Announcement

Collapse
No announcement yet.

Christmas build project - Bedroom active 2.1 system

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

  • #16
    Last jack and knobs installed... in place listening test, swapping in the RSS265 sub from my car, running on only one channel of the sub amp. Sounds good! No noise issues with the gain levels required for listening off the PC output. There's some broadband noise, sounds like brown noise if you put your ear up to the speaker, but nothing audible even in a quiet room. Reserving judgment on the sub, but at under 100W the SD215A should be pretty similar to that RSS sub, less a little extension under 40Hz.

    Here's the look of the plate powered up:
    Click image for larger version

Name:	2017-03-20 13.48.27.jpg
Views:	1
Size:	468.7 KB
ID:	1324828
    Electronics engineer, woofer enthusiast, and musician.
    Wogg Music

    Comment


    • #17
      Additional musings on the TDA7492 boards...

      One goal of this build is to eliminate turn on / off thump. Right now, the bedroom speakers run through a DTA-1 amp that is ALWAYS on. In the middle of the night, the PC often wakes up for backup / updates and throws a thump through the speakers. I'm a light sleeper, so this is pretty annoying. This new design will use the USB power to kick on the amp, so it will not be powered up all the time anymore.

      The TDA7492 board has a 1 second delay turn on using the chip's standby / mute pins, with that the turn on thump is completely eliminated. The problem now becomes turn off thump. The Meanwell 24V supply hangs on for a second or two, and during discharge the chip amp's mute pin remains high, so as the power supply drops a nasty thump happens. So instead of a PC turn on thump, I'll be treated to a PC turn off thump when it goes back to sleep.

      I thought about a relay to disconnect the speakers when the power goes off, but wasn't sure if it would cut fast enough to work. I watched the PSU and a single end of the output pins with the scope last night to get the timing of the power on / off sequences and see what I could do. I also perused the data sheet looking for the thump prevention circuits, and determined I have no way to force that mute pin low and avoid this thump. This weekend I'll be testing the turn off thump to determine if the amps themselves are generating the thump, or if it's generated by the crossover boards. If it turns out to be the crossover boards, I'll add a diode charged capacitor to the crossover power supply lines, which will keep them powered up for a second or two after the chip amps go down to eliminate the thump.

      On distortion...
      Another thing I've noticed on the scope is that these boards do not deal with clipping gracefully. That jives with the comments on PE that they sound bad when pushed for power. A good amount of HF noise appears as the voltage swings toward the supply before clipping. This was observed with the amp unloaded, so the filtering may not be as effective in that case, but nevertheless that's what the scope shows. I'll run that into a speaker this weekend as well to see if that noise appears under load.
      Last edited by wogg; 03-24-2017, 01:32 PM. Reason: Adding comment about distortion
      Electronics engineer, woofer enthusiast, and musician.
      Wogg Music

      Comment


      • #18
        Originally posted by wogg View Post
        Additional musings on the TDA7492 boards...

        One goal of this build is to eliminate turn on / off thump. Right now, the bedroom speakers run through a DTA-1 amp that is ALWAYS on. In the middle of the night, the PC often wakes up for backup / updates and throws a thump through the speakers. I'm a light sleeper, so this is pretty annoying. This new design will use the USB power to kick on the amp, so it will not be powered up all the time anymore.

        The TDA7492 board has a 1 second delay turn on using the chip's standby / mute pins, with that the turn on thump is completely eliminated. The problem now becomes turn off thump. The Meanwell 24V supply hangs on for a second or two, and during discharge the chip amp's mute pin remains high, so as the power supply drops a nasty thump happens. So instead of a PC turn on thump, I'll be treated to a PC turn off thump when it goes back to sleep.

        I thought about a relay to disconnect the speakers when the power goes off, but wasn't sure if it would cut fast enough to work. I watched the PSU and a single end of the output pins with the scope last night to get the timing of the power on / off sequences and see what I could do. I also perused the data sheet looking for the thump prevention circuits, and determined I have no way to force that mute pin low and avoid this thump. This weekend I'll be testing the turn off thump to determine if the amps themselves are generating the thump, or if it's generated by the crossover boards. If it turns out to be the crossover boards, I'll add a diode charged capacitor to the crossover power supply lines, which will keep them powered up for a second or two after the chip amps go down to eliminate the thump.

