This is a review and detailed measurements of Audio-gd NFB-28.28 DAC, headphone amplifier and pre-amplifier. It is on a kind loan from a member and retails for USD $750 plus shipping from China. The unit is quite heavy so I suspect shipping will cost a few dollars.
This is a deep enclosure but half the width of the larger Audio-gd products:
Navigating the unit is maddening. 7-segment LED digits are used to indicate input number, setting, etc. I searched for the manual but could not find any. Instead there is some random text on Audio-gd website that starts to deal with something about a problem they fixed. How a simple device is made so complicated is beyond me. The unit could stand serious cleaning up of the documentation and revamping of the user interface.
There are lots and lots of inputs and outputs. I let you read the documentation for yourself. For my testing I used the USB input and either balanced XLR output for DAC testing and front 1/4 inch for headphone. Thankfully the DAC was plug-and-play compatible with Windows 10 Creators Edition so I did not need to install any drivers.
When I unpacked the unit, there was something metallic rattling in it. I started to unscrew the top to find out what it was, only to notice one of those stickers on it saying warranty void if opened. This is silly since changing DAC filter setting is done by changing jumpers internally! Anyway, I took a chance and powered it up and seems to be working.
EDIT: see this tear down and cause of that rattle: https://www.audiosciencereview.com/...b28-28-dac-and-headphone-amp.5147/post-115292
Previous to this unit, I had tested and reviewed the Audio-gd NFB27.38. The distortion performance of that unit was exceptionally poor. Some suggested then that perhaps that unit was broken. This unit gives us another bite of that apple. So let's get into measurements and see how she did.
Measurements
There is a switch on the back of the unit that forces the line out to be at fixed level. I activated that and that produced this horror show:
We have large and massive amount of second and third order harmonics respectively. Experimenting with alternate mode of variable output and various gain and level settings did not produce anything better. This produces by far the worst SINAD rating of any DAC I have tested regardless of price:
This is the same problem I saw in Audio-gd NFB27.38. For that device, they had published far better measurements using audio precision analyzer. The same is true here:
We actually see some similarities to my graph with respect to distortion/jitter sidebands around the main tone. But theirs shows far less harmonic distortion. My guess is that this is the performance of the DAC subsystem sans the output stage. And it is the output stage that is adding the copious amount of distortion.
Given the fact that I have now tested two Audio-gd products with two different analyzers, I have no doubt that there are design issues in this product and their measurements do not reflect complete units they ship to customers.
Anyway, continuing with the measurements, lets look at jitter:
Nothing really broken here other than higher noise floor and sidebands that get somewhat too high. Fortunately audibility of all of this is very low to non-existent so not a problem.
Intermodulation distortion versus level reveals same problems we see in our dashboard:
We see much higher noise floor but then a sudden rise in distortion to very high levels.
Linearity is decent since the test removes distortion and noise before reading the level:
Usually the right side of the graph is ruler flat and at zero but we see a rise there due to distortion issues.
Let's switch to headphone out and measure power using 300 ohm load:
Whether we use low or high gain, noise and distortion levels are substantially higher than other products.
At this point I am in sufficiently bad mood to not do any more testing.
There is no listening test for that reason and for the fact that I am watching the election news on this machine.
Conclusions
These Audio-gd products are complex and have tons and tons of discrete components. They seem to be camping on audiophile myths that things like feedback are bad. As such, they are providing performance that in my opinion is beyond broken. No way are these models of fidelity or what an audiophile should be chasing.
I am also unhappy that the company provides no usable documentation and continues to show measurements that don't relate to products being sold. Bending the truth in advertising is something, but under-reporting distortion by so many miles is not correct or ethical.
Needless to say, I cannot in any form or fashion recommend Audio-gd NFB28.28. Or frankly any other product they make at this point.
-------------
As always, questions, comments, recommendations, etc. are welcome.
If you like this review, please consider donating funds using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
This is a deep enclosure but half the width of the larger Audio-gd products:
Navigating the unit is maddening. 7-segment LED digits are used to indicate input number, setting, etc. I searched for the manual but could not find any. Instead there is some random text on Audio-gd website that starts to deal with something about a problem they fixed. How a simple device is made so complicated is beyond me. The unit could stand serious cleaning up of the documentation and revamping of the user interface.
There are lots and lots of inputs and outputs. I let you read the documentation for yourself. For my testing I used the USB input and either balanced XLR output for DAC testing and front 1/4 inch for headphone. Thankfully the DAC was plug-and-play compatible with Windows 10 Creators Edition so I did not need to install any drivers.
When I unpacked the unit, there was something metallic rattling in it. I started to unscrew the top to find out what it was, only to notice one of those stickers on it saying warranty void if opened. This is silly since changing DAC filter setting is done by changing jumpers internally! Anyway, I took a chance and powered it up and seems to be working.
EDIT: see this tear down and cause of that rattle: https://www.audiosciencereview.com/...b28-28-dac-and-headphone-amp.5147/post-115292
Previous to this unit, I had tested and reviewed the Audio-gd NFB27.38. The distortion performance of that unit was exceptionally poor. Some suggested then that perhaps that unit was broken. This unit gives us another bite of that apple. So let's get into measurements and see how she did.
Measurements
There is a switch on the back of the unit that forces the line out to be at fixed level. I activated that and that produced this horror show:
We have large and massive amount of second and third order harmonics respectively. Experimenting with alternate mode of variable output and various gain and level settings did not produce anything better. This produces by far the worst SINAD rating of any DAC I have tested regardless of price:
This is the same problem I saw in Audio-gd NFB27.38. For that device, they had published far better measurements using audio precision analyzer. The same is true here:
We actually see some similarities to my graph with respect to distortion/jitter sidebands around the main tone. But theirs shows far less harmonic distortion. My guess is that this is the performance of the DAC subsystem sans the output stage. And it is the output stage that is adding the copious amount of distortion.
Given the fact that I have now tested two Audio-gd products with two different analyzers, I have no doubt that there are design issues in this product and their measurements do not reflect complete units they ship to customers.
Anyway, continuing with the measurements, lets look at jitter:
Nothing really broken here other than higher noise floor and sidebands that get somewhat too high. Fortunately audibility of all of this is very low to non-existent so not a problem.
Intermodulation distortion versus level reveals same problems we see in our dashboard:
We see much higher noise floor but then a sudden rise in distortion to very high levels.
Linearity is decent since the test removes distortion and noise before reading the level:
Usually the right side of the graph is ruler flat and at zero but we see a rise there due to distortion issues.
Let's switch to headphone out and measure power using 300 ohm load:
Whether we use low or high gain, noise and distortion levels are substantially higher than other products.
At this point I am in sufficiently bad mood to not do any more testing.
There is no listening test for that reason and for the fact that I am watching the election news on this machine.
Conclusions
These Audio-gd products are complex and have tons and tons of discrete components. They seem to be camping on audiophile myths that things like feedback are bad. As such, they are providing performance that in my opinion is beyond broken. No way are these models of fidelity or what an audiophile should be chasing.
I am also unhappy that the company provides no usable documentation and continues to show measurements that don't relate to products being sold. Bending the truth in advertising is something, but under-reporting distortion by so many miles is not correct or ethical.
Needless to say, I cannot in any form or fashion recommend Audio-gd NFB28.28. Or frankly any other product they make at this point.
-------------
As always, questions, comments, recommendations, etc. are welcome.
If you like this review, please consider donating funds using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
Last edited: