In both case on my RPi it works. But for USB there is that line you have to add in a file, linked to generic USB 2 DACs. I guess it's not linked to the daughter board you've chosen (which daughter board do you have?) but the case most of the time. In Moode 3.6 it's a thing to activate or not. In earlier version it was to be manually added in a file. Check screen copy to help you to search in, say, the moode audio thread.
BTW, I'm now on a SPDIF board, Allo DigiOne, and it's sounding very rock'n roll! Check this YouTube done for my eBay sell, it does lock on a RPi, even at 24/192, on SPDIF.
BTW, I'm now on a SPDIF board, Allo DigiOne, and it's sounding very rock'n roll! Check this YouTube done for my eBay sell, it does lock on a RPi, even at 24/192, on SPDIF.
Attachments
I'm using the L2706USB daughter board, now i haven't tried Moode yet so that might be worth a try. Certainly not working with Rune or Volumio.. well not for me at least. Would you say that avoiding the usb and using a spdif hat is the way to go sq wise. The iqaudio digi+ seems ok for a try out.
IIRC the "bug" is there in unix, not on the music soft. You may fix that first. But the DAC is seen, just sound NOK.
As for SPDIF I only advise the Allo DigiOne (I know, costy...) as this is the best one, or the Hifi Berry Digi + Pro, this one also has the 2 base clock freq. which the RPi can't handle. Other cards won't get accurate clock. And as this DAC doesn't reclock at all and is jitter sensitive it's the only way.
The Raspberry Pi: Audio out through I2S | Dimdim's Blog
As for SPDIF I only advise the Allo DigiOne (I know, costy...) as this is the best one, or the Hifi Berry Digi + Pro, this one also has the 2 base clock freq. which the RPi can't handle. Other cards won't get accurate clock. And as this DAC doesn't reclock at all and is jitter sensitive it's the only way.
The Raspberry Pi: Audio out through I2S | Dimdim's Blog
I have couple of questions regarding mods:
1. Increase Decoupling of Analog Stage
It is recommended to use several 820uf/35V low ESR caps. I saw people use from 2 to 6 caps here - can I use less caps with higher capacitance ? What total capacitance is optimal (still fitting inside original case) ?
2.Increase TDA1387 Chip Supply Capacitance
Abraxalito and some people use 4x15000uF 6.3V caps here, some other people use just one (but unknown capacitance due to making pics from top). It's hard to squeeze large caps here, what minimal total capacitance should be target here ?
3. Caps mentioned above should be audio grade or regular ones are ok ?
4. In later versions of this DAC there are two blue round Philips
22uF caps on output (just before sockets) - is there need to replace them with some other caps ?
1. Increase Decoupling of Analog Stage
It is recommended to use several 820uf/35V low ESR caps. I saw people use from 2 to 6 caps here - can I use less caps with higher capacitance ? What total capacitance is optimal (still fitting inside original case) ?
2.Increase TDA1387 Chip Supply Capacitance
Abraxalito and some people use 4x15000uF 6.3V caps here, some other people use just one (but unknown capacitance due to making pics from top). It's hard to squeeze large caps here, what minimal total capacitance should be target here ?
3. Caps mentioned above should be audio grade or regular ones are ok ?
4. In later versions of this DAC there are two blue round Philips
22uF caps on output (just before sockets) - is there need to replace them with some other caps ?
There's no optimal total capacitance, the aim is to get the lowest possible impedance for the opamp supplies at higher frequencies. By 'higher frequencies' I mean above 2kHz. A normal electrolytic cap has a corner frequency around 3kHz meaning above this frequency its impedance is dominated by its ESR rather than its capacitance. This is why its important to choose low ESR caps, they have higher corner frequencies. Using more caps of lower value in parallel usually gives a lower overall ESR than a single, high valued cap.
Have you pulled out some of the TDA1387 chips? Its worth reducing their number (and increasing the opamp feedback resistor to compensate) - this has a similar effect to increasing the supply capacitance.
Increasing the supply capacitance on the DAC chips is only worthwhile if you're using passive I/V. If you keep the opamps (meaning you're using active I/V), don't bother.
Regular low ESR caps.
Purely a matter of personal taste I'd say. Suck it and see.
Hi guys! I´m in the same boat than krzyh , I purchased mine and I´m waiting to receive it soon.
I´m an enthusiast and the research spent of all the members of this thread (special thanks to abraxalito and malefoda among others!) are impressive.
For now, I´m planning to do the two first (and most easy) mods: opamp swap and removal of 2n2 Opamp Feedback Capacitors.
