apple USB-C dongle to rescue!
-
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!
Periodic Audio Rhodium DAC Review
- Thread starter amirm
- Start date
Veri
Master Contributor
- Joined
- Feb 6, 2018
- Messages
- 9,599
- Likes
- 12,041
Love those. Own a couple
I'm sad this is a thing, especially among professionals. This is science, not politics (even then, why do people scrap relationships over that..)Hopefully I don't loose another professional colleague in the industry over this review.
Because one is a science tied to a luxury/discretionary item, whereas the other is (often/mostly) a non-discretionary tie to how one lives his or hers life and thickness of their pocket book. But I digress.
That thing is so puny I was wondering where the circuitry was. Than to Ivan for the image posts. Btw @IVX , what goodies have you been cooking up in your lab?
NevosUt0p1a
New Member
- Joined
- Feb 16, 2021
- Messages
- 2
- Likes
- 0
Measurements aside, it sound absolutely fantastic with my Audeze LCD-GX, great instrument separation and incredible (but a bit enhanced) sub-bass.
but yes, the volume is even weaker than my onboard OMEN laptop, weird.
but yes, the volume is even weaker than my onboard OMEN laptop, weird.
Jimster480
Major Contributor
Wow this is just terrible.... For $50? it looks like it is a $3 dongle DAC. I can get those to test all day through my amazon program. I got tired of picking and testing them since they are mostly all the same.
Ismapics
Active Member
- Joined
- Nov 21, 2020
- Messages
- 215
- Likes
- 285
This is a review and detailed measurements of the Periodic Audio Rhodium DAC and headphone amplifier/adapter. It was kindly purchased new by a member and drop shipped to me. It costs US $99 but company has it on sale for US $49.
The Rhodium looks like any other USB-C phone dongle:
View attachment 116260
It was plug and play on my Windows 10 PC and that is how I tested it.
Rhodium DAC Measurements
As usual, I test these dongles as DACs (with high impedance load) so that we can assess its performance without the headphone amplifier portion (even though it is still in the loop):
View attachment 116261
Ah, this is not very good. Third-harmonic is quite high at nearly -81 dB which dominates the SINAD to the same tune. Output voltage is only 1 volt which means it will not have much power to drive high impedance headphones either.
Jitter test showed nice, flat noise floor but some data dependent jitter (due to 250 Hz square wave in j-test signal):
View attachment 116262
Most important test here is amount of power so let's start with 300 ohm load:
View attachment 116264
As suspected, there is not much power here:
View attachment 116265
Switching to 32 ohm load we get:
View attachment 116266
View attachment 116267
Company spec is much higher (around 35 milliwatts). Maybe that is one channel with much more allowance for distortion.
Listening Tests
I tested the Rhodium DAC using my Sennheiser HD-650. On tracks with full amplitude, there was enough volume to enjoy so dynamic power should be higher than shown. But on any track with less than max digital levels, I didn't feel there was enough volume there.
Conclusions
I met the company at an audio show and was impressed with their objective approach to headphone design and measurements in general. I am therefore disappointed to see such a low performing dongle sold at rather premium prices. I hope they can upgrade the platform to something better.
As is, I can't recommend the Rhodium DAC. Hopefully I don't loose another professional colleague in the industry over this review.
------------
As always, questions, comments, recommendations, etc. are welcome.
Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
Well no better than the $9.99 Apple USB-C Dongle. Maybe construction, but you can get 5 of the Apple ones and throw them away if they break before you reach $50.
OP
- Thread Starter
- #29
It would measure the same.
If he always uses the DX3 then it is less time fiddling with graphs and it also gives us a straight comparison across the board. The Hidizs S8 is a similar device and it actually gets pretty close.
AudioSceptic
Major Contributor
Ken Rockwell measured one here <https://www.kenrockwell.com/apple/lightning-adapter-audio-quality.htm>. Some find him a bit much, but he does the measurements.
Ismapics
Active Member
- Joined
- Nov 21, 2020
- Messages
- 215
- Likes
- 285
I am an Apple user and I have 2 of these dongles in use all the time. I do shrink wrap them so they last. I also keep another one just incase I loose one.
- Joined
- Jun 24, 2019
- Messages
- 8
- Likes
- 38
Hi all,
I was asked to comment on this thread, so thus I am...
First, let me start by noting here are links to the measurement file and the results:
PDF of results
APx Measurement File
Those with APx units (v6 software) can run the project themselves; for everyone else, the PDF shows the configuration and the results.
I did some quick stepped sweeps: 20 Hz to 20 kHz, 61 steps. I ran it with two sweep levels: -1 dBFS and -15.45 dBFS. The second setting corresponds to an industry-standard (EN50332) 179 mV RMS output from the device. Thus we're measuring effectively at full output and at "normal" output levels (179 mV into a 32 Ohm load corresponds to 1 mW, and that's what EN50332 bases things around).
All measurements taken on my APx515, with a pair of our Ti IEMs connected (32 Ohm loads). We find the following:
Max output voltage is approximately 1 Vrms. Into a 300 Ohm load, you will get just about 3 mW (p = V^2/R). However, there are precious few portable audio headphones with that high of an impedance; most are in the 16-50 Ohm range (note: we're a portable audio company; we don't build gear targeted towards home use, but stuff you can use when you're out and about).
