I assume you do those continuity checks with everything off, as otherwise you might blow up your multimeter. Maybe the output impedance of the 3.3 V buck converter on the FPGA module is low enough when powered down to fool your continuity check into reporting it as a short. I don't know, I just can't think of any other explanation at the moment.
Regarding the issue of the I2S not working:
A. There could be some mix-up with logic levels, like an active-low output that's expected to be active high, although this seems unlikely as you only connected the data and clocks
B. There could be a soldering issue with J4 pins 39, 41, 45, 47, 69, 71, 75 or 77 or with P9
C. I may have messed something up in configuration 3
Regarding A, the correct levels for PCM or DoP playback via the I2S input would be:
P9 pin 1: sel4192in, low (about 0 V)
P9 pin 3: mutei2sin, low (about 0 V)
P9 pin 5: datai2sin, switching between low and high, so average voltage around 1.65 V when playing music (0 V when playing perfect silence)
P9 pin 7: deemni2sin, usually high (about 3.3 V), but low only reduces treble
P9 pin 9: bcki2sin, switching between low and high, so average voltage around 1.65 V
P9 pin 11: lrcki2sin, switching between low and high, so average voltage around 1.65 V
P9 pin 13: dsd64_128, doesn't matter
P9 pin 15: dsdon, low
P9 pins 2, 4, 6, 8, 10, 12, 14, 16: ground, so also low
P9 pins 17, 18, 19, 20: 3.3 V supply pins, so 3.3 V
Regarding B, all inputs of P9 have internal pull-ups except dsdon and dsd64_128, which have internal pull-downs. That means that with nothing connected to P9, you should measure 3.3 V at pins 1, 3, 5, 7, 9 and 11. When you do, that confirms that they make contact with the FPGA module and are not shorted to ground, although there could still be shorts between these lines. Pins 13 and 15 should be at 0 V with nothing connected to them.
Regarding C, as far as I recall, I did try the I2S interface on an earlier version of version 3, the one with the headroom issues, but only tried the very latest configuration file with S/PDIF. As I didn't touch the I2S part at all, it would be strange if this were the cause, but you never know.
Regarding the issue of the I2S not working:
A. There could be some mix-up with logic levels, like an active-low output that's expected to be active high, although this seems unlikely as you only connected the data and clocks
B. There could be a soldering issue with J4 pins 39, 41, 45, 47, 69, 71, 75 or 77 or with P9
C. I may have messed something up in configuration 3
Regarding A, the correct levels for PCM or DoP playback via the I2S input would be:
P9 pin 1: sel4192in, low (about 0 V)
P9 pin 3: mutei2sin, low (about 0 V)
P9 pin 5: datai2sin, switching between low and high, so average voltage around 1.65 V when playing music (0 V when playing perfect silence)
P9 pin 7: deemni2sin, usually high (about 3.3 V), but low only reduces treble
P9 pin 9: bcki2sin, switching between low and high, so average voltage around 1.65 V
P9 pin 11: lrcki2sin, switching between low and high, so average voltage around 1.65 V
P9 pin 13: dsd64_128, doesn't matter
P9 pin 15: dsdon, low
P9 pins 2, 4, 6, 8, 10, 12, 14, 16: ground, so also low
P9 pins 17, 18, 19, 20: 3.3 V supply pins, so 3.3 V
Regarding B, all inputs of P9 have internal pull-ups except dsdon and dsd64_128, which have internal pull-downs. That means that with nothing connected to P9, you should measure 3.3 V at pins 1, 3, 5, 7, 9 and 11. When you do, that confirms that they make contact with the FPGA module and are not shorted to ground, although there could still be shorts between these lines. Pins 13 and 15 should be at 0 V with nothing connected to them.
Regarding C, as far as I recall, I did try the I2S interface on an earlier version of version 3, the one with the headroom issues, but only tried the very latest configuration file with S/PDIF. As I didn't touch the I2S part at all, it would be strange if this were the cause, but you never know.
Last edited:
J4 pin 47 was not making good contact, so I fixed that.
I have the FPGA installed and the WaveIO connected. Every voltage on P9 is correct, other than pin 7 which is reading low, though that shouldn't stop playback. Pin 5 is about 1.3V and seems to change based on the music. Pins 9 and 11 stay at 1.65V.
Do the lights on the FPGA always turn off eventually? I thought I noticed them turn off when I selected USB on my input selection switch, but it could have been coincidence. The lights turning off on the FPGA did not prevent the s/d and filter lights from working, and BNC input still played.
The WaveIO is detected in Windows and appears to work both through Windows Mixer and as an ASIO device. The driver is reporting when ASIO is active and the correct sample rate, and it's definitely outputting voltage on the I2S pins when music plays.
