New Wondom ADAU1701 DSP Unit

[Audiomaniac3]

What do you mean by this? If I add gain to the output will it clip the RCA output?

Okay,made this video:

to make it clear, I have positive gain in my project +3.3db, but I still have plenty of headroom before clipping output signal, just put level detector in end of your project,run some sweeps or listen to music,and if it stays under 0db all is good.

Edit:Actually no more than -3db on signal detector scale,otherwise signal will be clipped.

 

[manokaiser]

Okay,made this video:

to make it clear, I have positive gain in my project +3.3db, but I still have plenty of headroom before clipping output signal, just put level detector in end of your project,run some sweeps or listen to music,and if it stays under 0db all is good.

Edit:Actually no more than -3db on signal detector scale,otherwise signal will be clipped.

Almost so. These level detectors are RMS level detectors:
"The Single-Level Detector w Numeric Display module calculates and displays the RMS level of the signal, shown in dB."
They do not display the Peak signal accurately, so if you see 0 dB on them in real music program you might still be clipping some higher level peak transients.

 

[manokaiser]

You can use REW's Tone generator to check for gain staging, as long as your source will be your PC after-all.
You can also check the internal gain with a Tone generator block in SigmaStudio at your input.
In either case you must set the generator at 0dB gain, play some sine wave tones at 1KHz and other frequencies. Your PC or any other source should also be at max level, or 0dB level. From there you can use several Level detectors, at the input, output and wherever gain is added to the signal so you have a full picture of the gain structure.

 

[Audiomaniac3]

You can use REW's Tone generator to check for gain staging, as long as your source will be your PC after-all.
You can also check the internal gain with a Tone generator block in SigmaStudio at your input.
In either case you must set the generator at 0dB gain, play some sine wave tones at 1KHz and other frequencies. Your PC or any other source should also be at max level, or 0dB level. From there you can use several Level detectors, at the input, output and wherever gain is added to the signal so you have a full picture of the gain structure.

Yep,done that,already,I've set tone generator to 0 db in rew and manually started selecting from 20hz-150hz,and strangly highest peak was at 75 hz and few other frequencies,under 75hz all other frequencies were almost the same on the level detector scale.

I have enough headroom,but probably I think it would be more beneficial,if I remoove gain from my project,and just crank subwoofer output from my receiver,currenly it is set to +2,maybe I'll need to put it +6 or +8.

 

[Audiomaniac3]

be at max level

Actually,my receiver onkyo840,when it receives multichannel audiotrack,it adds +3db to bass output,so when I measure my system in stereo mode on graph bass output is lower,so when I measure sub only it adds +3db and on my tests neither multichannel music,neither 0db tonesweeps didn't exceed -10db on level detector scale.

 

[Audiomaniac3]

input

Also I have this board in my hands:
https://store.sure-electronics.com/product/746

CS5343 Datasheet(PDF)

Part #: CS5343. Download. File Size: 370Kbytes. Page: 21 Pages. Description: 98 dB, 96 kHz, Multi-Bit Audio A/D Converter. Manufacturer: Cirrus Logic.

www.alldatasheet.com

Do you have any thoughts,would this analog to i2s converter board,be beneficial over stock adc input on adau1701 DSP?

 

[manokaiser]

Also I have this board in my hands:
https://store.sure-electronics.com/product/746

CS5343 Datasheet(PDF)

Part #: CS5343. Download. File Size: 370Kbytes. Page: 21 Pages. Description: 98 dB, 96 kHz, Multi-Bit Audio A/D Converter. Manufacturer: Cirrus Logic.

www.alldatasheet.com

Do you have any thoughts,would this analog to i2s converter board,be beneficial over stock adc input on adau1701 DSP?

There's not much information in the specifications of this board you linked.
ADAU1701 ADCs have SNR of 100dB and THD +N -83dB.

I have this board to play with :

*The Big Winter Deals* WONDOM OFFICIAL SHOP - Amplifier Board - Sure Electronics - ADAU1701 - 18650 charger - Sigmastudio

Sure Electronics is a professional audio amplifier supplier that provides electronics hobbyists and business partners the middleware which can help reduce customers' development difficulties and save their precious time-to-market by making the products directly released in the market.

store.sure-electronics.com

"High Audio Performance with SNR up to 119dB
Low Distortion with an Impressive THD+N as Low as 0.00047% and Noise Level as Low as 0.97uVrms
High Dynamic Range as 111dB."

This one in theory has an advantage on the ADC side. The DACs remain the limiting factor though, at 104dB SNR. I don't think that any of this improvement, if it actually plays out, would be really noticeable, though it could be measured with appropriate instruments.
The differential input could also be an advantage in some cases where EMI noise is becoming a factor.

 

[Audiomaniac3]

DACs remain the limiting factor though

The Svs SB2000 I own,have pretty good input sensitivity,so I guess for most subwoofers adau1701 with dac maxing out at 0.9v will be enough.

Well there is always room to grow,and probably Wondom will release their DSP unit V2 with adau1452 on board and it will be really unique and exceptional item.

 

[edc]

Jumping in with some info: ADAU1701 is capable of using FIR filters exported from REW, there's no need for PEQ filters to EQ frequency response.

It's limited to about 330 samples per input / 6ms IR window. If ran in mono then you can have 660 sample which covers the full range.

More details here: https://audiosciencereview.com/foru...-baffle-for-coaxial-driver.60019/post-2204997

 

[Audiomaniac3]

ADAU1701 is capable of using FIR filters

I've made Fir crossovers with Rew and rephase:

But for adjusting bass frequency response even 1000 taps is not enough,so I doubt that anyone will use it in mono mode,it will be better just to use equalizer apo on PC with any amount of taps.

 

[edc]

I've made Fir crossovers with Rew and rephase:

That's cool. I didn't do the REW EQ and rephase part and just inverted the original IR response to get linear results. Didn't pay too much attention to phase, so I'll follow your method as well to see if it sounds different. Thanks for the video!

 

[edc]

I designed cases for the Wondom JAB4 and ICP5, if anyone has access to a 3D printer and has one of these boards.

Wondom ICP5 case, two parts, press fit without screws.

ICP5 .STL Download (Google Drive)

Wondom JAB4 case, two parts, secured using 4 screws (UNC6-32/M3.5 or use a drill to enlarge.) This case has airflow openings + room for heatsinks and can be mounted to a speaker using the holes on the back (50mm between horizontal/vertical mount points)

JAB4 .STL Download (Google Drive)

I'll make another case for the APM2 once it arrives.

 

[Kyleag89]

I have a wondom adau1701 DSP the one that comes in an aluminum enclosure that's a 2 in 4 out. I'm wondering if there's a way to get more output from it because the adau1701 is supposed to be capable of 0.9 rms and I'm getting less output than I put in. If I put 600mv signal input I only get around 350mv out. Has anyone changed the gain of the sgm8904 line drivers on the board? If I identified the correct resistors that set the gain then it's only set to a gain of 1.4. There's a confusing arrangement of resistors and capacitors around the 8904 ICs and they're all smd.

 

[Kyleag89]

The gain settings in sigma studio for the adau1701 are already set to +5db and +10db so I don't understand where all the loss of signal is happening.

 

[Audiomaniac3]

The gain settings in sigma studio for the adau1701 are already set to +5db and +10db so I don't understand where all the loss of signal is happening

Maybe it's because of the stock project,try making simple project with input and output and level detector and see what is happening,also on the back there is relative gain (pot5)
Maybe try adjusting it.

 

[Weeb Labs]

Another interesting unit has appeared on AliExpress. It appears to offer multichannel output via USB audio, with per-channel PEQ.

 

[FrankV]

The gain settings in sigma studio for the adau1701 are already set to +5db and +10db so I don't understand where all the loss of signal is happening.

