New Wondom ADAU1701 DSP Unit

[net-david]

Have you tried it yet? I also need 5v for an aditional DAC .

@Hue was able to test his at 5v, see post 44 above.

 

[net-david]

I believe I traced the SIGNAL LED on mine back to one of the audio output pins on the ADAU1701. It only monitors one audio channel, OUTB R. I would have thought they would have used one of the ADAU’s GPIO pins, but there wasn’t anything in the SigmaStudio project Wondom posted.

Please ignore this, it is apparently completely wrong and too late for me to edit it.

 

[Weeb Labs]

My Nvarcher unit arrived this morning. This little board has really impressed me.

It is capable of driving both my Sundaras and KSC75s to very reasonable levels with EQ. To my knowledge, this is the only USB DAC with a headphone amplifier (~40mW into 32 Ohms per datasheet), subwoofer crossover, Bluetooth (with TWS pairing), line level input (accessible via USB) and 40 bands of PEQ available under $15. The closest alternatives (MiniDSP IL-DSP and Quidelix-5K) both cost more than ten times as much.

This board represents absurd value for money. I'm going to order a few more!

 

[net-david]

Bypassing the voltage regulator is by far the easiest route to take. A jumper across connecting the two inductors (after removal of U203) does supply 5V to the analog output stage. Clipping still occurs around the final output line drivers. Is it possible there are diodes connected across the audio output lines? Those 3 legged devices might be diodes, or are they transistors?? With the 5 v modification the audio output is much higher on the stage prior to the output liner driver. I may opt to rewire the RCA output jacks accordingly (possibly with a series capacitor to preclude DC on the jack).

@Hue, thanks for testing this!

I think we need to optimize the gain structure of the ADAU1701 to SGM8904 to PT2259 signal chain. I've ordered a "3.7V to 12V Mini DC-DC Boost Module Support 5V/8V/9V/12V Output" power supply from eBay so that I can get 9v to the PT2259s.

The SGM8904s have a built in charge pump power supply and may be overdriving the inputs of the PT2259s even when they are powered at 5v.

The ADAU1701 is rated for 0.9Vrms full-scale output, if we want a 2v output we need to set the gain of the output stages to 2.22. I'm seeing a 470ohm input resistor and an 8.2k feedback resistor on the SGM8904 which would give a gain of 17.4x? Although I haven't traced out what all of the capacitors around the SGMs are doing, perhaps its an analog low pass filter?

 

[Weeb Labs]

Testing the output of this board with various headphones confirms my suspicion that the output coupling capacitors are undersized. They were not selected with the goal of utilizing all of the DSP's output capabilities in mind. Low impedance headphones suffer from significant low frequency rolloff. There is only 1.5V of bias on the output, so paralleling a pair of 10V 220uF electrolytics solves the problem and brings the RC cutoff down to 20Hz for 32 Ohm headphones. They also happen to fit perfectly between the 3.5mm jack housings, which is convenient.

With this modification, it drives my Sundaras to very reasonable volumes with EQ. I will be printing a simple enclosure for it.

 

[ClayR]

SPDIF would be nice but the ADAU1701 does not have onboard SPDIF RX/TX. Its ADCs and DACs (even cascaded) are perfectly adequate for almost any purpose and have been measured here. Four I2S IO are potentially internally accessible but this unit targets those in search of a more "plug and play" solution. It is effectively a cheaper, vastly more functional MiniDSP 2x4 (which also makes use of the ADAU1701).

Any opinions on using this with one of those I2S IO's to make a 3 way active crossover? https://store.sure-electronics.com/product/745 I plan to set up POTS 1,2 and 3 to control levels on each band. How does the CS4334 compare to the internal DAC's?

 

[Weeb Labs]

Any opinions on using this with one of those I2S IO's to make a 3 way active crossover? https://store.sure-electronics.com/product/745 I plan to set up POTS 1,2 and 3 to control levels on each band. How does the CS4334 compare to the internal DAC's?

If you need to make use of the I2S output, then you will no longer be able to use the leftmost potentiometer. Both are configurable functions of the same pin.

 

[ClayR]

My unit has the 2 switches labeled as
MP0 "SDATA_IN0/TDM_IN" and MP6 "SDATA_OUT0/TDM_OUT". The POT's are MP3,8 and 9. I was also looking at maybe getting one of these 2 board, since those pins are DTM. Note that store on Ali is kind of wonky. The top of the page lists various combo options but the bottom is the board I'm looking at by it self.

https://www.aliexpress.us/item/3256...PNZnEtUs&utparam-url=scene:search|query_from:

https://www.aliexpress.us/item/3256805939591185.html?spm=a2g0o.productlist.main.1.1457490498b8KU&algo_pvid=8df41644-2d0f-4605-8a82-03fd0f69cc3a&algo_exp_id=8df41644-2d0f-4605-8a82-03fd0f69cc3a-0&pdp_npi=4%40dis%21USD%2145.65%2138.8%21%21%2145.65%21%21%402103224117045657954253895ee451%2112000035871440066%21sea%21US%21109772322%21&curPageLogUid=U9JNPNZnEtUs&utparam-url=scene%3Asearch%7Cquery_from%3A

 

[ClayR]

https://www.aliexpress.us/item/3256805939591185.html?spm=a2g0o.productlist.main.1.1457490498b8KU&algo_pvid=8df41644-2d0f-4605-8a82-03fd0f69cc3a&algo_exp_id=8df41644-2d0f-4605-8a82-03fd0f69cc3a-0&pdp_npi=4%40dis%21USD%2145.65%2138.8%21%21%2145.65%21%21%402103224117045657954253895ee451%2112000035871440066%21sea%21US%21109772322%21&curPageLogUid=U9JNPNZnEtUs&utparam-url=scene%3Asearch%7Cquery_from%3A

 

[ClayR]

I would have to cut the switch or switches out and solder to the pins there.

