensive than the Apple dongle, but obliterates it in every respect.
As one might expect given progress in the category. Time marches on, and the Apple dongle remains a viable product whose main attraction is not as much as its respectable-if-not-SOTA measurenents and SQ as its value (probably due to Apple's ability to apply economies of scale). It's not a spring chicken in this rapidly evolving space though, and its bang-for-the-buck is now being challenged by new devices such as this. Here we have a device nipping at the SOTA for just a few bucks more. If all you need is something for your commute or listening at the library, the Apple device many seem to love to hate is still fine. This one brings to the table the ability to use it as a no-excuses DAC in a no-excuses system too, and for a pittance, no less,

I wonder how many could tell the difference in a double blind.
 
Last edited:
All these SINAD numbers mean nothing when you hear clicks and pops. I won't buy another CS43131 product after bad experience with Tempotec Sonata HD Pro
I have the same device and the problem went away after replacing the (thankfully) interchangeable defective original USBC input cable. Oddly, cable orientation has an effect- one way pops on insertion, one doesnt. That shouldn't be possible but there it is.
 
All these SINAD numbers mean nothing when you hear clicks and pops. I won't buy another CS43131 product after bad experience with Tempotec Sonata HD Pro
In case you are referring to this phenomenon:
These clicks that occur on some CS43131 based products seem to be related to the class H mode being used for the output stage. There is a register setting in the CS43131 that allows to choose between class H and class A/B. The manufacturer can decide for power saving (class H) or optimal SQ (class A/B).

--> All I want to say is, that it's not fair to blame the CS43131 (and CS43198) in general - it depends on the implementation of the firmware

I have not yet found time to listen carefully and just did a few measurements with 650 Ohm load (E1DA ADCiso, mono) because operation with a high- ohmic load will be my use- case. These measurements did look fairly clean (post #147) so I cannot really imagine there's something that is audible.
 
By the way it is available for $16.99 ($19.99 + $3.00 coupon) here (at least for me in the States):


I got this unit that was reviewed for this price back in June so it seems to be often on sale.
By the way it is available for $16.99 ($19.99 + $3.00 coupon) here (at least for me in the States):


I got this unit that was reviewed for this price back in June so it seems to be often on sale.
Amazon does not have it and before ordering it from Aliexpress I would like to understand why it appears repeated many times with different prices. Which of them is the real one?
 
Back
Top Bottom