Announcement

Collapse
No announcement yet.

Need advice: chip amps + DSP

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Need advice: chip amps + DSP

    I'm looking to use Sure chip amps and the new Sure ADAU1701 Audio Digital Signal Processor in a soundbar design. Each "speaker" will be a WFW composed of two Dayton RS100-4 woofers in series, and the 46mm BMR fullrange. IIRC the woofers would use at most 30 watts, but I want that to be clean. I'll be using analog RCA inputs.

    From the Sure DSP link, can anyone tell me what I all need to order? Also if anyone can provide a walkthrough on how to program it or is willing to help that would be great. I've never dealt with any of this before so it's a bit of a learning curve for me.

    Which Sure amps should I get? The speaker is sealed and so I don't think the ones with fans will do much good. There are several with the same power rating and similar price, but using different chips, so I'm pretty lost.

    Thanks!

    <div id="HtmlAreaEditor" style="display: none; color: white; font-size: 1px;">[zipedit]</div> <p style="margin: 0px 25px; padding: 15px 0px; border-width: 0px 0px 1px; border-bottom: 1px solid #464646; font-weight: normal; font-size: 28px; font-family: C
    ~Brandon

    Soma Sonus
    DriverVault

  • #2
    Should probably add that the internal height limit is 4.5", but that should fit the amps I'm looking at since I designed with idea that active would be a possibility.
    ~Brandon

    Soma Sonus
    DriverVault

    Comment


    • #3
      I'm a fan of the tripath based amps. There not perfect, but for the $, there not too shabby.

      NeilDavis has the information you need to make the adau work.

      Comment


      • #4
        Given the drivers you mention, I might use two of these. At 24V they'll deliver 36 W rms into 8 ohms. Since they're two independent units (i.e., chips not sync'd) I'd place them at opposite ends of the sound bar.

        Amo.jpg

        Comment


        • #5
          Following... This should be interesting and the sigma dsp should be a more cost effective solution

          Comment


          • #6
            Forgot about Neil's articles. Yep those look like they will be a big help!

            So by Tripath I assume something like this: https://www.parts-express.com/sure-e...3116--320-3342

            TDA7492 is another suggestion. Which is better, or more likely which implementation is better?
            ~Brandon

            Soma Sonus
            DriverVault

            Comment


            • #7
              Another that caught my eye: https://www.parts-express.com/sure-e...ogy)--320-3340
              ~Brandon

              Soma Sonus
              DriverVault

              Comment


              • #8
                Originally posted by augerpro View Post

                The 3342 is a good option as well. The 3340 won't buy you anything power wise, unless you go to a 30V PS where PS choices will be limited. But it's gain is higher at 30 dB (32x) compared to the 3342's gain of 26 dB (20x). That may make a difference depending on the amplitude of the DSP module feeding the amps.

                Neil's may be able to comment on that. EDIT: The ADAU1701 chip's output is spec'd at 0.9 V rms. That should drive any of these amps to full power.

                At 24V all of these amps will deliver the same nominal 36 W rms per channel into an 8 ohm speaker given a sufficient input signal. I don't think you would need to enable the fan on the 3340 at this power level.
                Last edited by Millstonemike; 03-10-2017, 09:15 AM.

                Comment


                • #9
                  On the ab32971, is it supposed to use the 'tk2050' (tc2001 and st508)chipset? I'm using the 4x100 watt version of the tk2050 to drive the large Dayton rs build. They're not bad sounding, good base, don't expect anything near 100 watts from them though, maybe 40 or so real watts.

                  Comment


                  • #10
                    I'm going to be adding more articles that will walk you through the gory details of programming the ADAU1701. This will be part of an extended project for the PE SBDT effort. There is also an Active Speaker Test Bed kit that will be available as a learning aid. I've got those boards made and tested and I'll be selling them at cost -- probably $30. I've already given two of them away.

                    The SBDT "contract" required a speaker using a PE cabinet, then one with your own cabinet, then an unrestricted design. I finished the first two, and I'm now working on the unrestricted designs, which will use the ADAU1701 and the 320-307 6-channel amp. The next design will focus on bass enhancement algorithms, and then after that there will be an active 2.1 system that will eventually get WiFi.

                    ​The Sure DSP board is a great price but it has two limitations. First, it doesn't have the output reconstruction filters, but that's probably not a big problem if you are using a class D amplifier. The second is that their I/O board with the RCA connectors uses only 3 outputs. So if you are using the board as a stereo active crossover, you need to pick off the signals from the main board, as there aren't enough outputs on that I/O board. But there are other alternatives to the Sure DSP board...

                    I can provide you more details of the design I'm working on next week. Our grandkids are coming today for the weekend, so I'll be busy doing other things for a while.
                    Free Passive Speaker Designer Lite (PSD-Lite) -- http://www.audiodevelopers.com/Softw...Lite/setup.exe

                    Comment


                    • #11
                      Originally posted by neildavis View Post
                      ​The Sure DSP board is a great price but it has two limitations. First, it doesn't have the output reconstruction filters, but that's probably not a big problem if you are using a class D amplifier.
                      Have you seen this board in person? I would assume they are using passive reconstruction filters.

                      Originally posted by neildavis View Post
                      ​The second is that their I/O board with the RCA connectors uses only 3 outputs.
                      I noticed this too Neil. I also wish they provided more information for their programmer options. One is listed as an "open source programmer for Sigma Studio" and one is listed as "In-Circuit Programmer for ADAU1701", but no further detail is provided. I have an Analog Devices USBi programmer, but I ordered one of the Kernel boards and one of the "open source programmers" to check them out and compare to some of the AD devices I've used before.

                      Comment


                      • #12
                        Mike> What caught my eye with the 3340 was the distortion. A lot of these supposed 50 watt amps have 10% distortion at that power! And 1% still at only 10 watts!

                        Neil> I'll probably stay with the commercially available Sure DSP for the moment. Now if you do a 3 channel in/ 6 out DSP I would be interested! Seems impossible to find 3/5 channel DSP, amps, or BT for this stuff. About the 3 outputs on the Sure board, this is really too bad. How does a stereo input only end up as 3 channels out? Where would you "pick" the signal off the board for a 4th output?

                        Could someone take a look at the Sure DSP I linked to and tell me what addons I would need to order?
                        ~Brandon

                        Soma Sonus
                        DriverVault

                        Comment


                        • #13
                          Originally posted by 1100xxben View Post
                          Have you seen this board in person? I would assume they are using passive reconstruction filters
                          Nope--no filters at all on the main board. I've got one and traced the circuit. They might have filters on the I/O board, but I don't think so. It turns out that most class D amps have anti-aliasing filters on their input--at least, that is what several of the block diagrams show. So it's probably not a problem if you are using a class D amp. However, there are two others problems, as identified in this thread: http://www.diyaudio.com/forums/digit...-module-4.html. First, it uses a strange oscillator speed: 12MHz instead of the "standard" 12.288Mhz​. So instead of sampling at 48KHz, it samples at 46.875KHz. Second, the board is set up for 2V input, which results in 6dB insertion loss. Strange decisions on Sure's part, but maybe the newer boards are different.

                          I noticed this too Neil. I also wish they provided more information for their programmer options. One is listed as an "open source programmer for Sigma Studio" and one is listed as "In-Circuit Programmer for ADAU1701", but no further detail is provided. I have an Analog Devices USBi programmer, but I ordered one of the Kernel boards and one of the "open source programmers" to check them out and compare to some of the AD devices I've used before.
                          ​People are using SigmaStudio as a user interface for the DSP, but the program wasn't intended to be used that way. ADI recommends you use SigmaStudio to define the DSP architecture, and that you develop your own user interface. The approach I'm documenting only uses SigmaStudio to design the DSP flow. My user interface processes the SigmaStudio output files, and uses a microprocessor to load and control the ADAU1701.

                          You don't really need that Sure programmer if you control the ADAU1701 with your own CPU. Since I'm using an Arduino board to load the DSP and update the Parameter RAM, my approach doesn't need that programmer. It's a much more flexible approach. My approach also doesn't require a PC for local control. You can select different crossover slopes and frequencies from an LCD display mounted on the speaker cabinet.

                          The Arduino code is downloadable, so you don't need to understand the programming details unless you want to change the DSP on your own. Those articles and upcoming tutorials are intended to de-mystify that code and provide all of the tools needed to add new DSP algorithms. It's still going to require some computer programming skills, but the really difficult stuff such as loading DSP registers and understanding details of the hardware or microprocessor will all be hidden.
                          Free Passive Speaker Designer Lite (PSD-Lite) -- http://www.audiodevelopers.com/Softw...Lite/setup.exe

                          Comment


                          • #14
                            Originally posted by augerpro View Post
                            Neil> I'll probably stay with the commercially available Sure DSP for the moment. Now if you do a 3 channel in/ 6 out DSP I would be interested!
                            ​What I'm using now is a stereo in/6 channel out. So feed it left and right and the output will be tweeter, woofer and sub for the left channel and tweeter, woofer and sub for the right channel. The volume, crossovers, EQ, delay and bass enhancement are all adjustable using that Nextion touch-screen display, which only costs $16. Since the display only requires 4 wires and the data rate is low, it can be mounted on the back of the cabinet.

                            The picture shows the ADAU1701 in the center, a teensy Arduino board toward the left, and the DAC for the subs at the upper right of the board. No working code yet for the display, but I'm getting there...

                            Click image for larger version

Name:	teensy_ADAU1701.jpg
Views:	1
Size:	127.8 KB
ID:	1323057
                            Free Passive Speaker Designer Lite (PSD-Lite) -- http://www.audiodevelopers.com/Softw...Lite/setup.exe

                            Comment


                            • #15
                              Originally posted by augerpro View Post
                              Mike> What caught my eye with the 3340 was the distortion. A lot of these supposed 50 watt amps have 10% distortion at that power! And 1% still at only 10 watts!


                              FYI: the 2x50 W limitation is a thermal power limit only.

                              I would ignore the 10% distortion figure on the 3340. That figure is with a 30 V PS into 4 ohms. The TPA3116 chip itself can deliver 70W rms per channel into 4 ohms with a 24V PS at less than 1% distortion. That's just before clipping sets in. But the thermal limit would be exceeded. With your 8 ohm configuration, you'll get 30 W rms per channel at 0.4% distortion with no issues.

                              I've worked with the TPA3116 chip amps and like them. Some say the TDA7492 is warmer in the mid-range but doesn't produce the low end as well as the TPA3116. But with your sealed RS-100s, you won't be getting that low for it to matter. Hence the TDA7492 recommendation.

                              Comment

                              Working...
                              X