My WaveIO card arrived just before Xmas and on the very same day I had it up an running since the chassis/PSU was ready for it to be dropped in.
For anyone interested, here it is: Hi-Fi Projects USB to S/PDIF Converter
It sounds exactly the same as a Berkeley Alpha USB.
Thank you Lucian. It is really very good.
For anyone interested, here it is: Hi-Fi Projects USB to S/PDIF Converter
It sounds exactly the same as a Berkeley Alpha USB.
Thank you Lucian. It is really very good.
Today I swapped out my Windows OS disc for a fresh ssd and loaded Daphile, which is a Linux/squeezeserver build designed for a headless server streamer approach.
Took all of 10 minutes to get it working. My WaveIO was recognised without issue. This is my second play with this package. Now I have tweaked the power arrangement for the WaveIO/Buffalo this is working very well. Sounds just as good as the Windows/Jplay install it replaces. And I've been able to take a computer monitor/keyboard & mouse out of the music room.
I can connect my android phone wirelessly and browse my library and so on.
Took all of 10 minutes to get it working. My WaveIO was recognised without issue. This is my second play with this package. Now I have tweaked the power arrangement for the WaveIO/Buffalo this is working very well. Sounds just as good as the Windows/Jplay install it replaces. And I've been able to take a computer monitor/keyboard & mouse out of the music room.
I can connect my android phone wirelessly and browse my library and so on.
Squeezelite
Result - bloody fantastic! Best I have yet heard. And it's all free.
Coincidentally, I just tried Squeezelite/Logitech media Server on my old Win 7 laptop. Took all of 10 mins to get up and running and recognised my WaveIO no problem and Logitech do an excellent Android tablet/phone controller app which works really well. Squeezelite is a headless SB Touch replacement (software) and can act as a streamer or a player. Large screen, keyboard and mouse ejected.
Result - bloody fantastic! Best I have yet heard. And it's all free.
The onboard clock is pretty well specced. There might be more to gain in feeding the onboard device with its own offboard regulated power.
I'm sure Lucian will be along with more detailed reasoning...
Hello Guys! I promise myself to test drivers in Windows 8.1 and I did in the first Christmas day. Besides the fact that my personal opinion about this new MS OS is quite mixed up (liked few features but hate the rest
) I can confirm that the WaveIO drivers were installed quite easy from the first time and without any problems. I want to take a snapshot to prove what I'm saying but I have to boot in Win 8.1 for that...Now, it would be wise from my side to repair my early mistake in which I blamed Thesycon for doing something that cannot be entirely true: releasing a faulty driver pack. I apologies for my behavior!
Thank you Ian, I'm glad you like it! On your blog, I also do love to see (among other things) that heatsink glued on top of XMOS processor
You did noticed that it goes pretty hot out there but be without any worries because that particular XMOS package does have a thermal relief pad under the chip itself, pad which is directly soldered to the PCB using multiple vias. Taking into account that WaveIO does have four PCB layers from which two are planes with 70um copper thickness and adding the fact that the vias under the uP are free of thermal reliefs ... you can imagine that it will be quite hard for that chip to be damaged by high temps.I know that from my experience: sometimes, without proper temps set on the preheater and hot air gun, it is excruciatingly painful to solder that chip without damaging it. Anyway, after blowing few of them, I've learned how things should be done properly
The onboard clock is pretty well specced. There might be more to gain in feeding the onboard device with its own offboard regulated power.
I'm sure Lucian will be along with more detailed reasoning...
Well, in my honest opinion I should say this: every oscillator on that board does have its own low-noise PSU and associated decoupling network. Yes, its not perfect but the layout was made in such a way that the track length from the PSU to the oscillator should be as short and wide as possible. There are few other things that are very important to mention but I'll end here.
On the other hand, feeding any of the oscillators from external PSU could be appealing BUT you should place it as close to the oscillators as possible to minimize the wires' parasitic effects. Looking at how WaveIO is built I doubt that you can do that easily. I remember I did this for someone: an modified WaveIO version with three shunt regulators (78xx replacements like). I do not know about the sonic improvements but from the built side of things I wasn't very pleased about the results: the wires were too long for me and I had to place those regs on a metal frame because the risk to damage the copper pads was high!
After all, what I'm saying can only be taken as suggestion and if you want to do it then feel free to test your skills
Happy holidays to all of you,
Lucian
Last edited:
what about Black Gate cap on the clock pin
If we wanted to experiment with possible clock improvement, would not a simple approach be to add a Black Gate N (or NX) cap right on the power pin of the clock? A small value cap like BGN 4.7u on pin 4 of X1?? I assume this is the clock used when listening to 44.1 material?
If we wanted to experiment with possible clock improvement, would not a simple approach be to add a Black Gate N (or NX) cap right on the power pin of the clock? A small value cap like BGN 4.7u on pin 4 of X1?? I assume this is the clock used when listening to 44.1 material?
Cheers Lucian. I don't know how you solder like you do. I have enough trouble with through-hole!