I had thought to remove the stock IEC connector and put a Schaffner iec connector with integrated filter (1A value, mouser part 631-FN9244B-1-06).
What do yo think? It would be an useful addition or a waste of time/money instead?
Thank you in advance!
I´m an enthusiast and the research spent of all the members of this thread (special thanks to abraxalito and malefoda among others!) are impressive.
For now, I´m planning to do the two first (and most easy) mods: opamp swap and removal of 2n2 Opamp Feedback Capacitors.
I had thought to remove the stock IEC connector and put a Schaffner iec connector with integrated filter (1A value, mouser part 631-FN9244B-1-06).
What do yo think? It would be an useful addition or a waste of time/money instead?
Thank you in advance!
Hi again, Well, I have bad news... today I have received the dac, but.... when I connected it to my system, after pushing the play button from the very first seconds I noticed some strange noises coming from the right channel, as if something were injecting and awful low-level noises to the signal (the noise changes with the sound, as if it were a reflect of it) . It´s very annoying becouse these noises are more evident with calm music like classical, piano or a intro of one song with with few sounds for example... however the other channel is dead-silent and sounds good, without any artifacts.
I opened my unit in order to inspect more carefully all the components soldered on the pcb, but all seems to be in good state. However I re-soldered the joints of the rca socket becouse they were a bit flimsy.
Even I swapped the opamps from one channel to another, to determine if a fault opamp would be the reason, but the noise is still present...
What can I do? I have very basic electronic skills but I´m feeling lost with this situation... I bought the already assembled version believing that at least it was a minimum QC control and the unit would sound ok but this is frustating for me, frankly.
I really wanted to do this little adventure and test this promising dac, but definitely I must have a bad luck problem... Why god, why me??? :<
I opened my unit in order to inspect more carefully all the components soldered on the pcb, but all seems to be in good state. However I re-soldered the joints of the rca socket becouse they were a bit flimsy.
Even I swapped the opamps from one channel to another, to determine if a fault opamp would be the reason, but the noise is still present...
What can I do? I have very basic electronic skills but I´m feeling lost with this situation... I bought the already assembled version believing that at least it was a minimum QC control and the unit would sound ok but this is frustating for me, frankly.
I really wanted to do this little adventure and test this promising dac, but definitely I must have a bad luck problem... Why god, why me??? :<
thunderdomme - have you tried swapping left and right at the RCA outputs of the DAC? That's just a sanity test that the problem is indeed in the DAC.
Second, what is your source? How are you connecting it to the DAC? USB, optical SPDIF, coax? Is it possible for you to try a different DAC input?
I vaguely seem to recall I had a problem like that once, but now I can't remember the details... could be my imagination!
Also, can you tell us a bit about your downstream components (e.g. preamp, amp, etc)?
Second, what is your source? How are you connecting it to the DAC? USB, optical SPDIF, coax? Is it possible for you to try a different DAC input?
I vaguely seem to recall I had a problem like that once, but now I can't remember the details... could be my imagination!
Also, can you tell us a bit about your downstream components (e.g. preamp, amp, etc)?
Hi Matt, thanks for your response.
Yes, I did the channel swapping as you say, but the problem is 100% with the dac. Besides, I have connected it in the same way as my other sources, with the same cables and configuration.
I´m an hifi freak and I really love to try and test many kinds of hifi stuff, so I have a lot of sources and equipment.
I have mainly commercial dacs as the Rega Dac-R , the newest Pro-ject s2 digital (dual ESS9038K2M, it really sounds very good with full MQA decoding and DSD512 capable) or my cd player Arcam CD36.
My preamp is an upgraded Audio research LS2-B mk and for amplification I´m using two 250MR monoblocs "top of the line" from the golden era of Tag Mclaren audio. My speakers are B&W 805S.
For me, combining british sources/amp/speaker with american cables & preamp gives me a very balanced and engaging sound.
I´m using a PC running W10 with foobar2000/spotify, connected by USB to a Gustard U12 digital interface that feeds my dacs either via optical or SPDIF.
I have tried both inputs of the TDA1387 , and the noise is still present. Swapping with my other sources, it sounds like a charm, zero problems.
At this point, if I send the unit back I would pay again for the shipping cost to China (not to mention the time to get other unit back) and maybe I´m thinking if it would be a better option to order only the DIY board (blue version) and then use the missing components (transformers) of the actual faulty board as the donor...
The blue version has any component pre-soldered on it?
One other thing I thought of: did you get the right power transformers for your country's mains? When I first got into this DAC, they were hard to find with 110v mains transformers, which I need here in the USA. If you got the 220 version and are running on 110 mains, it probably won't even turn on, but it's worth asking.
Do you by chance have a cheap amp and cheap speakers to use instead? I don't think there's anything wrong with the rest of your "ecosystem", but right now the focus is on getting the thing to work. The cheap speakers are mainly insurance against the unexpected.
Does your tda1387x8 DAC have USB input? By default they do not, but it's an upgrade option when you buy. Does your PC have its own coax or optical audio out that you could plug straight into the DAC? I.e., cut out the Gustard? You'll have to use software volume control, make sure you turn the volume all the way down and slowly increase it when experimenting. Again, the goal here is to try to find a way to get non-messed up sound. If we can achieve that, then we can compare what works and what doesn't, and hopefully have a clue as to the cause of the noise.
Where are you located? I have the "blue" kit version, that I started but never finished. I don't think I'll ever finish it (now that I make my own tda1387 PCBs). If you cover shipping, I'll send it to you. By default, nothing is soldered. But I think I got most of the SMD stuff soldered.
Also, I have a completely stock version of this DAC... if you're in the states, maybe I could ship it to you to borrow, just to compare. If that might work for you, PM me and we can work out the details.
Do you by chance have a cheap amp and cheap speakers to use instead? I don't think there's anything wrong with the rest of your "ecosystem", but right now the focus is on getting the thing to work. The cheap speakers are mainly insurance against the unexpected.
Does your tda1387x8 DAC have USB input? By default they do not, but it's an upgrade option when you buy. Does your PC have its own coax or optical audio out that you could plug straight into the DAC? I.e., cut out the Gustard? You'll have to use software volume control, make sure you turn the volume all the way down and slowly increase it when experimenting. Again, the goal here is to try to find a way to get non-messed up sound. If we can achieve that, then we can compare what works and what doesn't, and hopefully have a clue as to the cause of the noise.
Where are you located? I have the "blue" kit version, that I started but never finished. I don't think I'll ever finish it (now that I make my own tda1387 PCBs). If you cover shipping, I'll send it to you. By default, nothing is soldered. But I think I got most of the SMD stuff soldered.
Also, I have a completely stock version of this DAC... if you're in the states, maybe I could ship it to you to borrow, just to compare. If that might work for you, PM me and we can work out the details.
Hey guys, I don´t see much movement here, what´s going on?
Thanks to the priceless help from Matt, finally I was able to replace and populate the other board, and I did most basic mods (opamp upgrade, 2n2 caps removal, increase decoupling stage with Nichicon UHW caps. Pending to to the pin7 mod and NOS droop filter).
The dac sounds very correct, but compared to the other "reference" dacs I own, it has a noticeable less output volume, as if the voltage output were less than 2V (I don´t have measured it, however).
Can I correct it by replacing the values of the output capacitors? I remember a post from this thread advising about change some values...
Apart from the low volume output, the sound is somewhat a bit dark, not as open, resolved or airy as my Rega DAC-R (ok, the price difference between them is offensive, but this was my benchmark ). However, as I don´t have done the nos droop filter yet, maybe this could be minimized later...
On the other hand, I`ve replaced the AMS1117 3.3v / 5V voltage regulators in the board with two ultra low noise TPS7A4700 modules TPS7A4700-LM31X Ultra Low Noise LDO Voltage Regulator and after that I noticed an audible change for the better, as if the sound were more a touch "more real". I´m still testing them. Any of you have tried to upgrade those voltage reg?
Lastly, I wanted to know what´s your opinion about a new version of this dac that it was released recently, the "deluxe" edition: 2018 New Listing HiFi Eight Parallel TDA1387 Decoder CM6631 24bit 192k USB Asynchronous Audio DAC Finished-in Digital-to-Analog Converter from Consumer Electronics on Aliexpress.com | Alibaba Group
For the price (very good) and seeing it by naked eye It has several upgrades, and I would not be surprised if all these changes are based in all the mods posted here from all your contritutions.... This could be (maybe) the more mature or advanced version of the previous one, and to me is very tempting. What do you think?
Thanks to the priceless help from Matt, finally I was able to replace and populate the other board, and I did most basic mods (opamp upgrade, 2n2 caps removal, increase decoupling stage with Nichicon UHW caps. Pending to to the pin7 mod and NOS droop filter).
