My personal preference in Foobar is to upsample all audio to either 352.8/384khz using the Sox resampler first. This way you bypass the digital filtering on the PCM5102A chip. I don't know whether sending more data via USB first minimises the effects of jitter, or it's the lack of digital filtering, but the soundstage is definitely more coherent.
: ура:I did everything as suggested by bjd, replacing only the switch on the relay, which automatically turns off the battery when the computer is turned off. The sound is very good, thank you!
Good to hear you modded DAC is working a treat.
I managed to improve the sound quality even further by finding a cheap USB hub that has a 3.3v chip being fed by USB power via a voltage regulator. I removed the voltage regulator and fed it's power from the DAC's battery directly using some hook up wire. The hub connects via an usb adaptor to the back of the DAC and from the hub a usb cable is connected to the PC. This cheap hub fed with clean LiFePo4 power works a treat a 'regen' clone.
I managed to improve the sound quality even further by finding a cheap USB hub that has a 3.3v chip being fed by USB power via a voltage regulator. I removed the voltage regulator and fed it's power from the DAC's battery directly using some hook up wire. The hub connects via an usb adaptor to the back of the DAC and from the hub a usb cable is connected to the PC. This cheap hub fed with clean LiFePo4 power works a treat a 'regen' clone.
hello all! I just ordered one of these dac's to use on an android car stereo that's connected to a class a/b MB Quart amp. Will i still benefit from removing those coupling caps in this scenario?
One thing I have noticed lately, working with the 1117 regs is that they work much better with a good quality, standard 100uf 50v electrolytic or so directly across the out pins and gnd.
Without that, it’s anyone’s guess as to what they are doing, and the noise created.
Without that, it’s anyone’s guess as to what they are doing, and the noise created.
@bjd:
I just found this topic...
I have bought a DAC board from China, containing the PCM5102 DAC chip. I will use it with Raspberry Pi as a media player, connected its I2S pins to the RPi's GPIO.
I want to upgrade a bit, to have the best possible sound quality. Your improvements sounds very interesting for me, but aren't exactly clear, which capacitors have you not removed (because I have a different board). Could you please clarify this?
Do you still use this board? Have you made further improvements?
I just found this topic...
I have bought a DAC board from China, containing the PCM5102 DAC chip. I will use it with Raspberry Pi as a media player, connected its I2S pins to the RPi's GPIO.
I want to upgrade a bit, to have the best possible sound quality. Your improvements sounds very interesting for me, but aren't exactly clear, which capacitors have you not removed (because I have a different board). Could you please clarify this?
Do you still use this board? Have you made further improvements?
Hi.
@chatziva.
I use two an orange pis as audio/web radio player (mpd+rompr), it works great and play DSD over pcm as well.
Mine are Sa9227 + pcm5102a.
I replaced ams1117 with lp5907, removed the two output capacitors, added 10uf on power line and it sounds nice, not as good as my ak4490 diyinhk kit but I find it really nice for price.
The only cons I see is they overheat a bit and I drilled the enclosure with many holes to ensure an air Flux.
@chatziva.
I use two an orange pis as audio/web radio player (mpd+rompr), it works great and play DSD over pcm as well.
Mine are Sa9227 + pcm5102a.
I replaced ams1117 with lp5907, removed the two output capacitors, added 10uf on power line and it sounds nice, not as good as my ak4490 diyinhk kit but I find it really nice for price.
The only cons I see is they overheat a bit and I drilled the enclosure with many holes to ensure an air Flux.
Has anyone tried a basic linear supply with these boards?
Wall wart, reg, caps...
Wondering how much improvement could be had just from eliminating the usb as it’s power supply.
Wall wart, reg, caps...
Wondering how much improvement could be had just from eliminating the usb as it’s power supply.
Pics.
Since I moded it a log time ago, I forgot that one had a heatsink and no 10u cap on supply line, the one with that cap is for my step father.
I'll add some caps to it too.
Only lp5907 instead of ams1117 and shunted output caps.
I got one with sa9023 + es9023 and, with the same mods, don't like it so much, bass and mids lacks a lot of "body" to my taste.
Since I moded it a log time ago, I forgot that one had a heatsink and no 10u cap on supply line, the one with that cap is for my step father.
I'll add some caps to it too.
Only lp5907 instead of ams1117 and shunted output caps.
I got one with sa9023 + es9023 and, with the same mods, don't like it so much, bass and mids lacks a lot of "body" to my taste.
Attachments
Do not expect a Rolls-Royce though.
On windows it needs savitech bravoHD drivers for asio, and you must plug it at boot in order to works as expected (in my case).
You can find it for cheap on aliexpress.
On windows it needs savitech bravoHD drivers for asio, and you must plug it at boot in order to works as expected (in my case).
You can find it for cheap on aliexpress.
The pcm5102 is distinctly a mid-fi part it seems, have some experience with it.