Looking at page 4 of the measurements (remember, these were run with a 32 Ohm IEM attached) we see it generates 1 Vrms. That would be about 30 mW (p = V^2/R, 1*1/32 = 31.25 mW), not 7 mW as stated earlier.
At a -1 dBFS sweep, we see on page 5 that THD is below 0.008% THD down to about 250 Hz, and slowly rises below that, to 0.05% THD at 20 Hz. This is at 31 mW output into a real load; the Ti IEMs are 106 dB SPL @ 20 Hz at 1 mW, so this would represent an output level of 120 dB SPL. Rather loud, I think most would agree.
Page 8 shows the 3HD for the unit, swept from 20 Hz to 20 kHz, at that same -1 dBFS level. We see that it is the main driver of THD at the high output voltage; this is an effect of current pumping and we've done the best we can, within the constraints of the size and budget, to minimize that. However, 3HD is 0.005% at 1 kHz, and 0.015% at 100 Hz, despite our best efforts.
Page 9 shows the SINAD (basically S/N + THD) of the system. It's pretty good, being > 80 dB until the deep frequencies (where our sensitivity to THD is reduced). I would like it to be better down below 100 Hz, but we would have had to add some pretty big film caps and more shielding, both of which would double or triple the size of the USB end. And we deemed it better to stay small and pretty good down there, versus being a lot bigger and better down there.
Pages 10 to 16 repeat the measurements, but at the -15.45 dBFS level. On page 11 we see the result - we have 179 mV RMS output (EN50332 compliant - 1 mW into 32 Ohms). Page 12 shows THD is a bit higher across the band, but lower in the bass range. This is an effect of less voltage stress on the capacitors in the audio path - they're not pushed as hard (high voltage, film caps would help here - but those are quite large and expensive for the size of capacitor we need). So while midrange/treble THD "rises" from 0.005% to 0.01%, bass THD drops from 0.05% to 0.02%. And note - this measurement is a LOT closer to where you'll be listening.
Page 14 is also pretty instructive. This is "just THD" - no noise. Here we see THD at 1 kHz is 0.0035%! So at the lower output level (179 mV, not 1 V), our THD+N is dominated by noise. Not really a surprise; it doesn't take much noise to generate 0.005% when your output level is less than 200 mV! So at normal listening levels, we'll have a higher noise floor, but the THD is really non-existent for pretty much all the range, not breaking 0.008% (rated) until 80 Hz.
Page 15 also shows a pretty significant thing - 3HD is quite a bit lower at the 1 mW output level. At full output, 3HD drove most of the THD; here we see it's less than 1/5th of the THD. Again, this is voltage-on-cap based issues, and shows that at usual listening levels, the THD is not just low, the dominant odd harmonic (3HD) is a LOT lower and essentially a non-issue.
The last page, page 16, shows the "big picture". SINAD is quite a bit better, not really changing much at all in the mid/treble, but significantly improving in the bass range, gaining over 10 dB down below 50 Hz.
So, to sum up, Rhodium:
- can output 1 Vrms into a 32 Ohm load
- will generate 3 mW into a 300 Ohm load, and 30+ mW into a 32 Ohm load
- will have < 0.008% THD @ 1 kHz @ 30 mW output (1Vrms into 32 Ohms)
- has even better performance at typical 1 mW output listening levels
- was designed so that tradeoffs benefitted typical listening levels rather than rarely-used full output levels
- is not optimal for high impedance, large/inefficient cans
- is extremely portable
Oh, lastly: we use the ALC5686, not the 4050 as shown in a previous picture. It's a bigger package (6mm x 6.5mm) but with finer pitched pins. We also have a higher tolerance oscillator, and selected every cap in the system with an eye towards superior performance in the small package, but when tradeoffs arose we chose to give priority to typical listening levels (-25 dBFS to -10 dBFS) rather than full/max output (0 dBFS).
I hope this is instructive; for anyone wanting additional tests or information about our APx setup, please feel free to ask.
Dan Wiggins
Periodic Audio Inc.
I was asked to comment on this thread, so thus I am...
First, let me start by noting here are links to the measurement file and the results:
PDF of results
APx Measurement File
Those with APx units (v6 software) can run the project themselves; for everyone else, the PDF shows the configuration and the results.
I did some quick stepped sweeps: 20 Hz to 20 kHz, 61 steps. I ran it with two sweep levels: -1 dBFS and -15.45 dBFS. The second setting corresponds to an industry-standard (EN50332) 179 mV RMS output from the device. Thus we're measuring effectively at full output and at "normal" output levels (179 mV into a 32 Ohm load corresponds to 1 mW, and that's what EN50332 bases things around).
All measurements taken on my APx515, with a pair of our Ti IEMs connected (32 Ohm loads). We find the following:
Max output voltage is approximately 1 Vrms. Into a 300 Ohm load, you will get just about 3 mW (p = V^2/R). However, there are precious few portable audio headphones with that high of an impedance; most are in the 16-50 Ohm range (note: we're a portable audio company; we don't build gear targeted towards home use, but stuff you can use when you're out and about).
