I have an RPi3B+ running Moode with an Allo Boss DAC from its RCA output jacks into a simple headphone amp and headphones. I never notice any clicks, pops, bumps, or other noises when changing tracks, loading albums, or whatever. My amp doesn't have much gain (about +6dB), so for louder listening I run it with the volume control at about 2 to 3 o'clock.
Interesting
My first DAC was 5122 based, which must be run single ended. Most of the anomalies are noticed on the woofer for me using that DAC. There's so much talk about jitter and PSU noise, where for me it has been by far, the little clicks and pops etc.. I suppose that if I used a separate volume control that I could mute that somewhat. Going balanced solved it for me anyway..
Thx
My first DAC was 5122 based, which must be run single ended. Most of the anomalies are noticed on the woofer for me using that DAC. There's so much talk about jitter and PSU noise, where for me it has been by far, the little clicks and pops etc.. I suppose that if I used a separate volume control that I could mute that somewhat. Going balanced solved it for me anyway..
Thx
While tracks are playing it's more insignificant. Most of the sounds are heard as you say in switching tracks. I've still got my 5122 working on another speaker and I can notice sounds at times when the OS is reading data from a drive. Volumio. I can hear the dithering of any software system that uses it with my setup. This whole aspect has been my focus, well more than what the others are talking about. I've had to chip away at these things and know there are others that notice the same. It just doesn't rear it's head as much when you have a potentiometer in front of your amp, but you will notice it without on good speakers. What happens is that when the DAC goes open and close, there's a small change in the DC, which shows up on single ended outputs. Balanced out both exhibit the same change and is nulled, as you know how. While not entirely necessary, it has given my system a much more professional performance. It's dead quiet now even while wide open.
So much to respond to here...
First, @olsond3, you MIGHT want to find out a person's experience is with a certain area before insulting them arond it. I built my 1st DIY'd I2S-connected DAC in 1992, using the I2S feed from a CD player, with very short (<2") leads to from the CD player's filter chip outputs to the DAC chip's inputs, about the same as on the CD player's board. Worked quite well and I used that setup for over 10 years.
In the past 10 years, I've DIY's a number of other I2S-connected DAC setups, starting with several in typical slave mode, then moving to more traditional master mode using logic-based master signal generation along with isolation and local reclocking such as with Acko's S03 board (see here: Amanero Isolator/Reclocker GB ). I was on the product development team for a master mode DAC setup using a SBC similar to the RPi where it could not directly input the master clock, but required a bit clock signal (generated in logic on the DAC board interface section) to generate the I2S Data and LRCK feeds. Sadly this product never came to market due to some unforeseen limitations in the chosen SBC.
You can learn more about the limitations of master mode processing with the RPi in this article:
Raspberry Pi version B+ | H i F i D U I N O .
You can read more about the various master mode operational options of the PCM5122 in the datasheet (Rev1-16), pages 46-51.
AND you can see the various options Allo tested in post 601 of this thread from Allo:
New FIFO buffer for RPI/SBCs
These were the basis for my understanding of the PCM5122 / RPi master mode. AND my comments were based on actual listening trials. Have you done the same?
AND the results were different than I expected, which I often find and have worked hard to learn to hear over the years. In the case of master mode versus slave mode for this setup as implemented in Allo's Boss 1.2, as Dimdim so correctly stated, "... there exists no perfect solution. Only compromises. Engineering 101."
@Dimdim, I love your DIY'ing a solution around a RPi compute module. Being less technically adept, I've used the separate linear-regulated power supply board for the RPi 2B as outlined here: Mezzanine Power board for Raspberry Pi . I describe my implementation and experiences with it in several posts starting at post 25. The two RPi's I've modified this way are by far above my best stock RPi setups with roughly equivalent power supplies. Getting rid of the DC-DC converters on the RPi 2B board and using linear regulators instead have a very positive inpact, just as not using the +-15V DC-DC converters on the Allo Katana and instead feeding good linear-generated +-15V lifts this DAC from ok mid-fi into the top mid-fi to low hi-fi realms.
Sadly, the last few posts in that thread suggest we won't easily be able to do the same with the latest generation of RPis, based largely on the article cited in post 43.
AND both @Dimdim and @Soundcheck, my experience mirrors yours, in that the opsys and SW configurations make significant contributions to the system's SQ. @Soundcheck, I particularly appreciate your contributions to this area, especially the work you've documented on your blog.
Now back to the topic of this thread. @twocents, you could do much worse than an Allo original Boss as a starting point for RPi audio. I DID find the later, 1.2 version to be a significant upgrade. BUT I don't know of any other RPi DACs that sound better than either the original Boss or the Piano 2.1 (in dual-mono mode) for the price that you can get them for today.
You will get a SQ lift from 2 supplies. The RPi supply should be 5V/2A. I've used 1A supplies, but the RPi does not boot every time with them. The Boss does not require much... I don't know the current requirements, but I know it is <500mA. DO note that the original Boss took the movement of a surface mount resistor on the board to use a separate supply. Check with Allo for the technical manual.
You also get a lift from using an Isolator board, either Allo's original one or Ian Canada's IsolatorPi (I slightly prefer the former). IF you use an isolator board, then it is easier to use 2 seperate supplies for the RPi... you feed 5V/2A to the RPi as normal and feed 5V/1A to the Isolator board which will also feed the Boss with no modifications.
As for supply options, I have a thread 'Getting the best out of Allo.com's new Katana DAC' (Getting the best out of Allo.com's new Katana DAC... ), where in the first post I list a number of power options, some ready-to-use, more DIY. Hopefully that'll be helpful for you.