        On distortion...
        Another thing I've noticed on the scope is that these boards do not deal with clipping gracefully. That jives with the comments on PE that they sound bad when pushed for power. A good amount of HF noise appears as the voltage swings toward the supply before clipping. This was observed with the amp unloaded, so the filtering may not be as effective in that case, but nevertheless that's what the scope shows. I'll run that into a speaker this weekend as well to see if that noise appears under load.
        I pretty sure it's the amps - it's been reported on DIY audio. There might be a way to schedule a daily task(s) in the PC to mute the sound card during the middle of the night and restore it in the morning. That may or may not work for a wakeup / sleep issue.

        One thing I notice by accident on another chip ... if the amp does have an on-board PS bypass cap, there's no energy to power the thump when power is cut to the amp. You might try putting the PS bypass cap outside the board and cut power to the board and the power supply this would cut power to the PS and the bypass cap energy to the board.
        Last edited by Millstonemike; 03-24-2017, 05:09 PM.

        Comment


        • #19
          Originally posted by Millstonemike View Post
          I pretty sure it's the amps - it's been reported on DIY audio. There might be a way to schedule a daily task(s) in the PC to mute the sound card during the middle of the night and restore it in the morning.
          Just tested.... ​Actually with the gain pot right before the amp all the way down (shorting the amp input) the turn off thump disappears. There's a small bit of noise, but nothing like the WATHUMP that is coming through the crossover, the woofer cone goes almost full exertion. The next trick is that I put 660 ohms across the crossover supply for the voltage splitter that created the ground reference for the op-amps. That means I need near 10,000 uF to get a charge big enough to keep them powered up for a few seconds. ...maybe a mod later.

          ​Here's some images of the clipping distortion I was talking about. At 12V RMS output it looks good
          Click image for larger version

Name:	2017-03-24 16.13.09.jpg
Views:	1
Size:	566.2 KB
ID:	1325372


          ​At 16V RMS, before clipping we get this high frequency noise at the tops of the waveform. It's not really audible through just the woofer but it's really significant distortion and will likely cause that crappy sound at power relative to other amps for transients that push the output voltage.
          Click image for larger version

Name:	2017-03-24 16.13.30.jpg
Views:	1
Size:	722.7 KB
ID:	1325374


          ​And here's what actual clipping looks like. The flat lines appear, but they're surrounded by the same high frequency noise that appears just before clipping.
          Click image for larger version

Name:	2017-03-24 16.13.39.jpg
Views:	1
Size:	761.8 KB
ID:	1325373


          ​Based on this, the board has 20W continuous RMS of real clean power into 8 ohms with a 24V supply, and a little over 30W of "dirty" peak power.
          Electronics engineer, woofer enthusiast, and musician.
          Wogg Music

          Comment


          • #20
            I was fooled by the blue PCB. It's the Sanwu TPA3116 board that has the thump (and turn on easy to correct, turn off, not so easy).

            The scope shot of the ringing at clipping; is that a 4 ohm woofer? The board's output filters are optimized for 8 ohm and I wonder what difference that might make.

            I'm currently investigating a 2.1 amp board with an on-board active pre-amp. The designer used a 5V regulator to power the single supply op-amps. I wonder if that would keep the op-amp signals intact until most of the 24 V supply has bled away reducing/eliminating thump. A PS bypass cap on the regulator would help in that regard with the op-amps' low current draw.

            Comment


            • #21
              The scope shot is on the SD215-88, 8 ohm coil and probably higher than that at 45Hz free air. 5V seems low for the op amps, they'd start clipping at 2V peak. I'm going to go with a pair of 4700uF caps when I get a chance. First... make woofer box. Sawdust commenced this morning. Sent from my SM-G920P using Tapatalk
              Electronics engineer, woofer enthusiast, and musician.
              Wogg Music

              Comment


              • #22
                Originally posted by wogg View Post
                The scope shot is on the SD215-88, 8 ohm coil and probably higher than that at 45Hz free air. 5V seems low for the op amps, they'd start clipping at 2V peak. I'm going to go with a pair of 4700uF caps when I get a chance. First... make woofer box. Sawdust commenced this morning. Sent from my SM-G920P using Tapatalk
                Yes, that's an issue with the amp I'm looking at. I think the original intent was to avoid clipping in the power stage by limiting the pre-amp signal. The 5532 op-amps have kind of a soft clipping inherent in their design.

