Nothing different that I can think of. All tracks were 16bit 44.1. The noise happened immediately as soon as the first song started to play. After the incident, I plugged everything back in to see if there was any damage and the song played fine without the screech. I can't help but wonder if there are newer versions of the bin files that I should have been using.
Around the date you mentioned, Allo changed the crossover filters after I pointed out to them that the response was way down at the crossover frequency and hence the bass was virtually absent. If you weren't experiencing very little bass you presumably had the modified filters. Either way, I don't see how that could be related to your issue though.
I don't have any issues with the Piano 2.1 - with DSP off and without firmware of course.
I also can't understand why anybody would use such a basic DSP approach as
supplied by Allo or TI.
Beside the very limited functionality, the soundquality IMO suffers big time when using that DSP.
However.
You seem to have quite an unusual and complex setup. Running JRiver !?!?
And a screen attached?? And quite a process chain.....
If you fry your tweeters it's usually your own fault. I guess you feed your amps straight from the DAC, aren't you??
@0db or 2V blown right into the amp, things can get quite nasty.
And not to forget - just pulling the plug is a very dangerous method on it's own!! If the noise wouldn't have killed your tweeters. Pulling the plug might
have accomplished it.
I also can't understand why anybody would use such a basic DSP approach as
supplied by Allo or TI.
Beside the very limited functionality, the soundquality IMO suffers big time when using that DSP.
However.
You seem to have quite an unusual and complex setup. Running JRiver !?!?
And a screen attached?? And quite a process chain.....
If you fry your tweeters it's usually your own fault. I guess you feed your amps straight from the DAC, aren't you??
@0db or 2V blown right into the amp, things can get quite nasty.
And not to forget - just pulling the plug is a very dangerous method on it's own!! If the noise wouldn't have killed your tweeters. Pulling the plug might
have accomplished it.
Certainly "hot plugging" will cause a lot of HF transients that destroy tweeters. I have had to rebuilt tweeters that "burst" when a family member hot plugged into an amp. The ferrofluid was blown out and the coil wires snapped but no heat damage. They were salvaged by unwinding 1 turn of the coils to reattach them.
Speculation1 : I have noticed (but not related to the DAC itself) that sometimes when I stream over a poor wifi connection I get the usual drop outs but more alarming is I've seem my amp(s) inputs go into "clipping" , the fans start up, and the sound is torn up. I can only speculate that sound bursts are being created from the error prone connection. I can hear rasping sound and subwoofer bursts during these events. Resetting the PI or switching to a wired connection returns to normal.
Speculation2 : I have also had nasty noise bursts when mucking about with a convolution engine while changing data formats trying to debug it. This is an unusual circumstance so the volume is turned down, I have filters, as well as fuses installed. Is it possible the data formats changed (ie. BE to LE would do it).
Speculation1 : I have noticed (but not related to the DAC itself) that sometimes when I stream over a poor wifi connection I get the usual drop outs but more alarming is I've seem my amp(s) inputs go into "clipping" , the fans start up, and the sound is torn up. I can only speculate that sound bursts are being created from the error prone connection. I can hear rasping sound and subwoofer bursts during these events. Resetting the PI or switching to a wired connection returns to normal.
Speculation2 : I have also had nasty noise bursts when mucking about with a convolution engine while changing data formats trying to debug it. This is an unusual circumstance so the volume is turned down, I have filters, as well as fuses installed. Is it possible the data formats changed (ie. BE to LE would do it).
I see now that it was premature of me to blame the piano for the damage to my speakers and that most likely the damage was done by me. I had a rough night.
I do run my pi over wifi so that might be something to look into. Currently I am not running any EQ (though I hope to).
I also have a thread going over at the JRiver forum to see if there is anything I might have messed up there. I have two other pis which both run JRiver and I have not had this issue before but that doesn't mean that I don't have a setting wrong.
Tonight, I spent some time unsuccessfully trying to recreate the issue. I turned the input gain way down on my amp so as not to damage anything further. Despite changing tracks 50 times maybe, starting then pausing/stopping, playing different sample and bit rate tracks I couldn't get it to act up.
As for my choice to run the piano using the DSP, my desire is to take advantage of proper bass management in a simple, reasonably priced system. Previously I had PC to USB/SPDIF converter to DSP to two DACs to an analogue multi-channel preamp to power amps. So pi + piano to power amp is very appealing from a simplicity standpoint. Unfortunately there are not a lot of options for a 2.1 system unless you go with home theater gear.
I appreciate the suggestions.
I'm still kind of curious to know for sure that I have the correct firmware and DSP files as early on there were issues with track changes (from what I can tell anyway).
Matt
I do run my pi over wifi so that might be something to look into. Currently I am not running any EQ (though I hope to).
I also have a thread going over at the JRiver forum to see if there is anything I might have messed up there. I have two other pis which both run JRiver and I have not had this issue before but that doesn't mean that I don't have a setting wrong.