Greg in Mississippi
First, @olsond3, you MIGHT want to find out a person's experience is with a certain area before insulting them arond it. I built my 1st DIY'd I2S-connected DAC in 1992, using the I2S feed from a CD player, with very short (<2") leads to from the CD player's filter chip outputs to the DAC chip's inputs, about the same as on the CD player's board. Worked quite well and I used that setup for over 10 years.
In the past 10 years, I've DIY's a number of other I2S-connected DAC setups, starting with several in typical slave mode, then moving to more traditional master mode using logic-based master signal generation along with isolation and local reclocking such as with Acko's S03 board (see here: Amanero Isolator/Reclocker GB ). I was on the product development team for a master mode DAC setup using a SBC similar to the RPi where it could not directly input the master clock, but required a bit clock signal (generated in logic on the DAC board interface section) to generate the I2S Data and LRCK feeds. Sadly this product never came to market due to some unforeseen limitations in the chosen SBC.
You can learn more about the limitations of master mode processing with the RPi in this article:
Raspberry Pi version B+ | H i F i D U I N O .
You can read more about the various master mode operational options of the PCM5122 in the datasheet (Rev1-16), pages 46-51.
AND you can see the various options Allo tested in post 601 of this thread from Allo:
New FIFO buffer for RPI/SBCs
These were the basis for my understanding of the PCM5122 / RPi master mode. AND my comments were based on actual listening trials. Have you done the same?
AND the results were different than I expected, which I often find and have worked hard to learn to hear over the years. In the case of master mode versus slave mode for this setup as implemented in Allo's Boss 1.2, as Dimdim so correctly stated, "... there exists no perfect solution. Only compromises. Engineering 101."
@Dimdim, I love your DIY'ing a solution around a RPi compute module. Being less technically adept, I've used the separate linear-regulated power supply board for the RPi 2B as outlined here: Mezzanine Power board for Raspberry Pi . I describe my implementation and experiences with it in several posts starting at post 25. The two RPi's I've modified this way are by far above my best stock RPi setups with roughly equivalent power supplies. Getting rid of the DC-DC converters on the RPi 2B board and using linear regulators instead have a very positive inpact, just as not using the +-15V DC-DC converters on the Allo Katana and instead feeding good linear-generated +-15V lifts this DAC from ok mid-fi into the top mid-fi to low hi-fi realms.
Sadly, the last few posts in that thread suggest we won't easily be able to do the same with the latest generation of RPis, based largely on the article cited in post 43.
AND both @Dimdim and @Soundcheck, my experience mirrors yours, in that the opsys and SW configurations make significant contributions to the system's SQ. @Soundcheck, I particularly appreciate your contributions to this area, especially the work you've documented on your blog.
Now back to the topic of this thread. @twocents, you could do much worse than an Allo original Boss as a starting point for RPi audio. I DID find the later, 1.2 version to be a significant upgrade. BUT I don't know of any other RPi DACs that sound better than either the original Boss or the Piano 2.1 (in dual-mono mode) for the price that you can get them for today.
You will get a SQ lift from 2 supplies. The RPi supply should be 5V/2A. I've used 1A supplies, but the RPi does not boot every time with them. The Boss does not require much... I don't know the current requirements, but I know it is <500mA. DO note that the original Boss took the movement of a surface mount resistor on the board to use a separate supply. Check with Allo for the technical manual.
You also get a lift from using an Isolator board, either Allo's original one or Ian Canada's IsolatorPi (I slightly prefer the former). IF you use an isolator board, then it is easier to use 2 seperate supplies for the RPi... you feed 5V/2A to the RPi as normal and feed 5V/1A to the Isolator board which will also feed the Boss with no modifications.
As for supply options, I have a thread 'Getting the best out of Allo.com's new Katana DAC' (Getting the best out of Allo.com's new Katana DAC... ), where in the first post I list a number of power options, some ready-to-use, more DIY. Hopefully that'll be helpful for you.
Greg in Mississippi
Just a subjective opinion here. Recently acquired my first rpi...a 3b+ that came with a hifiberry dac+pro.
Used to stream Spotify, using the standard rpi 2.1(?) Amp wall wart.
It sounds noticeably worse than my other means of streaming Spotify....which is from a Sky Q set top box...optical into either my Denon Avr or via Cambridge DAC magic plus.....they both sound better than the rpi.
So maybe better rpi dac performance is out there for not a lot more money?!
Just my two bobs worth.
Used to stream Spotify, using the standard rpi 2.1(?) Amp wall wart.
It sounds noticeably worse than my other means of streaming Spotify....which is from a Sky Q set top box...optical into either my Denon Avr or via Cambridge DAC magic plus.....they both sound better than the rpi.
So maybe better rpi dac performance is out there for not a lot more money?!
Just my two bobs worth.
I've used a number of methods to get spotify working well. In the end I have had to settle with shairport-sync and streaming from the spotify app on the computer or phone etc..
Every other method has had issues, lag, no gapless etc.. The app in Kodi will have occasional pops and cracks when apparently, after the track is loaded, it gets taped together with what is already playing.
There are no professional solutions to my knowledge, with the only method I haven't attempted yet is to run the x86 Eltechs ExaGear Desktop and set spotify to open up automatically, headless. In this way, one can connect to it as a spotify connect. It'l never happen that spotify releases something to run on the pi for us, as it would undermine with other manufacturers products.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.