Fixed the connection, had a loose wire. Is there a default project for the ADAU1452_DSP CS42448 combo ?
Did some more digging, I see that the 1452 uses the default settings of the CODEC and will have to find my own way to control the CS42448. Managed to execute a crossover for the Directiva R1 speaker and seems to be working OK. The info on this forum is certainly adequate the moment I took the time to dig into the PCB schematics and the 1452 specs.
The xover pipeline uses General 2'nd order IIR filters fed with biquad coefficients from Vituix simulation. I use a differential output to negate common mode noise and gain signal.
PC --> USB_audio-coax-->1452_coax_in-->AsrcIn[1:0]-->Filters.SecondOrder.NoSlew.GeneralEq(2nd order)-->Positive/Negative-->Ouput[5:6]
Thank you all
Hi, hi have both boards: 1452+ad1938 and 1452+cs42448. Actually only the 1938 is working at 48khz. Can you explain me more the following?Did some more digging, I see that the 1452 uses the default settings of the CODEC and will have to find my own way to control the CS42448. Managed to execute a crossover for the Directiva R1 speaker and seems to be working OK. The info on this forum is certainly adequate the moment I took the time to dig into the PCB schematics and the 1452 specs.
The xover pipeline uses General 2'nd order IIR filters fed with biquad coefficients from Vituix simulation. I use a differential output to negate common mode noise and gain signal.
PC --> USB_audio-coax-->1452_coax_in-->AsrcIn[1:0]-->Filters.SecondOrder.NoSlew.GeneralEq(2nd order)-->Positive/Negative-->Ouput[5:6]
Thank you all
How do you "control" your codec? have you added some extra commands to manually enable 192khz (or others samplibg rates)after fw downloading on dsp?
What frequency do you use in sigma studio project?
Hi @Flavio88 , I ran 1452+cs42448 at 96KHz and 192KHz. Did not meddle with DAC controls as I reached the conclusion that there's no need. There are no controllable filters/features for the DAC worthy of an effort. Seems the DAC is working in sample rate auto-detect mode by default, 192KHz being "QSM" or quad speed mode, DAC filter de-emphasis is disabled by default for the sampling rates I used (96, 192)
At the moment the only thing that is worth tinkering is to patch the PCB to output differential outputs of the DACs and gain a (not negligible) 3dB across THD and also DNR
At the moment the only thing that is worth tinkering is to patch the PCB to output differential outputs of the DACs and gain a (not negligible) 3dB across THD and also DNR
Hey julbo, I have the same combo 1452 + cs42448 but i cant manage to get sound out of it
Is possible to send a working "demo" project based on this combo? Thanks a lot, SorinHi Sorin,
I contacted directly the vendor that I purchased the board from - nvarcher on aliexpress - and they kindly sent me a link to download the package from. Mind that your firmware might be different than mine, slim chance as these boards are pretty generic. In any case if you can't get a hold of the vendor to send you the files I can upload into your storage. I don't have this file in the cloud, it's ~0.5GByte compressed. I can say that if you follow the steps described in this thread, it works pretty good
I contacted directly the vendor that I purchased the board from - nvarcher on aliexpress - and they kindly sent me a link to download the package from. Mind that your firmware might be different than mine, slim chance as these boards are pretty generic. In any case if you can't get a hold of the vendor to send you the files I can upload into your storage. I don't have this file in the cloud, it's ~0.5GByte compressed. I can say that if you follow the steps described in this thread, it works pretty good
hi, I bought it from a different aliexpress vendor that wont reply to my private message..can you please upload it to a free service like wetransfer? I am a new user and seems that i cannot send forum private messages unfortunately.
Ok, i noticed the same about autosetting while reading cs42448 datasheet. Still no sound from the board. I think my project has errors but still don't understand where.
I precise that in my project i can see level detectors on SPDIF input moving so i think that the top part (dsp) of the board is correctly working. But no sound from analog outputs.
I attach my base terst project. Can you check if the yours is configured correctly? Or just post here your working project.
@mambus feel free to try my project and if you hear white noise from the outputs.... just to see if my DAC is clearly broken....
Thank you Flavio88, will try and post backOk, i noticed the same about autosetting while reading cs42448 datasheet. Still no sound from the board. I think my project has errors but still don't understand where.
I precise that in my project i can see level detectors on SPDIF input moving so i think that the top part (dsp) of the board is correctly working. But no sound from analog outputs.
I attach my base terst project. Can you check if the yours is configured correctly? Or just post here your working project.
@mambus feel free to try my project and if you hear white noise from the outputs.... just to see if my DAC is clearly broken....