                Comment


                • #23
                  Sawdust created... All I've got on hand to use for this is a 7 1/4" skill saw, an aluminum fence and c-clamps, a router, and a Jasper jig. I was super careful about the measurements and even bought a new 40 tooth blade, but it seems +- 1/16" is as close as I can get with accuracy on these cuts. A table saw would be way better for repeatable cuts. I see a lot of filler and sanding in my future.

                  ​This is the box on it's side. I plan on fabricating a circular grill over the woofer that snugs right in that outer recess.
                  Click image for larger version

Name:	2017-03-25 16.02.24.jpg
Views:	1
Size:	502.3 KB
ID:	1325458


                  Click image for larger version

Name:	2017-03-25 16.02.31.jpg
Views:	1
Size:	558.7 KB
ID:	1325459

                  Electronics engineer, woofer enthusiast, and musician.
                  Wogg Music

                  Comment


                  • #24
                    Now the vision is coming together. It's all glued up, the Precision port is flush mounted. I press fit and tested the tuning with a scope, which turned out to be 30Hz, just a teeny bit below the model. I'll keep it, liking the EBS alignment for the room.

                    Click image for larger version

Name:	2017-04-07 21.10.18.jpg
Views:	1
Size:	343.4 KB
ID:	1327471


                    After Bondo, sand, round-over, then sand again, ​now we have a cabinet!

                    Click image for larger version

Name:	2017-04-08 14.36.58.jpg
Views:	1
Size:	517.3 KB
ID:	1327472


                    ​I don't have the 1x2's I thought I did, so it's a trip to the store for a internal cross brace and the mounting surface for the plate amp in the back. Up next is Duratex! The Precision port will be painted like everything else with the goal of being completely seamless.
                    Electronics engineer, woofer enthusiast, and musician.
                    Wogg Music

                    Comment


                    • #25
                      Now with Duratex! I used a white foam roller typically for doors or trim this time for a less aggressive texture. I was also down to my last bit of Duratex, so the roll isn't the best. The heavy texture really tends to hide things a bit better I think. This has a few minor but odd texture issues, but otherwise looks alright.
                      Click image for larger version

Name:	2017-04-09 18.02.14.jpg
Views:	2
Size:	482.7 KB
ID:	1327573


                      Here's the backside with the rails to mount the plate amp.

                      Click image for larger version

Name:	2017-04-09 18.01.53.jpg
Views:	1
Size:	468.9 KB
ID:	1327571



                      Attached Files
                      Electronics engineer, woofer enthusiast, and musician.
                      Wogg Music

                      Comment


                      • #26
                        And we're done! I'll have to condense this into the Project gallery soon.

                        For a grille I'm feeling a little lazy about a complex circle jig creation with expanded sheet metal on the front covered by grill cloth. Instead I'm thinking about one of these. The mounting ring wouldn't be used, instead I'd apply foam weather stripping around the outside to squeeze in the 8.75" inset for the woofer. The outside of the grille is 8.375", a little smaller than the 8.5" OD of the woofer but still outside the surround. Weather stripping that is 1/4" thick should close that gap and give it some pressure to stay in place without rattle.
                        Click image for larger version

Name:	2017-04-09 22.09.09.jpg
Views:	1
Size:	418.7 KB
ID:	1327608


                        And the business end with the custom 2.1 plate amp.
                        Click image for larger version

Name:	2017-04-09 22.08.45.jpg
Views:	1
Size:	489.3 KB
ID:	1327607
                        Electronics engineer, woofer enthusiast, and musician.
                        Wogg Music

                        Comment


                        • #27
                          Did you cure the thump?

                          Comment


                          • #28
                            Originally posted by Millstonemike View Post
                            Did you cure the thump?
                            Yes sir. I added a diode charged pair of 4700uF caps to power the pre-amp boards. There's a small cone movement as the system powers down, but no audible thump.

                            When I publish this as a project, I'm going to provide a crossover board re-design purpose built for this application. I can raise the voltage divider resistors and only use one set for all op-amps on the same board and accomplish that with a whole lot less capacitance.
                            Electronics engineer, woofer enthusiast, and musician.
                            Wogg Music

                            Comment

                            Working...
                            X