Can you share the schematic(s) for the mono PCM1794A DAC's/IV convertors?
I assume that the SPDIF receiver design is Bryan's? I wonder why external buffers were used instead of using the 8 dedicated SPDIF internal to the WM8805?
No need for a transformer isolated SPDIF input?
Inputs should be terminated in 75 ohm R if using 75 ohm coax.
Phono pre-amp design for those that want one?
4+1 internal analog inputs might be a bit short, I guess it depends on each user and their setup.
What instrument did you use to measure 25uV of noise?
I assume that the SPDIF receiver design is Bryan's? I wonder why external buffers were used instead of using the 8 dedicated SPDIF internal to the WM8805?
No need for a transformer isolated SPDIF input?
Inputs should be terminated in 75 ohm R if using 75 ohm coax.
Phono pre-amp design for those that want one?
4+1 internal analog inputs might be a bit short, I guess it depends on each user and their setup.
What instrument did you use to measure 25uV of noise?
Last edited:
We're still working on the DAC I/V converters. Valery's having trouble finding any useful info on the output signal from the DAC so we're prototyping and measuring for our self, but it's slow going.
I designed the SPDIF switch. WM8804/5 becomes stupid in software mode, much better IC in hardware mode.
The coax inputs are Rod Elliot's design. They are 75 ohm impedance. No transformer required. This is a prototype to see if the design will operate with all coax input SPDIF sources. I have a feeling it might not work with weak sources.
4 analog inputs is lots for most people these days. Everything is digital now. Phono preamp is coming.
I used a Rigol DM3058E to measure output noise.
I designed the SPDIF switch. WM8804/5 becomes stupid in software mode, much better IC in hardware mode.
The coax inputs are Rod Elliot's design. They are 75 ohm impedance. No transformer required. This is a prototype to see if the design will operate with all coax input SPDIF sources. I have a feeling it might not work with weak sources.
4 analog inputs is lots for most people these days. Everything is digital now. Phono preamp is coming.
I used a Rigol DM3058E to measure output noise.
I got WM8804 mixed up with WM8805.
Yeah I guess making a discrete I/V is a lot more involved than just following the app note examples
I do wonder how one can tell/hear the difference in performance between discrete and opamp implementations.
I have not looked at this digital audio stuff in a long while. I know Cirrus bought out Wolfson. I thought that they were going to discontinue CS8406,8416, not the case.
I do see that Onsemi is sell the Sanyo part LC89091JA-H at a fraction of the cost of a WM8804, wonder what the differences are, have to look further, no H/W mode only S/W mode, maybe not brain dead though. Lots of work to evaluate each offering.
I am wrong, LC89091JA-H offers a H?W op mode.
Yeah I guess making a discrete I/V is a lot more involved than just following the app note examples
I do wonder how one can tell/hear the difference in performance between discrete and opamp implementations.I have not looked at this digital audio stuff in a long while. I know Cirrus bought out Wolfson. I thought that they were going to discontinue CS8406,8416, not the case.
I do see that Onsemi is sell the Sanyo part LC89091JA-H at a fraction of the cost of a WM8804, wonder what the differences are, have to look further, no H/W mode only S/W mode, maybe not brain dead though. Lots of work to evaluate each offering.
I am wrong, LC89091JA-H offers a H?W op mode.
Last edited:
I built a first prototype for the DAC following the datasheets. Output was very low and the op amps were running far too hot for my liking. We've since figured out why the output was low. They don't tell you enough on the datasheet to be useful. Their suggested 10k resistor for reference current is way off.
Makes you wonder how they can document something that does not work per the datasheet app note.
Did you look at
http://www.ti.com/lit/an/sboa237/sboa237.pdf
Did you look at
http://www.ti.com/lit/an/sboa237/sboa237.pdf
Hi Rick!
Thanks a lot for the link - we have found out a number of things, hitting some issues and finding the solutions, collecting information bit by bit from different sources - this paper finall confirms we were moving in the right direction
We were just exchanging our thoughts with Jeff about a few surprises, related to the way our custom I/V converters work, coming to some conclusions, when your link came in to clarify those points.
Now we know for sure how to make the thing working properly
Cheets,
Valery
Hi Jeff,
So that's what you've been busy at!
I am interested with your project. As always, my bench is your bench if you ever need. One big change is I switched to LED overhead lights, so distortion readings don't have to be done in the dark anymore.
Just call, I'll make the time for you. I agree with your desire to have a better quality of optical encoder - strongly!
-Chris
So that's what you've been busy at!
I am interested with your project. As always, my bench is your bench if you ever need. One big change is I switched to LED overhead lights, so distortion readings don't have to be done in the dark anymore.
Just call, I'll make the time for you. I agree with your desire to have a better quality of optical encoder - strongly!
-Chris
Hi Valery and Jeff
Glad I can help out in a small way, can never have too much design use information.