The unit uses 18k resistors on the input to the ADC which gives 7dB attenuation for the ability to input up to 2V RMS. If 2V RMS handling is not required, the input resistors (R108 & R111) could be reduced to 7k for example, which would allow for a maximum input of 0.9V RMS. See Table 13, page 20 of the ADAU1701 datasheet https://www.analog.com/media/en/technical-documentation/data-sheets/ADAU1701.pdf

The SGM8904's in the outputs provide low-pass filtering and also around 3.5dB gain, resulting in a net throughput loss of around 3.5dB. Unfortunately, the Volume control ic is well under voltage supplied and is the limiting factor such that the DSP DAC cannot be driven to full scale without the input circuit in the volume control ic distorting. It is clear it is the input as reducing the volume setting does not change the distortion. I dropped the gain of the output buffers down on mine (solder a 20k resistor across R301, R321, R310 and the equivalent one in the fourth stage (I can't see a board marking for its reference number). This allows full scale output from the DACs without the distortion but doesn't help with much else. After this mod, adding +7dB digital gain in the DSP to the input signal will at least give unity gain input to output but will not allow more than 1V RMS at the output sockets before clipping.

Ideally, the volume stage requires a higher voltage supply, then the buffer gains could be increased to +7dB, rather than decreased, for 2V in to out capability.

 

[mcdn]

Hi all! I bought one of the Wondom units to use in my garage (just basic main/sub splitting and some PEQ), but I'm going to give it a thorough test first. Some of my reviews have got onto the front page so I wanted to ask in advance to see what interested people would like me to measure/check.

 

[FrankV]

Hi all! I bought one of the Wondom units to use in my garage (just basic main/sub splitting and some PEQ), but I'm going to give it a thorough test first. Some of my reviews have got onto the front page so I wanted to ask in advance to see what interested people would like me to measure/check.

Check out the load driving capability. I had to substantially modify a friend's unit as he was driving into a couple of Topping PA5 ii power amplifiers and noticed heavy distortion setting in at less than 300mV (RMS) output. This was down to a bad combination of low input impedance on the PA5 ii (estimated at less than 2k) plus poor drive capability from the PT2259 volume control ic in the Wondom output stage. I thought at first it was down to the PT2259 being supplied at 4.6V (at best, the PSU filtering in the Wondom is also a problem) but having removed them to fix my friend's problem I bench tested the PT2259 at it its recommended 10V power supply and it did not help a lot with load drive. To be fair, the PT2259 has a minimum output load specification of 6k ohms, which makes the Wondom a poor choice for directly driving a lot of power amplifiers.

If anyone is stuck with one of these Wondom units driving a low impedance amplifier and able to work with surface mount components then the best modification is to remove the two PT2259 ics. Alternatively, they could be left in situ and isolated by removing the 10uF series input and output capacitors. 100uF leaded capacitors can then be used to bridge input to output which effectively makes the SGM8904's be the output drivers. This is fine because that is what they are designed to do! They are specified to drive 3V RMS into a 2.5k load with a 5V power supply, they have their own internal charge-pump supply rail generators.

Another problem my friend noticed was a mechanical noise plus some noise on the output when sometimes powering up. I discovered this was down to the inductors used in the power regulating circuits. There are four of these and their resistance is too high for the current being taken by the Wondom circuitry. This was causing the first regulator in the DSP PSU supply chain to be fed with less input voltage than it was supposed to give output voltage (4.6v) and this was occasionally causing it to oscillate. Simplest fix was to add 1.8 ohm resistors across the first two inductors (the two nearest the rear) which then allowed the regulators to actually regulate.

I would expect the unit to provide unity gain insertion performance. As mentioned in my previous post, Wondom chose to use 18k input resistors which attenuates the input by 7dB and then to add around 3.5dB output gain with the SGM8904's. This is not good for end to end background noise performance as the 7dB loss on the input has to be made up for on the output and following amplifier, which means 7dB amplification of the ADC and DAC noise. Best to drop the 18k to 7k and reduce the output gain around the SGM8904's to unity as mentioned in my previous post.

More detail on the above can be provided if requested.

 

[mcdn]

@FrankV thank you! I’ll definitely test with a variety of loads, but if 6k is spec then I won’t mark it down for issues below that . I’ll see what I can do about measuring coil whine as well.