 This is the story of a 24 euro, 24 bit, 192 kHz digital to analog converter that has a remarkable sound quality, a peculiar emmer that blocks that and how to work around that. Two weeks ago I received a mail from my ex-colleague Peter van Willenswaart. A friend of his had a very good experience with a small 24 euro duck, of course of Chinese origin. Peter has ordered an example and could confirm the inspiring sound quality. Only on CDs that hold very hot levels, the sound quality was clearly lower. I have worked with Peter for five years and know how to judge his judgment, so I ordered one too at bold.com, the Dutch online shop according to the same model as Amazon. As soon as it arrived I connected it to my Setup 2 and the affordable Allo Nirvana Audio Great Power supply. I came to the same conclusion as Peter and his friend. Interesting. But let's first see where the duck fits into your stereo. You of course need an amplifier and loudspeakers. The duck with the rebarcable name Merchandise is to be connected over RCA cables to the analog input on the amp, for instance AUX or CD. The digital source is to be connected to the Merchandise duck. That can be a digital output of a CD player, a network player, to play from internet or your computer or NAS, or any other digital source with a speedy for tossing digital output. Despite the box it comes in doesn't mention 192 kHz, the duck accepts sampling rates from 32 to 192 kHz 24 bit depth. The all metal housing measures 57 by 74 by 22 mm and weighs 73 grams. On one side there is a 5 volt DC power input, a speed of input on RCA and an optical input on tossing here surprisingly called speed. On the opposite side we find a 3.5mm headphone jack, a power led and the left and right analog outputs on RCA. The power supply is of the biops type, bring your own power supply. Inside we find a small circuit board. Let's take a closer look. The incoming 5 volts DC is regulated down to 3.3 volts so the internal circuit after this voltage regulator runs on 3.3 volts. Next to it the Rui-Meng MS8416T digital interface receiver that translates the incoming AES3 signals from either RCA or tossing to the I2S signal that is sent to the upsampling CirrusLogic CS4344 DAC chip. So far so good. But when we look at the circuit board we see some nasties. Two of the four holes are very close to the edge of the board and the power led needed to be twisted slightly to fit the hole in the housing. And when we look at the back of the circuit we see far from tidy manual soldering. I started in my setup 3 where the NADC 316BE amplifier drives the modern short Avant902 loudspeakers and the REL T5 subwoofer that is also connected to the loudspeaker terminals. The merchandise was powered by the Allo Nirvana power supply and connected to the AUX input using simple RCA cables and to the Allo's US Bridge Signature Lux Raspberry Pi based Room Bridge over a 75 ohm video cable. The US Bridge Signature Lux internal is connected to the Accu-Fox Accu-Switch SE using normal CAT6 patch cable. It was immediately clear what Peter meant. The merchandise has an elegance and friendliness you normally find in clearly more expensive products. At least most of the time. Some CDs like those that are compressed to hell and back during mastering sound extra nasty and on tracks that have fierce snare or glockenspiel these instruments got their trenches killed. There is a fair stereo image, focusing is somewhat vague while sibilance is fairly controlled. Time to measure. I always measure equipment I review but rarely publish the measurements since they usually don't show why equipment sound is good as it sounds. This time it's different. You'll understand later why. Let's start with the frequency response, that is within half a dB from 31 Hz to over 40 kHz at 96 kHz sampling. At 44.148 kHz it of course stops at 20 kHz. And that's good enough. Then the linearity. This shows whether the levels in the digital domain are translated precisely enough to levels in the analog domain. The yellow line shows this. Minus 30 dB in the digital domain, the left scale and minus 30 dB on the bottom scale meet at the right point in the graph. The same goes for minus 60 dB's. From minus 90 dB's downward, that one on one relation gets somewhat lost and below minus 105 dB's the analog signal cannot get deeper since the noise is taking over. Very unusual is the non-linear behavior between 0 and minus 10 dB. I have never seen a DAC behave like this. Minus 0.6 dB at 0 dB full scale. When we look at the total harmonic distortion and noise versus amplitude, we see that the distortion is just below minus 100 dB's for all levels below minus 20 dB's to get seriously higher from there upwards. This is exactly what I heard. Measurement and listening outcomes confirm each other. The TAD plus N measurement shows a 1 kHz tone that is filtered out. It should show only noise like it does at higher frequencies. But here we see distortion products, tones, multiples of the 1 kHz stimulus that are added to the output due to distortion. Another way to measure distortion is intermodulation distortion. Here two tones are sent to the device under test. Distortion