Tonight, I spent some time unsuccessfully trying to recreate the issue. I turned the input gain way down on my amp so as not to damage anything further. Despite changing tracks 50 times maybe, starting then pausing/stopping, playing different sample and bit rate tracks I couldn't get it to act up.
As for my choice to run the piano using the DSP, my desire is to take advantage of proper bass management in a simple, reasonably priced system. Previously I had PC to USB/SPDIF converter to DSP to two DACs to an analogue multi-channel preamp to power amps. So pi + piano to power amp is very appealing from a simplicity standpoint. Unfortunately there are not a lot of options for a 2.1 system unless you go with home theater gear.
I appreciate the suggestions.
I'm still kind of curious to know for sure that I have the correct firmware and DSP files as early on there were issues with track changes (from what I can tell anyway).
Matt
I just said (or meant) that I'd consider the Piano21 rather unlikely to be the cause for your issue.
I didn't say it's impossible.
When running this or that scenario in the past not only me ran into several driver issues with the Piano21. We've seen several driver updates over time.
I also tested some DSP setups (to mute this or that channel) in the early days. I did experience weird behavior. However. I never followed up on that. Because I let the whole DSP approach alone afterwards.
And as DonVK suggested, if feeding a DSP with an incompatible or corrupted stream-format it might starts puking.
Bottom line. Don't exclude the Piano from your root-cause analysis.
****
And. You should also make sure you've got the latest DSP firmware on your image. It's stored under /lib/firmware/allo .
The firmware doesn't come with the kernel respectively driver though.
Your OS maintainer (Volumio,Moode/PcP,...) has to make sure that
you'll have it - if he supports the Piano.
Of course, you can also do it by yourself.
The FW can be downloaded from the Piano FW git.
Something like this should do to get it installed:
Code:
sudo su
git -C "/tmp" clone "https://github.com/allocom/piano-firmware"
cp -r /tmp/piano-firmware/lib/firmware/* /lib/firmware
rm -rf /tmp/piano-firmware@Allo.
Would be nice if you could add a small "HowTo" to your git README.md.
****
The Piano also requires a proper setup/init with alsamixer.
In the beginning I relied on Alsa storing my setup. And it did it.
However. I also realized that settings got changed after a while - due to whatever reason.
From that experience I figured the most secure way to achieve a reliable DAC/driver init, is to run an alsamixer init script - per DAC - that's being started "prior" to the audio app (e.g. mpd/squeezelite).
"Prior", because some mixer controls might be locked if the audio app is up'n running already.
If your favorite OS mirrors all alsamixer controls into the GUI, it'll most probably do the job (save and restore) for you.
Though, once in a while, you'd better check if nothing has changed.
Good luck.
Last edited:
Hi.
Just a hint from my side.
I think I mentioned that I'm running Kali and Piano21 (dual-mono).
Clocked by the Kali MCLK.
I power the Kali with iPower and a 47kuF Buffer.
The Piano21 gets powered over the GPIO PINS from Kali.
There were comments that powering the DAC separately would further
improve its performance.
I've been reluctant to attach a third PS to my PI-Tower.
That now led to another quick hack.
I had several Oscons 330uf around. I just attached 3 of them to the Piano2.1 power GPIO Pins.
Result:
I think it's worth a try. Things IMO sound more relaxed and solid.
That's what you usually experience when stabilizing the power situation.
Enjoy.
Just a hint from my side.
I think I mentioned that I'm running Kali and Piano21 (dual-mono).
Clocked by the Kali MCLK.
I power the Kali with iPower and a 47kuF Buffer.
The Piano21 gets powered over the GPIO PINS from Kali.
There were comments that powering the DAC separately would further
improve its performance.
I've been reluctant to attach a third PS to my PI-Tower.
That now led to another quick hack.
I had several Oscons 330uf around. I just attached 3 of them to the Piano2.1 power GPIO Pins.
Result:
I think it's worth a try. Things IMO sound more relaxed and solid.
That's what you usually experience when stabilizing the power situation.
Enjoy.
What best oscillators can I apply for Kali in place of 44.1/48? What about the socket for an external clock, can I use it? Why is there only one socket?
I have RPi3, Kali, Allo isolator, battery pack, iFi PS and Piano 2.1. What is the best stack of them and what is the best combination of power supply? Piano is not so important as I going to use it with an external DAC through I2S.
Is there a more convenient solution for charging the LIFEPO batteries?
I have RPi3, Kali, Allo isolator, battery pack, iFi PS and Piano 2.1. What is the best stack of them and what is the best combination of power supply? Piano is not so important as I going to use it with an external DAC through I2S.
Is there a more convenient solution for charging the LIFEPO batteries?
Are you referring to the MCK socket? That's the OUTPUT of the MCK signal, not input.
Size is not the same. We use much bigger inductor and there are 2x tpa3118 ICs