Doesn’t seem to do anything bad when it’s set up right, but isn’t stellar either.
I’m looking at the clock decoupling capacitor, and it’s a long ways from the power pin it looks like.
Was thinking that may be an area for potential improvement, along with a ferrite head.
Maybe I’ll get one to play around with.
Doesn’t seem to do anything bad when it’s set up right, but isn’t stellar either.
I’m looking at the clock decoupling capacitor, and it’s a long ways from the power pin it looks like.
Was thinking that may be an area for potential improvement, along with a ferrite head.
Maybe I’ll get one to play around with.
Some refreshs to this thread here.
I use this DAC with an OTG cable from my android phone. This needs a little different arrangement in power supply modifications.
1. I seperated the power supply completely from the USB source with an external cable, not connecting the positive supply. (As the DAC draws 220mA this will protect your phone from draining ...)
2. I used some mods discussed in this thread (removed output coupling capacitors, direct output straight out of the PCM5102, removed LDOs, used LiFeIon accu with 3.2V)
3. It is important for the completly seperated power supply to set the old 5v rail to 3.2V too. There are some circuit parts, which will need the supply. This mod can be simply done by "shortening" the IN and OUT of a removed LDO contact.
(The capacitor at LDO position in the picture has no meaning here).
Never plug in a USB cable to the old socket in this modification, this will directly feed 5V to the chip and probably destroy it.
Sound ist great from mobile phone, of course.
I use this DAC with an OTG cable from my android phone. This needs a little different arrangement in power supply modifications.
1. I seperated the power supply completely from the USB source with an external cable, not connecting the positive supply. (As the DAC draws 220mA this will protect your phone from draining ...)
2. I used some mods discussed in this thread (removed output coupling capacitors, direct output straight out of the PCM5102, removed LDOs, used LiFeIon accu with 3.2V)
3. It is important for the completly seperated power supply to set the old 5v rail to 3.2V too. There are some circuit parts, which will need the supply. This mod can be simply done by "shortening" the IN and OUT of a removed LDO contact.
(The capacitor at LDO position in the picture has no meaning here).
Never plug in a USB cable to the old socket in this modification, this will directly feed 5V to the chip and probably destroy it.
Sound ist great from mobile phone, of course.
Attachments
Not sure if I have some counterfeit version of the item. Can anyone test if this works correctly on linux? The usb device id I get is 262a:9217. Normally it should say 262a:9227, which is what my hifimediy and another sa9227 device I have returns back.
Because of I don't have the right device id, perhaps the wrong driver is being loaded, ,and the usb will only run at 1.0 speeds rather than 2.0 speeds, so no support for anything over 96khz, unlike my other 9227 DACs. The sound though, after shorting the caps sounds similar to my hifime uh1, which also uses the sa9227 + pcm5102 combo.
Because of I don't have the right device id, perhaps the wrong driver is being loaded, ,and the usb will only run at 1.0 speeds rather than 2.0 speeds, so no support for anything over 96khz, unlike my other 9227 DACs. The sound though, after shorting the caps sounds similar to my hifime uh1, which also uses the sa9227 + pcm5102 combo.
Found out that the issue might be the kernel. I was using 5.10, when I rebooted with lts 5.4, I got the right device id with normal driver support, able to play higher bitrates than 96khz. Still no alsa hardware volume control though.
Strange issue. Considering that might other sa9227 devices work fine with 5.10.
Listening closer, bass isn't as strong, and definition and imaging isn't as good as the hifime uh1. Perhaps after putting caps on the rectifer, it will help.
Strange issue. Considering that might other sa9227 devices work fine with 5.10.
Listening closer, bass isn't as strong, and definition and imaging isn't as good as the hifime uh1. Perhaps after putting caps on the rectifer, it will help.
Not sure where the edit button is, so I couldn't edit my previous message.
Found out that it isn't a kernel issue. The dac is weird, where when its plugged into the USB port at reboot, it loads the correct usb device id 262a:9227. If you unplug it afterwards and replug it in, it reads as 262a:9217, and you lose high speed usb/bit rates above 96khz.
I think this is what also windows users are having issues with, where the dac is working until unplugged, and then you need to reboot to get it working again. You should be able to see the same thing happening if you check the usb device id using a program e.g. usbtree view (USB Device Tree Viewer)
Perhaps the sa9227 chip isn't a real one?
Found out that it isn't a kernel issue. The dac is weird, where when its plugged into the USB port at reboot, it loads the correct usb device id 262a:9227. If you unplug it afterwards and replug it in, it reads as 262a:9217, and you lose high speed usb/bit rates above 96khz.
I think this is what also windows users are having issues with, where the dac is working until unplugged, and then you need to reboot to get it working again. You should be able to see the same thing happening if you check the usb device id using a program e.g. usbtree view (USB Device Tree Viewer)
Perhaps the sa9227 chip isn't a real one?