3e Audio SY-DAPx002 TPA3250/TPA3255 DSP Amplifier

[manokaiser]

I don't know which, if any, of the inputs are available or exposed. You probably need to look closely at the PCB and the datasheet to see. If none are usable it should be possible to add a small mcu to do the job via the programming interface - there are some HifiBerry docs that have some coverage of doing this on the DAC+DSP and BeoCreate covering this with the expectation that you'll be using a Pi to make the changes, but it should work similarly from other things with an I2C interface. I haven't actually tried it though...

Edit: https://www.hifiberry.com/modify-dsp-registers-on-the-fly/ - covers determining the register used by a control, but uses a tool on the Pi to do the register reads and writes rather than doing it directly. More digging needed to work out how to do it with an arduino or similar.

Hey Somebodyelse, thanks for input.
I'm not at all familiar with PI or hifiberry so I should skip that.
In the attached file there's an Application Note from Analogue Devices that describes how it's done. Seems very easy and straightforward to attach a switch and a resistor to one of the input pins and do the trick.

Where I am lost is at the settings that need to be arranged in the DSP program for the pin to function accordingly.
I have no idea what it takes and I wouldn't want to mess with it. I don't even know if this kind of functionality is available or is it locked in the inaccessible sections of the DSP programming.

I believe 3eAudio could give us this information.

 

[3eaudio]

View attachment 126472

Well I've made a simplified version of it, with a two way switch (yellow highlight), which should be fairly easy to implement with an "Input GPIO Debounce" kind of thing, as I read on the web. Fact is I do not know if it's possible with this amplifier or how to do it if it is possible.

Can anyone help with this please? It's a simple outboard hardware switch to bypass the subwoofer filters and have the amp in full frequency operation. Like a "direct mode".



Also, I have another question, regarding the internal sampling rate frequency (highlighted on the top).
Default is 48KHz, is it OK to use 96KHz sampling for these filters?

Anyone to the resque? 3eAudio, please?

your idea is ok and please see there is 3 PIN you can use,but need some skill on soldering.
in sigma studio you can add a counter to IO and then switch by 2 or more.

 

[Yiba]

In addition to what 3e Audio stated above, you'll need:
1. ADAU1701/IO/GPIO/Input/General Purpose Input for whichever pin your pullup/down switch is connected, GPIO_1(MP1), GPIO_2(MP2) or GPIO_7(MP7). Add a 0.1uF cap to the switch as the hardware debouncer.
2. /Basic DSP/Logic/Invert/Buffer Gate (this does not invert the signal even though it is classified under Invert) to translate the above on/off analog signal to integers 0 and 1
3. /Muxes Demuxes/Multiplexers/Index Selectable/No Slew/Stereo/Index Selectable to accept the integer as the index (0 or 1) to select the signal path.

The lower right hardware resource section of the main window does not work (yet). Look for compiler_output.txt file in a sub-directory (forgot the name) under Sigma Studio/yourprojectname/. Opening the .txt file shows you the resource usages.

 

[manokaiser]

Dear 3eAudio and Yiba,

Thank you for the responces.
So if I get it right this should do the trick, along with the appropriate hardware switch and RC network,
is that correct?

 

[Yiba]

Yes, the orange circled area looks good to me. This is how external control of signal paths is done. Try it with the hardware connected.

I don't understand the use of bass boost though. What are you trying to achieve there? If it is for some fancy volume control of sub-woofers, 3e Audio could give you a prototype .dspproj file for logarithmic volume control, which gives you access to DAP2002 primary volume control connected to AUX_ADC_2.

 

[manokaiser]

Yes, the orange circled area looks good to me. This is how external control of signal paths is done. Try it with the hardware connected.

I don't understand the use of bass boost though. What are you trying to achieve there? If it is for some fancy volume control of sub-woofers, 3e Audio could give you a prototype .dspproj file for logarithmic volume control, which gives you access to DAP2002 primary volume control connected to AUX_ADC_2.

Thanks for the confirmation.

The bass boost is there to correct the natural response of the speakers I have designed. I use a couple of H Frame, open baffle bass drivers, which have a mild roll off starting around 50 Hz and lower.
So depending on their final responce in the room a slight bass boost can be used at the lowest frequencies ~ 30-40 Hz to level up the frequency responce.

After that, there's the high pass rumble filter at ~ 25- 30 Hz to protect the driver from overexcursion.
And finally the low pass filter at ~ 100-125 Hz to integrate them with the full range speakers.

The 3eaudio amplifier input would come from the mains amplifier's sub output, so it's volume control will be used only to balance the two amplifiers together.
After that, system volume will be controlled by the mains Amp.

The logarithmic volume control will probably be useful in any case, so I will request it.
Thank you so much!

 

[Yiba]

As it is almost impossible without measuring in an anechoic chamber to separate "natural response" from room acoustics, what you are trying to do is better incorporated with REW / AutoEQ tuning for room acoustics in the 20 - 125Hz range. I have discussed the details in https://www.diyaudio.com/forums/class-d/369917-3e-audio-tpa3250d2-sy-dap1002-mods-dsp-tuning.html
Do the analysis with AutoEQ replacing the Bass Boost for 20 - 125Hz and when you are satisfied with the filters, replace 2 AutoEQ blocks (and why not the HP and LP filters as well) with a single general stereo filter block
/Filters/Second Order/Single Precision/2Ch/General(2nd Order) --could be double precision or Nth order if you want--
by manually transferring the filter specs in AutoEQ to it, with 4 - 6 filters as you require, to save resources so that you could use 96 or 192kHz system sampling rate without hitting the limit, or overloading ADAU1701.