products show up as multiples of these tones but also as differences between these tones and distortion products. The measurement here is not very brilliant but above minus 10 dB the distortion goes sky high. Within a few days Peter discovered a serious problem in the design. Both the left and right analog output circuits contain a FET, a field effect transistor, that functions as a mute switch when an unstable situation or lack of digital signal occurs. Every DAC has a mute provision to kill the output when there might be a loud click or plop since that might damage the loudspeakers. But here, according to Peter, a FET is used that is designed for a voltage line of 5 volts while, as we have seen, the voltage line after the voltage regulation is 3.3 volts. That doesn't work and to get it working the designer or builder tried to solve that by connecting the gate and the source to earth. It seems that the FETs no longer work as mute switches. Unfortunately we forgot to test that before we desoldered the FETs by warming the middle pin and prying it loose using a miniature screwdriver. That can easily be done but if you want to attempt that, pause the video and read the warning on the left side of the screen. Peter sent me a photo of his oscilloscope showing one original and one modified output. Where the upper trace shows clipping in a negative part of the waveform, the lower trace shows the output of the modified channel and that's not clipped. I've modified both channels and measured it again. Linearity now shows linear behavior between 0 and minus 10 dB. Total harmonic distortion and noise versus amplitude now shows a more natural behavior at high levels and the FFT of the total harmonic distortion and noise now shows no distortion products. The intermodulation distortion, although not brilliant, shows a normal behavior at high levels. The overall sound remains identical at lower levels but now high level signals stay in DAC 2. Again a direct link between measurement and listening. So let's try the DAC in my setup too. Here the Marantz Ki-Perl light drives the acoustic energy radiance 1 loudspeakers over Kimber 4PR loudspeaker cable. As usual the REL T5 is connected to the loudspeaker terminals too. Here the merchandise DAC is powered by the Upturn Audio UltraCaps LPS 1.2 supercapacitor power supply and connected to the AUX inputs of the amp over Siltek London interlink and to the Singser DDC, see last week's review, over Vanderhulp Video 75 cable. The Intel NUG running RoonROG was connected to the Singser over an AudioQuest Pearl USB cable. On the other side the RoonROG NUG was connected to the AcuVox AcuSuite HE over Cat6 patch cable. Both the Singser and the NUG used the power supplies they came with. Now the high amplitude problem is solved at the expense of the Mute function, transient stay intact. That makes it a remarkably well sounding DAC especially if you consider the €24 buying price. This is a good example of a product that measures bad because it had an error but when that was solved it measures like a cheap DAC but sounds clearly better. It sounds relaxed without nasty artefacts. Of course, resolution, although not bad, is not of the level of the denaflips areas too that normally can be found in my setup too. The same goes for the depth of the serial image or the bass reproduction. But I have heard €1000 DACs sound worse. The main thing here is that it makes music. Let me start by warning you that you can order such a DAC and modify it all at your own risk. I'm not liable for possible problems that might arise. I cannot offer support either because I simply lack the time. It's a jungle out there. When I search for DACs at AliExpress there are at least four DACs that looks identical to the one I reviewed here. Apart from the branding of course. But that is easily printed. Does that mean that all are the same product? Hard to tell. There is one visually identical for sale at €2.43 plus €1.80 shipping. Is that the same and is it on discount since they discovered the dysfunction in FETs? Another one is €5.41 including power supply and free shipping. In that light, €24 excluding the power supply as delivered by MerchandiseBV is expensive. If it is internally identical or even better it has the right type of FETs, I can't say. Anyway, you don't know what you're buying and how long it takes to get it delivered. At the other hand, at that money and on that bombshell we come to the end of this video. There will be a new video next Friday at 5pm central european time. If you don't want to miss that, subscribe to his channel or follow me on the social media so you will be informed when new videos are out. Help me reach even more people by giving this video a thumb up or link to this video on the social media. It is much appreciated. Many thanks to those viewers that support this channel financially. It keeps me independent and lets me improve the channel further. If that makes you feel like supporting my work too, the links are in the comments below this video in youtube. I am Hans Beekhuyzen, thank you for watching and see you in the next show or on the HB project.com. And whatever you do, enjoy the music.