Substandard 28 Euro costing ultra low budget stuff apparently works OK. I have heard worse (ESS) DACs to be honest. Indeed NE5532 does not cut the cake, as was expected. Will check the design of CheapESS later on. It does not have the typical ESS sound signature like ES9038Q2M and its predecessors had.
So it is worth its price if only to get to know ES9039Q2M. Eval board on the cheap 😉
So it is worth its price if only to get to know ES9039Q2M. Eval board on the cheap 😉
Last edited:
250 bucks is not cheap enough, I mean in the spirit of the thread ?
Anyine aware of a square tip for standard Aoyue soldering station to remove soic 8, please (sorry for the off topic) ?
Anyine aware of a square tip for standard Aoyue soldering station to remove soic 8, please (sorry for the off topic) ?
If one doesn't favor ESS sound signature, I have found a more "analogue" yet plenty detailed and sparkling sound in the 2024 version of the oldie Xiangsheng Dac05a II 4x Pcm1794 DAC. The tube output is just tacked on, which I find more forgiving but less resolved, so I just removed the tube completely. 4x LME49720 in I/V stage are already excellent, and the 3x OP2604AP in output stage can be rolled to one's tastes. If one uses balanced output, 3rd opamp can be removed, which converts balanced to single-ended. For about $250 for PCM1794 x4 with XMOS XU208 USB input with chassis and power supply included, a great deal IMO.
0401251534_HDR by drjlo2, on Flickr
0401251538_HDR by drjlo2, on FlickrI
I agree, the PCM1794 is a fine sounding dac. I have two in a differential config driving my IV transformers - no opamps needed.
Aha , I'm listening to it right now in Antelope Zodiac Gold. but I do like more expensive or obscenely expensive ESS dac implementations of the cheapest "car audio" version ESS9018 . Cheap would be Auralic Altair and expensive Weiss Dac202. I'd still pay $2k for Weiss if I had the money.
You can use a chisel tip wide enough so it can melt solder on all 4-pins at once on one side of the IC. Add some flux or some extra solder, stick a little jeweler screwdriver (or else something thin enough) under the IC and pry it up a little on one side as you melt the solder. Then unsolder the other side.
If no tip that wide then you could probably melt the solder on an end pin while bending the pin up with a little tool. That would leave fewer pins to do at once.
or just tin the solid core electrical wire , lay it around the pins in a loop apply solder to solder the pins to attach it to the wire and then heat the whole shebang so the heat distributes to all pins and the socket either fall off by itself or you will be able to pull it . Clean the mess
Yeah that's what I use but it always stress the other raw when you lift and finally break or weak too much the pads... I should invest on an air station, but for the few diy I make, it has no sense but fulling the shelves.
Thanks guys.
What temp do you use for your iron, and what type of flux and or solder do you add (you need to assure good heat transfer)? How far do you lift up the the first side you unsolder (it shouldn't be very far)?
Of course, if you try the trick too many times then you may start to damage the glue that holds the pads to the PCB.
Of course, if you try the trick too many times then you may start to damage the glue that holds the pads to the PCB.
I would use paste flux for that rather than a rosin flux pen. Paste will hold up to the heat better. Also, for lead-free solder I set my Hakko iron to 820F degrees. That's what it takes to work fast, including to solder SMD grounds with a small tip. Also, if you are adding leaded solder to lead-free solder for better heat transfer then you still have to get the lead-free hot enough to melt so the leaded solder can mix with it (and its better not to mix different types of solder, if possible).
Also, with SMD its really important to keep your solder clean and free of even too much finger oil. For leaded solder I would wipe it with a paper towel until the dark oxide layer is removed. For unleaded, I at least use an alcohol pad to wipe down the solder before use.
Also, it turns out thinner diameter solder may have a high percentage of flux, which can be helpful. For small SMD pads I use about .25mm diameter lead-free solder most of the time.
Also, with SMD its really important to keep your solder clean and free of even too much finger oil. For leaded solder I would wipe it with a paper towel until the dark oxide layer is removed. For unleaded, I at least use an alcohol pad to wipe down the solder before use.