Anyway I use an IR thermometer on my projects - especially the PSU regulator heatsink and heatsink in Class-A amps etc. I pointed it at the "naked" XMOS and it was a bit warm - nothing too serious though - maybe 15 degrees or so above ambient. Of course the reading off the shiny copper stick-on heatsink is irrelevant (since its emissivity coefficient is different to a black surface) but hey at 99c that heatsink seemed a good insurance.
All the best with your new board and for 2014.
Anyway I use an IR thermometer on my projects - especially the PSU regulator heatsink and heatsink in Class-A amps etc. I pointed it at the "naked" XMOS and it was a bit warm - nothing too serious though - maybe 15 degrees or so above ambient. Of course the reading off the shiny copper stick-on heatsink is irrelevant (since its emissivity coefficient is different to a black surface) but hey at 99c that heatsink seemed a good insurance.
All the best with your new board and for 2014.
Last edited:
I´ve have some trouble with noise on the output of this rakkdac/ waveIO combo when I use my Squeezebox Touch as digital input. My combo looks like this:
Pic1.
If I use win7/ jriver media/ asio as digital input to WaveIO I have no trouble or noise, but with SBT (with EDO-applet) I get noise on the output of the rakkdac.
The noise looks like this on my scope when I run a 1kHz sinewave from the SBT at 44.1kHz samplrate:
Pic2
The SBT seems to recognize the waveIO card just fine, but if I scope the "I2S data" connection the frequency seems low, at around 300kHz. Bitclock and LRclock looks the same with SBT as with pc-input.
Any ideas that could help me sort this out?
Pic1.
If I use win7/ jriver media/ asio as digital input to WaveIO I have no trouble or noise, but with SBT (with EDO-applet) I get noise on the output of the rakkdac.
The noise looks like this on my scope when I run a 1kHz sinewave from the SBT at 44.1kHz samplrate:
Pic2
The SBT seems to recognize the waveIO card just fine, but if I scope the "I2S data" connection the frequency seems low, at around 300kHz. Bitclock and LRclock looks the same with SBT as with pc-input.
Any ideas that could help me sort this out?
Attachments
Don't want to sound very smart here, but I see you are using the supplied probe ground lead. To minimize interference I use something like shown here :
terminology - What is the name of this springy type oscilloscope probe accessory? - Electrical Engineering Stack Exchange
I just remove the original ground lead and use a few turns of bare copper wire. It does work against picking interference. I am not sure if it will help here, but you can try.
terminology - What is the name of this springy type oscilloscope probe accessory? - Electrical Engineering Stack Exchange
I just remove the original ground lead and use a few turns of bare copper wire. It does work against picking interference. I am not sure if it will help here, but you can try.
That is what I am going to try. I need to get the connectors and cable in house first.
Kind of strange that it works really well with win7/jriver/asio. I was up all night playing music. I worked with the SBT/ EDO trying different versions of the software/firmware and options for buffersize but noise remained the same.
First attempt, problems
My wave came in today, and I hooked it into my DAC (buffalo IIIse, placid HD, 10cm uFL cables. 5.05 V dc I had set the output to 580ma, measured a shunt current of 280 ma, so it was getting enough power. System is a Mac Mini, set Luckit as source, at 384, integer mode. Both the Audio/midi panel and Audirvana recognized it right away. Used a run of the mill 1 meter USB cable (will be eventually attaching to the board rather than using on board connector, with data and ground cable). Sound was staticky very much like a fading radio station. At moments it sounded okay, (not good, just not horrible) but especially on quiet passages the static overtook the signal.
Is this a symptom of too long uFL cables, or ?? Not sure the best way trouble shoot this. Shorter cables will take a week to get in; minimum I can fit in will be 6 to 7 cm. anything else worth looking at? I was surprised that the card only seemed to draw 300 ma, but this May be due to the signal it had to work with. Of course, I haven't tested the buffalo except via the toslink input, so there might be an issue there. Any suggestions would be appreciated.
My wave came in today, and I hooked it into my DAC (buffalo IIIse, placid HD, 10cm uFL cables. 5.05 V dc I had set the output to 580ma, measured a shunt current of 280 ma, so it was getting enough power. System is a Mac Mini, set Luckit as source, at 384, integer mode. Both the Audio/midi panel and Audirvana recognized it right away. Used a run of the mill 1 meter USB cable (will be eventually attaching to the board rather than using on board connector, with data and ground cable). Sound was staticky very much like a fading radio station. At moments it sounded okay, (not good, just not horrible) but especially on quiet passages the static overtook the signal.
Is this a symptom of too long uFL cables, or ?? Not sure the best way trouble shoot this. Shorter cables will take a week to get in; minimum I can fit in will be 6 to 7 cm. anything else worth looking at? I was surprised that the card only seemed to draw 300 ma, but this May be due to the signal it had to work with. Of course, I haven't tested the buffalo except via the toslink input, so there might be an issue there. Any suggestions would be appreciated.
Last edited:
Be sure your I2S wires are well away from any power wires. Shorter I2S lines are always better.