TI also has a AEC-Q100 qualified part, the DS for that details their recommended layout guideline as shown on p17 and maybe some other info since they had a second crack at a DS.
http://www.ti.com/lit/ds/symlink/pcm1794a-q1.pdf
I am puzzled as to why the opamps are running so hot in the PCM1794A application ckt?
What was the opamp supply voltages?
The same ckt is used in DEM-PCM1792, DEM-DSD1792, DEM-PCM1794, DEM-DSD1794 EVM User's Guide
Glad I can help out in a small way, can never have too much design use information.
TI also has a AEC-Q100 qualified part, the DS for that details their recommended layout guideline as shown on p17 and maybe some other info since they had a second crack at a DS.
http://www.ti.com/lit/ds/symlink/pcm1794a-q1.pdf
I am puzzled as to why the opamps are running so hot in the PCM1794A application ckt?
What was the opamp supply voltages?
The same ckt is used in DEM-PCM1792, DEM-DSD1792, DEM-PCM1794, DEM-DSD1794 EVM User's Guide
HI Chris. I'll likely be taking you up on this offer soon. Hopefully I can have some panel ready for your rack this time.
Hi Jeff,
You mentioned to me that you plan on using the Si4703 FM chip.
Nothing inherently wrong with the part as it is part of the Si47xx family. It uses the same programming guide (AN332) so they are s/w compatible.
The Si4703 has the same basic specs as a Si4735 but the Si473x family offers a few features that might be of benefit, Si4735-D60 is available in a SSOP package for easier soldering over a DFN. The Si4735-D60 offers a digital audio interface which could be used with your high end DAC.
One design issue I found with the Si473x digital I/F is that it works in I2S slave mode. Thus it requires a master device for its interface. I was thinking that one solution is to convert the I2S to SPDIF, so that it can be used as another digital audio source for the SPDIF digital inputs. Thus one could use a SPDIF transmitter like a the TI DIT4096/4196 to be the I2S master for the Si4735 device.
P.S. My Win XP computer died so I have got a Win10-64 computer going and have been playing with DIPtrace, since I can not get my old orcad 16 to run on win10
Rick
You mentioned to me that you plan on using the Si4703 FM chip.
Nothing inherently wrong with the part as it is part of the Si47xx family. It uses the same programming guide (AN332) so they are s/w compatible.
The Si4703 has the same basic specs as a Si4735 but the Si473x family offers a few features that might be of benefit, Si4735-D60 is available in a SSOP package for easier soldering over a DFN. The Si4735-D60 offers a digital audio interface which could be used with your high end DAC.
One design issue I found with the Si473x digital I/F is that it works in I2S slave mode. Thus it requires a master device for its interface. I was thinking that one solution is to convert the I2S to SPDIF, so that it can be used as another digital audio source for the SPDIF digital inputs. Thus one could use a SPDIF transmitter like a the TI DIT4096/4196 to be the I2S master for the Si4735 device.
P.S. My Win XP computer died so I have got a Win10-64 computer going and have been playing with DIPtrace, since I can not get my old orcad 16 to run on win10
Rick
I'm sill looking at different options for tuners. I have a breakout board for the SI4703 to experiment with though. I was thinking of going I2S straight into the DAC and skipping SPDIF completely. I want to do this with USB audio as well.
I actually like soldering QFN. I find it doesn't bridge as easily. I use a CP2102 in my serial / USB adapter.
I actually like soldering QFN. I find it doesn't bridge as easily. I use a CP2102 in my serial / USB adapter.
Soldering a DFN/QFN, I guess it depends who is doing the soldering, is the project plan to be built by others as in DIY?
I'd like to hear about how you plan to design the interface for two I2S slaves tied together? I thought about it, determined I would need a DSP or FPGA to do the task. A lot more involved than a DIT4096.
I used the PCM2902B/C on the portable, it can act as a USB/SPDIF bridge. Worked exactly as published in the DS and eval pcb from TI. I found if I looped back the SPDIF, I had a way of recording streaming audio with wavelab or similar s/w.
I'd like to hear about how you plan to design the interface for two I2S slaves tied together? I thought about it, determined I would need a DSP or FPGA to do the task. A lot more involved than a DIT4096.
I used the PCM2902B/C on the portable, it can act as a USB/SPDIF bridge. Worked exactly as published in the DS and eval pcb from TI. I found if I looped back the SPDIF, I had a way of recording streaming audio with wavelab or similar s/w.
Anyone can solder QFN as long as the layout is done properly. just extend the pads out for enough to get an iron on them. There are lots of my serial/USB adapters out there. I've never heard any complaints about assembling them.
74VHCT244 is popular for switching I2S.
74VHCT244 is popular for switching I2S.
Attachments
Unless I do not understand how it works, I think that you are missing my point about I2S master/slave operation.
For example the PCM1794A acts as a I2S slave, the bit clock is an i/p.
The SPDIF receiver WM8804 is configured as a master, thus it generates the I2S bit clock, for the PCM1794A DAC, but as an example
The Si4735 is a I2S slave, its DCLK is also an input, what device in this case generates the bit clock? since you can't tie to I2S clock inputs together without a ref clock source.