the next step along the way.
I have discrete input, middle and output boards. it gives me a chance to change things out or play around before I settle on something final.
input board is passive in its current state but will get some 8pin opamps as input buffers. the caps (green nichicon muse) will be after the input op-amp but before the cirrus input pins. that will hopefully offer the most click-free volume changes.
the output board has room for opamps as well but I didn't finish all the wiring for them and so I have jumper stubs in place, for now. the relays do work and they do a great job of muting if 'bad things' happen or even just at turn-on and at turn-off time. even abrupt turn-off does a fast mute, so things are safe and quiet. these relays also act as a way to do output selection.
the rear has a firewire port (6pin standard using standard pc style cable) that goes to my DIY spdif switch. or it goes to my DIY port selector on my dac
either way, when I select inputs, it changes the spdif source. I have an option for analog relay switching too but these days, I do everything digital as a source.
the ribbon cable on the far left goes to a db9 on the back. there's a plastic dongle that holds an xbee rf modem. that gives me wireless remote control via a peer xbee that is connected to my pc via usb. a webserver connects to that usb/serial port and they talk. I threw together a shell script that keeps the local alsa system vol control in sync with the serial port cirrus chip. it works pretty nicely and its a nice new way to control things, using existing gui apps that others have written for their phones.
I'm very happy with this cirrus. very. even with all this crazy hand wiring, it still sounds really good and has none of the trouble that I was getting with the pga chip.
I have discrete input, middle and output boards. it gives me a chance to change things out or play around before I settle on something final.
input board is passive in its current state but will get some 8pin opamps as input buffers. the caps (green nichicon muse) will be after the input op-amp but before the cirrus input pins. that will hopefully offer the most click-free volume changes.
the output board has room for opamps as well but I didn't finish all the wiring for them and so I have jumper stubs in place, for now. the relays do work and they do a great job of muting if 'bad things' happen or even just at turn-on and at turn-off time. even abrupt turn-off does a fast mute, so things are safe and quiet. these relays also act as a way to do output selection.
the rear has a firewire port (6pin standard using standard pc style cable) that goes to my DIY spdif switch. or it goes to my DIY port selector on my dac
either way, when I select inputs, it changes the spdif source. I have an option for analog relay switching too but these days, I do everything digital as a source.the ribbon cable on the far left goes to a db9 on the back. there's a plastic dongle that holds an xbee rf modem. that gives me wireless remote control via a peer xbee that is connected to my pc via usb. a webserver connects to that usb/serial port and they talk. I threw together a shell script that keeps the local alsa system vol control in sync with the serial port cirrus chip. it works pretty nicely and its a nice new way to control things, using existing gui apps that others have written for their phones.
I'm very happy with this cirrus. very. even with all this crazy hand wiring, it still sounds really good and has none of the trouble that I was getting with the pga chip.
Attachments
taking the free software 'mpdremote' (php web app) and hacking it to be a bit more usable for my own preamp.
not a native android app but it didn't take much effort to get this going (and I am not yet an android developer).
when the rasp pi board is stable enough to use, it will be a nice little web back-end processor. for now, though, I still use a mini-itx intel board (dn2800 atom mobo).
not a native android app but it didn't take much effort to get this going (and I am not yet an android developer).
when the rasp pi board is stable enough to use, it will be a nice little web back-end processor. for now, though, I still use a mini-itx intel board (dn2800 atom mobo).
Attachments
Hi Bryan, I am following your steps on more forums incl. AMB as I intend to follow similar concept. As I am in pure stereo listening I wonder if going cirrus way makes sense? What was the trouble you got with pga? For my purpose pga seems easier to handle (soldering, less pins, ready kits on eB...;-)
Thanks
mba
even for just 2ch, I would prefer the cirrus. it was just easier to get better sound with this chip, for some reason. I tried a few pga variations but always had some noise bursts in some channels. could have been layout (they were mostly ebay boards with no understanding of how layout should be done) but its hard to believe the performance would be THAT bad. but it was. otoh, the cirrus with just hand wiring and perf board construction gives me no noise or issues.
the only thing I don't like is that the voltage is a bit low (9v) and so you can clip the input if you have a very hot (+22db style hot) input. for normal 2 or 4v signals, the cirrus is fine and there won't be over-driving.
soldering can be hard but using the schmart board makes it easier than it looks, in fact.
the only thing I don't like is that the voltage is a bit low (9v) and so you can clip the input if you have a very hot (+22db style hot) input. for normal 2 or 4v signals, the cirrus is fine and there won't be over-driving.
soldering can be hard but using the schmart board makes it easier than it looks, in fact.
Well, if an expert recommends the more difficult way...;-) I don't think I would overdrive it with my Carat Ruby USB DAC. Is the Atom 2800 based mobo still your first choice for a music server or are you aware of any new mobo type? I am not skilled in Linux so I hesitate if to get into it (mpd) or stick to W7/foobar (which I like for the customizable frontend). But some flexibility and remote control via TCP would be nice in the future...
I'm not sure I can recommend that 2800 board by intel. it now 'sings'! ;( in a bad way. you can *hear* the windings from the onboard components making noise as the system 'computes'. its insane. from a foot away, my average hearing can hear the damned thing computing. no AM radio needed. they must be using really poor dc-dc parts on that board.
the video was not supported by linux, but I knew that and was going to use this just for audio, anyway. but its still a strike against it that its not intel GMA open source video, its another video chip that intel just uses. the driver won't be open for a long time, if ever.
there's a lot I like about the n2800 atom board but also a lot I don't like. its not a pure win.
the video was not supported by linux, but I knew that and was going to use this just for audio, anyway. but its still a strike against it that its not intel GMA open source video, its another video chip that intel just uses. the driver won't be open for a long time, if ever.
there's a lot I like about the n2800 atom board but also a lot I don't like. its not a pure win.
I hoped the computing noise had gone with the ole cash register
...have to look elsewhere!
U used to build loudspeakers in the past, right ;.-)? I wonder if AMD Brazos-T platform will become available for DIY folks...
Nope, I got one of these: 2 Pcs Universal QFN/QFP/TQFP/LQFP 16-80 44 64 to DIP gold adapter PCB convertor | eBay
the chip was about $30 a year ago, but now its a bit lower.
fwiw, the schmartboard is very high quality. I can't always say the same about unbranded china stuff.
but its just an adapter, not too much to go wrong. when I get things like this, I check for continuity on all lines before I solder down. just to be sure.
fwiw, the schmartboard is very high quality. I can't always say the same about unbranded china stuff.
but its just an adapter, not too much to go wrong. when I get things like this, I check for continuity on all lines before I solder down. just to be sure.