The dac sounds very correct, but compared to the other "reference" dacs I own, it has a noticeable less output volume, as if the voltage output were less than 2V (I don´t have measured it, however).
Can I correct it by replacing the values of the output capacitors? I remember a post from this thread advising about change some values...
Apart from the low volume output, the sound is somewhat a bit dark, not as open, resolved or airy as my Rega DAC-R (ok, the price difference between them is offensive, but this was my benchmark ). However, as I don´t have done the nos droop filter yet, maybe this could be minimized later...
On the other hand, I`ve replaced the AMS1117 3.3v / 5V voltage regulators in the board with two ultra low noise TPS7A4700 modules TPS7A4700-LM31X Ultra Low Noise LDO Voltage Regulator and after that I noticed an audible change for the better, as if the sound were more a touch "more real". I´m still testing them. Any of you have tried to upgrade those voltage reg?
Lastly, I wanted to know what´s your opinion about a new version of this dac that it was released recently, the "deluxe" edition: 2018 New Listing HiFi Eight Parallel TDA1387 Decoder CM6631 24bit 192k USB Asynchronous Audio DAC Finished-in Digital-to-Analog Converter from Consumer Electronics on Aliexpress.com | Alibaba Group
For the price (very good) and seeing it by naked eye It has several upgrades, and I would not be surprised if all these changes are based in all the mods posted here from all your contritutions.... This could be (maybe) the more mature or advanced version of the previous one, and to me is very tempting. What do you think?
I can't speak for everyone, but I know I took what I learned from this DAC and started creating my own simplistic tda1387 implementations. Abraxalito is working on a very sophisticated tda1387 implementation.
That is a legitimate limitation of the tda1387: the output is fairly weak. You can overvolt the tda1387 chips as Abraxalito does (that is, power them with 6v instead of 5v) and get a little more voltage swing. Paralleling chips will give you more current output. But, an active preamp or some kind of buffer is probably what you ultimately need.
I don't think so.
I think the high-frequency rolloff that all NOS DACs exhibit with 16/44 material could be perceived as a "dark" sound. If you have the means to re-sample your music, you could try oversampling at the source to see if that "brightens" the sound at all. IOW, keep the bit rate fixed at 16, but up-sample to 88.2khz or maybe even 176.4khz. I can't remember offhand how high a sample rate the tda1387 will take. I know it will do 88.2khz. As you increase the sample rate, the NOS "droop" effect is lessened.
I have not, but it sounds like a reasonable thing to do! One thing everyone seems to agree on in diyAudio world: better/cleaner/lower-impedance power is always a goal. (Exactly how good it needs to be is not decided.)
I've seen the listing (it's on ebay too), but have not purchased one. Going strictly from the pictures, it does appear to have incorporated at least some of the mods discussed here. If nothing else, the added size will make modding easier.
There's another new design that I'm currently playing with, see here. This is a single tda1387 chip, with a complete output stage: I/V, filtering, and buffer. (The x8 DAC discussed in this thread really only does I/V and a bit of filtering.) The low price makes it particularly appealing (at least to me). It's certainly the most comprehensive tda1387 implementation I've seen outside of this forum. It also uses really high quality regulators. I did some quick initial testing, and found the sound encouraging. Now I'm trying to back out a schematic.
Whilst this is all correct it applies only to using the tda1387 with passive I/V which (as I understand) isn't the case here as there are output opamps being used for I/V. So if the level is below 2V then there would appear to be something wrong with the DAC. I'd check that all 8 DACs are in fact contributing current and that the I/V resistor is the correct value (from memory it was something around 680R?). Given a current swing of 8mA the 'textbook' value is 700R.
tda1387 will definitely go up to 8X OS (not that I've tried it) - the limitation is likely in the S/PDIF receiver chip - I can't recall which one was being used. All will do 88.2kHz for sure.
I'm on the cusp of ordering one of these too to have a play. So many things to order...
Last edited:
Interesting... So talking about this particular DAC (subject of this thread, the x8 DAC), it has two op-amps, one for each channel. Those op-amps (along with the resistor) do the I/V conversion. (IIRC, I believe there are some extra caps and resistors for filtering as well. I drew up this output stage circuit once, but now can't find that paper!)
Are you saying that in addition to I/V conversion, the op-amps are also adding some gain?
In comparison to that new board I mentioned, the one with a single tda1387 chip, but a more complex output stage: it has four total op-amps, basically two per channel. (I'm currently working on trying to back-out the circuit schematic.)
Can you help me understand the difference between the one op-amp vs two-op-amp (per channel) approach?