I have the FPGA installed and the WaveIO connected. Every voltage on P9 is correct, other than pin 7 which is reading low, though that shouldn't stop playback. Pin 5 is about 1.3V and seems to change based on the music. Pins 9 and 11 stay at 1.65V.
Do the lights on the FPGA always turn off eventually? I thought I noticed them turn off when I selected USB on my input selection switch, but it could have been coincidence. The lights turning off on the FPGA did not prevent the s/d and filter lights from working, and BNC input still played.
The WaveIO is detected in Windows and appears to work both through Windows Mixer and as an ASIO device. The driver is reporting when ASIO is active and the correct sample rate, and it's definitely outputting voltage on the I2S pins when music plays.
Last edited:
The lights on the FPGA module are supposed to indicate the sample rate:
LED D3: lights up when the sample rate exceeds 90 kHz or the raw DSD interface is activated
LED D5: lights up when the sample rate exceeds 100 kHz or the raw DSD interface is activated
LED D6: lights up when the sample rate exceeds 190 kHz or the raw DSD interface is activated
The raw DSD interface is activated means P9 pin 1 low and pin 15 high.
Does the WaveIO support raw DSD? If so, it's easy to check if raw DSD comes out with a makeshift low-pass filter.
LED D3: lights up when the sample rate exceeds 90 kHz or the raw DSD interface is activated
LED D5: lights up when the sample rate exceeds 100 kHz or the raw DSD interface is activated
LED D6: lights up when the sample rate exceeds 190 kHz or the raw DSD interface is activated
The raw DSD interface is activated means P9 pin 1 low and pin 15 high.
Does the WaveIO support raw DSD? If so, it's easy to check if raw DSD comes out with a makeshift low-pass filter.
The LEDs are in an undefined state until the firclock is up and running, which only happens after the 27 MHz oscillator has started. That is, they can be either on or off until the ECC81 and EF80s have heated up sufficiently.
I asked about the raw DSD because of the debugging possibilities it provides. When the WaveIO supports raw DSD and you connect RC low-pass filters to the word clock and data outputs, you should get audio at the filter outputs if the WaveIO works well. If that works, it would also be interesting to see if the valve DAC can work with the WaveIO in raw DSD mode.
I asked about the raw DSD because of the debugging possibilities it provides. When the WaveIO supports raw DSD and you connect RC low-pass filters to the word clock and data outputs, you should get audio at the filter outputs if the WaveIO works well. If that works, it would also be interesting to see if the valve DAC can work with the WaveIO in raw DSD mode.
Last edited:
Ok that makes sense as to what I saw. The lights turning off when switching the input must have been coincidence since I waited for the initial relay click to change them. It also explains why they stayed off, because I had nothing playing.
Taking a break for now, I guess I'll check every connection point again. I made sure all of the J4 pins which connect to P9 are good, the cable I made from the WaveIO to P9 and SW3 is good, P9 pin 3 is grounded, and all voltages are good (other than deemni2sin, which is reading low). Inspecting things is a hassle given my case setup lol. I'll disassemble again later.
EDIT: The WaveIO does support native DSD.
Taking a break for now, I guess I'll check every connection point again. I made sure all of the J4 pins which connect to P9 are good, the cable I made from the WaveIO to P9 and SW3 is good, P9 pin 3 is grounded, and all voltages are good (other than deemni2sin, which is reading low). Inspecting things is a hassle given my case setup lol. I'll disassemble again later.
EDIT: The WaveIO does support native DSD.
Last edited:
I connected the WaveIO into my BNC port using the isolated spdif output on the WaveIO, and it works. So the WaveIO is successfully outputting PCM from my PC, but I can't get it to play using I2S. I'm sure it's something easy I'm overlooking. It's possible all the removing of the FPGA has caused another issue with those headers, because my lights for Apodizing and Steep filters have stopped working, and there's continuity from LED to ValveDAC PCB. Changing to those filters triggers the relay and sound continues to play, so I don't think anything too serious happened
I'm a little concerned I may have damaged the FPGA, which is why the Halfband and Apodizing lights no longer work. The connection from J4 to R59 and R61 is fine, as is the connection from U10 and U12 to the respective LEDs. The LEDs are also working. When I switch to Halfband or Apodizing, I get no voltage at U10 and U12, all the others show 3.3V when active.
When I was doing diagnostic stuff before, I slipped and may have shorted something on P9 (or maybe it was P7?). The FPGA reset when this happened. Could this explain the now-dead lights?
Everything else works fine, other than i2s input. The DAC still plays music when set to Apodizing or Halfband, there is just no indication light.