Looking at page 4 of the measurements (remember, these were run with a 32 Ohm IEM attached) we see it generates 1 Vrms. That would be about 30 mW (p = V^2/R, 1*1/32 = 31.25 mW), not 7 mW as stated earlier.
At a -1 dBFS sweep, we see on page 5 that THD is below 0.008% THD down to about 250 Hz, and slowly rises below that, to 0.05% THD at 20 Hz. This is at 31 mW output into a real load; the Ti IEMs are 106 dB SPL @ 20 Hz at 1 mW, so this would represent an output level of 120 dB SPL. Rather loud, I think most would agree.
Page 8 shows the 3HD for the unit, swept from 20 Hz to 20 kHz, at that same -1 dBFS level. We see that it is the main driver of THD at the high output voltage; this is an effect of current pumping and we've done the best we can, within the constraints of the size and budget, to minimize that. However, 3HD is 0.005% at 1 kHz, and 0.015% at 100 Hz, despite our best efforts.
Page 9 shows the SINAD (basically S/N + THD) of the system. It's pretty good, being > 80 dB until the deep frequencies (where our sensitivity to THD is reduced). I would like it to be better down below 100 Hz, but we would have had to add some pretty big film caps and more shielding, both of which would double or triple the size of the USB end. And we deemed it better to stay small and pretty good down there, versus being a lot bigger and better down there.
Pages 10 to 16 repeat the measurements, but at the -15.45 dBFS level. On page 11 we see the result - we have 179 mV RMS output (EN50332 compliant - 1 mW into 32 Ohms). Page 12 shows THD is a bit higher across the band, but lower in the bass range. This is an effect of less voltage stress on the capacitors in the audio path - they're not pushed as hard (high voltage, film caps would help here - but those are quite large and expensive for the size of capacitor we need). So while midrange/treble THD "rises" from 0.005% to 0.01%, bass THD drops from 0.05% to 0.02%. And note - this measurement is a LOT closer to where you'll be listening.
Page 14 is also pretty instructive. This is "just THD" - no noise. Here we see THD at 1 kHz is 0.0035%! So at the lower output level (179 mV, not 1 V), our THD+N is dominated by noise. Not really a surprise; it doesn't take much noise to generate 0.005% when your output level is less than 200 mV! So at normal listening levels, we'll have a higher noise floor, but the THD is really non-existent for pretty much all the range, not breaking 0.008% (rated) until 80 Hz.
Page 15 also shows a pretty significant thing - 3HD is quite a bit lower at the 1 mW output level. At full output, 3HD drove most of the THD; here we see it's less than 1/5th of the THD. Again, this is voltage-on-cap based issues, and shows that at usual listening levels, the THD is not just low, the dominant odd harmonic (3HD) is a LOT lower and essentially a non-issue.
The last page, page 16, shows the "big picture". SINAD is quite a bit better, not really changing much at all in the mid/treble, but significantly improving in the bass range, gaining over 10 dB down below 50 Hz.
So, to sum up, Rhodium:
- can output 1 Vrms into a 32 Ohm load
- will generate 3 mW into a 300 Ohm load, and 30+ mW into a 32 Ohm load
- will have < 0.008% THD @ 1 kHz @ 30 mW output (1Vrms into 32 Ohms)
- has even better performance at typical 1 mW output listening levels
- was designed so that tradeoffs benefitted typical listening levels rather than rarely-used full output levels
- is not optimal for high impedance, large/inefficient cans
- is extremely portable
Oh, lastly: we use the ALC5686, not the 4050 as shown in a previous picture. It's a bigger package (6mm x 6.5mm) but with finer pitched pins. We also have a higher tolerance oscillator, and selected every cap in the system with an eye towards superior performance in the small package, but when tradeoffs arose we chose to give priority to typical listening levels (-25 dBFS to -10 dBFS) rather than full/max output (0 dBFS).
I hope this is instructive; for anyone wanting additional tests or information about our APx setup, please feel free to ask.
Dan Wiggins
Periodic Audio Inc.
Last edited:
Torokun
New Member
Why such a negative reaction when the manufacturer actually responds?
Disclaimer: I own one of this company's IEMs (but not this product)(but wouldn't hesitate to throw them under the bus if I thought they deserved it)
Not only did they respond but they responded with data which contradicts ASR's. There would have been several possible ways to clear up the confusion this creates for the reader, none of which seems likely to happen given that this was a year ago, but here they are:
- Request a second sample from the company to rule out a defective unit (which they would probably have provided- its in their interest to provide a quality product and accurate data) and/or send this unit to Periodic to see how it tests on their AP setup (and rule out a defective unit there)
- Point out any observable flaws in the Periodic test procedures or interpretation of the data, or conflicts between the procedures (Note that the Periodic data was into a real-world load, not a 'dummy' load FWIW)
Lastly: Don't know what social media is being referred to but they certainly weren't defensive after getting panned here IMHO. My 2¢.
Last edited:
Similar threads
- Poll
- Poll