 

[MCH]

My Nvarcher unit arrived this morning. This little board has really impressed me.

It is capable of driving both my Sundaras and KSC75s to very reasonable levels with EQ. To my knowledge, this is the only USB DAC with a headphone amplifier (~40mW into 32 Ohms per datasheet), subwoofer crossover, Bluetooth (with TWS pairing), line level input (accessible via USB) and 40 bands of PEQ available under $15. The closest alternatives (MiniDSP IL-DSP and Quidelix-5K) both cost more than ten times as much.

This board represents absurd value for money. I'm going to order a few more!

with this one, do you still need a USBi programmer or can you just plug it to your computer via USB?
(i have been following your finds, but so far resisted because i have no use for them, but this one is simply too good to pass, thanks!)

 

[Weeb Labs]

with this one, do you still need a USBi programmer or can you just plug it to your computer via USB?
(i have been following your finds, but so far resisted because i have no use for them, but this one is simply too good to pass, thanks!)

USBi is not required. It is configured directly via USB (while also supporting simultaneous USB audio), using ACPWorkbench.

 

[ClayR]

USBi is not required. It is configured directly via USB (while also supporting simultaneous USB audio), using ACPWorkbench.How is the

Testing the output of this board with various headphones confirms my suspicion that the output coupling capacitors are undersized. They were not selected with the goal of utilizing all of the DSP's output capabilities in mind. Low impedance headphones suffer from significant low frequency rolloff. There is only 1.5V of bias on the output, so paralleling a pair of 10V 220uF electrolytics solves the problem and brings the RC cutoff down to 20Hz for 32 Ohm headphones. They also happen to fit perfectly between the 3.5mm jack housings, which is convenient.

With this modification, it drives my Sundaras to very reasonable volumes with EQ. I will be printing a simple enclosure for it.

Any plans for more testing? I'm wondering how clean the output is. How well does it work with an amp? Would a 10k input amp need a caps? On a 2.1 system I would think only the sub would need one, if at all.

 

[Weeb Labs]

Any plans for more testing? I'm wondering how clean the output is. How well does it work with an amp? Would a 10k input amp need a caps? On a 2.1 system I would think only the sub would need one, if at all.

Performance is about what one might expect. Schmoo is just under 80dB, with the usual spurious garbage.

Things become quite interesting at 40Hz, with a relatively loud 4KHz spur appearing. This can actually be heard accompanying loud kick drums if listening very carefully.

 

[ClayR]

I'm not following, since the 40hz peak and 4k spike appear unrelated on the graphs. But I take it you would not recoment we purchase one of thses?

 

[Weeb Labs]

I'm not following, since the 40hz peak and 4k spike appear unrelated on the graphs. But I take it you would not recoment we purchase one of thses?

The 4KHz spur decreases in amplitude as the fundamental frequency increases. I would fully recommend purchasing one of these. With the exception of this very minor issue, it represents incredible value for money.

 

[ClayR]

If you need to make use of the I2S output, then you will no longer be able to use the leftmost potentiometer. Both are configurable functions of the same pin.

Do the switches short the pin to ground? If so, I should be able to connect a twisted pair in cat5 cable to the pin and ground at that point to give the best noise rejection posible. Nothing is labeled as using pins 1 or 6 on my unit. One of those is surely for the bussy led. But I would asume the power led is just connected to one of the regulators. 4,5,10, and 11 are reserved for clocks. So doesn't that leave one pin unused?

 

[ClayR]

The 4KHz spur decreases in amplitude as the fundamental frequency increases. I would fully recommend purchasing one of these. With the exception of this very minor issue, it represents incredible value for mon

OK, now I see what the graphs are. I;m kind of new to thais subject. You have to be eally listening close to hear that note 40db down from the 40hz note. If that noise is comming from a dac, might it be mostly eliminatd if you were using t as a 2.1 output instead of for headphones? Since the 40hz note and the 10k note would be on dacs, or at least different seakers.

 

[ClayR]

The 4KHz spur decreases in amplitude as the fundamental frequency increases. I would fully recommend purchasing one of these. With the exception of this very minor issue, it represents incredible value for money.

Obviously an awesome value at 15 dollars. If I hadn't already purchased the Wondom one I would have used it for the active subwoofer crossover I hope to eventually get around to installing. Of course had the amp sitting around for years now and have bought another vehicle and am yet to get it done.

 

[ClayR]

Performance is about what one might expect. Schmoo is just under 80dB, with the usual spurious garbage.

View attachment 340352


Things become quite interesting at 40Hz, with a relatively loud 4KHz spur appearing. This can actually be heard accompanying loud kick drums if listeni

Performance is about what one might expect. Schmoo is just under 80dB, with the usual spurious garbage.

View attachment 340352


Things become quite interesting at 40Hz, with a relatively loud 4KHz spur appearing. This can actually be heard accompanying loud kick drums if listening very carefully.

View attachment 340353

When I use a distortion calculator I come u with 3% at 40hz. (if I'm doing it correctly)If you can hear that you have exceptional hearing. Distortion at that low a frequency is especially hard to hear. There is a post on ASR that references a study that says bellow 120 hz most people can't detect anything bellow something like 10%.