When I was doing diagnostic stuff before, I slipped and may have shorted something on P9 (or maybe it was P7?). The FPGA reset when this happened. Could this explain the now-dead lights?
Everything else works fine, other than i2s input. The DAC still plays music when set to Apodizing or Halfband, there is just no indication light.
Last edited:
I don't know. You probably shorted a supply to ground, which would usually not cause damage because the regulators (switched mode or not) are usually protected against it. A short from the 5 V to the 3.3 V supply at P7 would be more dangerous, but the FPGA would probably also survive that as long as it doesn't last too long.
I just listened to a few seconds of Adriaan Alsema and the good-looking intellectuals through the I2S interface of my valve DAC running the very latest configuration file version 3. I would have loved to listen to some more Adriaan Alsema and the good-looking intellectuals, but as I had forgotten to ground the mute input, I had to manually ground it via test leads that I had to keep in place while listening to get any sound out.
Anyway, the I2S input works with version 3, or at least it does on my DAC using 44.1 kHz, 16 bit data coming from a Raspberry Pi model 3B. Another file, 96 kHz 24 bit, also works.
Anyway, the I2S input works with version 3, or at least it does on my DAC using 44.1 kHz, 16 bit data coming from a Raspberry Pi model 3B. Another file, 96 kHz 24 bit, also works.
Last edited:
I checked every pin of J4 and J5 to make sure they were connected to their intended destination. One ground and pin 73 of J5 were not connected properly.
Something I noticed: when attempting to use i2s, it's dead silent if P9 pin 3 is open, but as soon as I ground it out, I hear that the output is enabled. There's a very quiet hiss (due to cheap tubes, it changes if I move tubes). So I know that the output is enabling because I hear that slight hiss. I have a meet this weekend, I'll look at it more after that. A new bottom panel for my case is also coming in, so I'll have it totally disassembled to re-evaluate then. For now, using the spdif output of the WaveIO on a "standard" input of the ValveDAC. Sounds amazing!
Something I noticed: when attempting to use i2s, it's dead silent if P9 pin 3 is open, but as soon as I ground it out, I hear that the output is enabled. There's a very quiet hiss (due to cheap tubes, it changes if I move tubes). So I know that the output is enabling because I hear that slight hiss. I have a meet this weekend, I'll look at it more after that. A new bottom panel for my case is also coming in, so I'll have it totally disassembled to re-evaluate then. For now, using the spdif output of the WaveIO on a "standard" input of the ValveDAC. Sounds amazing!
There is a loss of bit clock detector of which the output signal gets combined with the mute input, so the fact that the mute input changes the output noise level also means that the DAC sees a bit clock - not necessarily a good bit clock, but at least something that switches frequently.
I think it would be interesting as an experiment to try the raw DSD mode. The raw DSD mode uses the word clock line for the right channel and the data line for the left channel, so if one channel works, the problem is narrowed down a bit further.
I think it would be interesting as an experiment to try the raw DSD mode. The raw DSD mode uses the word clock line for the right channel and the data line for the left channel, so if one channel works, the problem is narrowed down a bit further.
I've had a chance to listen more and can't wait to share this with people at the meet this weekend.
I've been using it with a TU-8800 tube amp and HD580 and Borealis headphones. The natural stage and timbre are really impressive. Has really great bass texture as well.
I'm just starting to mess with the filters and s/d settings. So far, the three filters are pretty tough to distinguish, but I can tell between chaos and PWM9. Maybe not in a blind test, but going back and forth, Chaos doesn't sound as resolving as PWM9, hard to describe. Chaos sounds a little "stuffy" almost? PWM5 sounds like PWM9, though these comparisons are very early.
I've been using it with a TU-8800 tube amp and HD580 and Borealis headphones. The natural stage and timbre are really impressive. Has really great bass texture as well.
I'm just starting to mess with the filters and s/d settings. So far, the three filters are pretty tough to distinguish, but I can tell between chaos and PWM9. Maybe not in a blind test, but going back and forth, Chaos doesn't sound as resolving as PWM9, hard to describe. Chaos sounds a little "stuffy" almost? PWM5 sounds like PWM9, though these comparisons are very early.
Brought my ValveDAC to a meet with tons of top tier gear, and it was one of the stars of the show. It paired beautifully with an Eddie Current Studio T and a few other amps. Had an A/B test setup comparing it to other DACs, such as the Dangerous Music Convert 2, and the common opinion was how natural the and presentation and decay of the ValveDAC sounded, without sacrificing technicalities. Couldn't be happier with it, other than wanting to put nice tubes in. Marcel, you impressed everyone at this meet, several even asked what it would cost for me to make them one lol