 

[manokaiser]

Dear Yiba,

I've read your post in DIY Audio. Very interesting indeed.
Regarding the auto EQ filters, well... there's a long debate on how much and at what frequencies this can be effective, in doing more good than harm.
My conclusions, after lot's of reading and research is that in order to apply any filters to correct room interactions, first and most important is to know what the speakers are doing naturally, when the room is out of the equation. Then, any further room measurements can be interpreted correctly and filters can be manipulated accordingly.

The way to do that cannot blindly be left to any autoEQ algorithms for in room corrections, as there's great variability in room responses on different positions.
My priority here is to measure the Subs outside, flatten their response from 35-125 Hz, then set them in the room with that initially ideal curve to check what the room is doing to their near perfect response.
Then I can have a grater overview of the actual room contributions and adjust accordingly.

By the way, I am using the same tools as you do in your post, iMM6 mic, REW, along with EqualizerAPO and Voicemeeter Banana.

The difference in the approach is that any nulls created by the natural speaker response CAN be corrected, while those created by the room CANNOT and should not be corrected.
So, unless you separate the two you do not know what can be adjusted and what is better left alone. With autoEQ alone this can't be the case.
For that reason I have set the system as such, in order to accomplish the initial ideal FR for the speakers, out of the room. According to my speaker simulations I need the bass boost and the Hi & Low pass filters to accomplish that.
Afterwards, i will certainly rearrange the necessary filters and DSP programming to readjust for the actual in room response.
Thank you so much for your ideas and specific suggestions, i will definitely give them a go.

Well, so far, this is only a plan, as I am still waiting for the amp to be delivered , it takes so long...

 

[dtn]

аutoeq

 

[Weeb Labs]

This unit is simultaneously fun and disappointing. The ADAU1701 provides four I2S input channels (bypassing ADC) and four DAC channels, yet this unit exposes only the two ADC inputs and nothing else. Just one DAC channel in the form of an extra RCA output on the rear would have rendered this one of the most flexible pre-built amplifiers on the market, as it pertains to subwoofer integration.

Bodging one yourself is fairly straightforward but this unit could have been the recommendation for those with relatively sensitive speakers who either intend to purchase a subwoofer or make use of DRC. Had the manufacturer gone even further and exposed three DAC channels (at minimal additional cost) said flexibility would have extended to both active and passive speakers. This could quite easily have been the new "swiss army knife of audio devices".

 

[creed]

I ordered an unit of SY-DAP2002, however didn't realize whether the unit I ordered was 4ohm/8ohm version as i only came across here post the ordered, how do i check which version is that?

Quick question, i do have a spare pair of NE5532P (not the NE5532AP) and a spare pair of AD797ANZ opamps, wondering if these opamp will be able to fit into the SY-DAP2002?

 

[molexey]

This unit is simultaneously fun and disappointing. The ADAU1701 provides four I2S input channels (bypassing ADC) and four DAC channels, yet this unit exposes only the two ADC inputs and nothing else.

@Otaku+, you are misinformed. All of the four DAC channels are used for TPA325x balanced inputs (pretty clever solution) and two of I2S inputs are used for the Bluetooth module.

 

[Weeb Labs]

@Otaku+, you are misinformed. All of the four DAC channels are used for TPA325x balanced inputs (pretty clever solution) and two of I2S inputs are used for the Bluetooth module.

You are correct. I was unaware that the four DAC channels were being used to feed the 3255's balanced inputs but this still strikes me as unnecessary. The 3255 exhibits an unbalanced input sensitivity of about 2.8Vrms, so it would have been just as effective to buffer the ADAU's 0.9Vrms output as required.

I'm aware that two of the I2S channels are occupied by the Bluetooth radio but that still leaves two free.

 

[AdamG]

I ordered an unit of SY-DAP2002, however didn't realize whether the unit I ordered was 4ohm/8ohm version as i only came across here post the ordered, how do i check which version is that?

Quick question, i do have a spare pair of NE5532P (not the NE5532AP) and a spare pair of AD797ANZ opamps, wondering if these opamp will be able to fit into the SY-DAP2002?

Welcome Aboard @creed.

 

[creed]

I ordered an unit of SY-DAP2002, however didn't realize whether the unit I ordered was 4ohm/8ohm version as i only came across here post the ordered, how do i check which version is that?

Quick question, i do have a spare pair of NE5532P (not the NE5532AP) and a spare pair of AD797ANZ opamps, wondering if these opamp will be able to fit into the SY-DAP2002?

anyone can share some feedback if the AD797ANZ opamp can be use? since i've a pair if that's possible i might want to do some opamp rolling and see how it behave, just don't wish to break things up

Much appreciated!

 

[bloqhed]

anyone can share some feedback if the AD797ANZ opamp can be use? since i've a pair if that's possible i might want to do some opamp rolling and see how it behave, just don't wish to break things up

Much appreciated!

The AD797ANZ is a single channel opamp whereas the SY-DAP2002 uses dual channel opamps. In order to use the AD797ANZ you would need a total of 4 chips, and 2 single to dual adapters, for example the BrownDog 021001.

 

[creed]

thanks for the response, will scrap the idea for now , initially thought it can be a drop-in for testing

 

[AfterThought]

Does anyone know what this issue could be?

 

[Toku]

Does anyone know what this issue could be?

There are cases where the NE5532 used is of poor quality. (Fake NE5532) Try replacing the OP amplifier once.

 

[AfterThought]

There are cases where the NE5532 used is of poor quality. (Fake NE5532) Try replacing the OP amplifier once.

Opamp was replaced. I tried a few different ones from Digikey... Currently the LM4562 are installed.