Also, it turns out thinner diameter solder may have a high percentage of flux, which can be helpful. For small SMD pads I use about .25mm diameter lead-free solder most of the time.
Last edited:
Another option I found handy is a second inexpensive (e.g. T12) soldering station. Takes almost no space on the desk/shelf. Also useful to have a small and larger tip in each for standard work, no need to keep replacing the tip, just grab the other pen, it heats up from standby in single seconds. For desoldering just swapping both/the other tip for a suitable one, the chinese T12 tips cost next to nothing and come at any size and shape.
No the XO delivers 100 MHz to the ES9039Q2M. Apparently the IC has no issue with that despite the datasheet (for wussies as you said) not mentioning anything higher than 50 MHz. Reality versus datasheet. Sorry.
Just read the datasheet. ES9039Q2M has a register for dividing MCK to have lower SYS_CLK (register 5 CLK GEAR SELECT).
For the ES9039Q2M board, it can be helpful to find out how the MCU is programming the dac chip. Cheap and easy to monitor/capture I2C bus traffic at power-on. https://www.ebay.com/itm/124238666414?_skw=logic+analyzer+usb&itmmeta=01JTKHTA7D0ZW4VEYCVG68B9BC&hash=item1ced339aae:g:Xs4AAOSwQ7haxEa1&itmprp=enc:AQAKAAAA0FkggFvd1GGDu0w3yXCmi1d4nV26KAHfQEcbWEEjqv0lJub3a6c9GppDUWMwYbCupUi2NviB9p/kX5LBZ/qmhqd3ycol+UZq9ilbExWRKOZR7E8ohs2skduaAHB4KYZ+b1Ch5UULI76JS3dGMRc8HNECi5Nz228wWDdImuf/NtI7WqmoXuS39YFLa+/fwJvwDfBbvRX6K8OcHtmKScuDXxoJrffa1Ra20AfSu0aJrd2OGAhimgQ7ytHqsjLKuO1L+iXUMKUtfjSUOmjPtT3jYQA=|tkp:Bk9SR_Cj6fHUZQ
Well I am measuring not reading. As said it just works OK at 100 MHz. Opamp (voltage out mode) has some trouble with it though. Will try old friend LM6172 and some shielding when I get the time. Also a PSU with multi 3.3V regulators will be used. Newcomer (to me as I used to avoid recent ESS chips like the Black Plague) ES9039Q2M apparently has its own 1.2V LDO regulator onboard. Despite the beautiful cheapness of CheapESS one can hear the IC to be promising but this device sure needs work.
Nice piece of inexpensive Chinese DAC gear! Which happens to be the subject of the thread, how convenient 🙂
Last edited:
wow, this dac is tight!!! dream sound,
As the PCM53 is upper bass deep focused, the PCM58 is lower midrange focused, so it is easier on the ears, but it makes the mixes sound so TIGHT and the beat so great! it really sounds like a dream.
In a sense it has more details than the akm4396... which is completely brutal at revealing each sounds in its utter detail and spotlight.
on the PCM, its the opposite, everything is sacrificed for the enjoyment of the dynamics up down up down. it is a groovy fun dac to listen to.
after adjusting bit 4 and MSB with many signals, it sounds more focused.
WOW the sound is just WOW.
As the PCM53 is upper bass deep focused, the PCM58 is lower midrange focused, so it is easier on the ears, but it makes the mixes sound so TIGHT and the beat so great! it really sounds like a dream.
In a sense it has more details than the akm4396... which is completely brutal at revealing each sounds in its utter detail and spotlight.
on the PCM, its the opposite, everything is sacrificed for the enjoyment of the dynamics up down up down. it is a groovy fun dac to listen to.
after adjusting bit 4 and MSB with many signals, it sounds more focused.
WOW the sound is just WOW.
Yes, try the JCally jm20 dongle see audio science for review and testing DAC's are a done deal. For $ 20 - 30 you have high end.