Attached an active xover project for the ASR Directiva r1 open source platform speaker, very cool project with lots of learningOk, i noticed the same about autosetting while reading cs42448 datasheet. Still no sound from the board. I think my project has errors but still don't understand where.
I precise that in my project i can see level detectors on SPDIF input moving so i think that the top part (dsp) of the board is correctly working. But no sound from analog outputs.
I attach my base terst project. Can you check if the yours is configured correctly? Or just post here your working project.
@mambus feel free to try my project and if you hear white noise from the outputs.... just to see if my DAC is clearly broken....
https://www.audiosciencereview.com/...1-speaker-build-using-denovo-flat-pack.23535/
I checked and rechecked the DAC outputs by "brute force attack": instantiated a sine waveform generator in SigmaStudio, connected it to each DAC output and sampled each of the terminals with a scope. Also roughly checked DAC INL/DNL, co-channel interference, making sure there are no "dead bits", clamping, or any non-linearity nastiness. I did a first order mod, connected the DSP board and DAC board to separate power supplies; mainly to gain a slightly cleaner supply for the DAC - that's where it matters
Last edited:
Noise can source from a myriad of reasons. Try the following:
1. In sigma studio modify the default project and ground all the DAC inputs so no digital signal is output. If you still have noise then the DAC stage power supply is noisy or DAC is faulty.
2. Goes without saying check your amp without the dsp. Disconnect dsp , try short inputs to gnd
3. If 1 is ok, instantiate a sine wave and generate sound through the dsp at various frequencies amplitude to reach all DAC bit range - this should rule out issues with dac connectivity
1. In sigma studio modify the default project and ground all the DAC inputs so no digital signal is output. If you still have noise then the DAC stage power supply is noisy or DAC is faulty.
2. Goes without saying check your amp without the dsp. Disconnect dsp , try short inputs to gnd
3. If 1 is ok, instantiate a sine wave and generate sound through the dsp at various frequencies amplitude to reach all DAC bit range - this should rule out issues with dac connectivity
Hi,
i powered even from an external battery bank so is not likely is a power issue.
The amp is dead silent with the input shorted at full volume.
I removed also the analog inputs from the project, left only a sine wave generator, same thing.
So i think is clearly a DAC issue, I wonder..does this thing requires a metal casing? Maybe wifi interference or something..
Casing as a "Faraday cage" may help against RFI/EMI
I am powering the DAC and DSP boards with separate DC2DC converters, I do have some noise from the tweeter that is not audible at say 1.5m.
All the DAC outputs suffer from the same level of noise?
Can you look with a scope on the DAC output signal while digitally it is fed with constant 0x0000 and show us what you're measuring?
Huge noise might indicate a floating ground, missing/bad/wrong passive or bad power decoupling. Both boards have their own voltage regulator, try to measure the DC and AC levels on the regulators' output. Scope with AC coupling on the reg output may also shed some light whether it's a decap issue.
Your amplifier might be too sensitive (pre-amp stage?) with input range or impedance mismatch.
I think the DAC board outputs through passive filters (no opamps), not ideal driving long cables and an amp that might be a bit fussy.
Isn't DIY wonderful
I am powering the DAC and DSP boards with separate DC2DC converters, I do have some noise from the tweeter that is not audible at say 1.5m.
All the DAC outputs suffer from the same level of noise?
Can you look with a scope on the DAC output signal while digitally it is fed with constant 0x0000 and show us what you're measuring?
Huge noise might indicate a floating ground, missing/bad/wrong passive or bad power decoupling. Both boards have their own voltage regulator, try to measure the DC and AC levels on the regulators' output. Scope with AC coupling on the reg output may also shed some light whether it's a decap issue.
Your amplifier might be too sensitive (pre-amp stage?) with input range or impedance mismatch.
I think the DAC board outputs through passive filters (no opamps), not ideal driving long cables and an amp that might be a bit fussy.
Isn't DIY wonderful
Last edited:
So after some days of trial and error I have bypassed the internal AMS1117 3v3 regulators, and used a external 3v3 regulator and by doing this much of the noise have cleared up. Now I end up with a board that has audible noise at about 1 - 1.5m from the tweeter ( much like you said ) . Problem is that even in this setup, I measured the THD+N and the results are pretty bad... -50db Noise floor...the results with the built-in regulators is -25db lol... I must specify that the soundcard loopback is -95db..I have used REW for the measurements. So I think the analog board must have some design flaws...yea DIY is wonderful )) when you're not crying
EDIT: I must mention that the results are at 100Hz test tone ( Sigma Internal Generated ) , when I go up the results tend to get a little better
)) when you're not cryingEDIT: I must mention that the results are at 100Hz test tone ( Sigma Internal Generated ) , when I go up the results tend to get a little better
Last edited: