Announcement

Collapse
No announcement yet.

xmax of mid woofer

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

  • #16
    Originally posted by wavefunction View Post
    Cheap amps are known for not meeting their specs . And power specs are often given at 1 kHz. Below 40 Hz, your amp could easily be 2x25W or less. And that is without any high pass filtering. Just skimping on power supply capacitance.
    +1. Look at this graph of a 2.0 BT amp. The BT receiver chip output is flat down to 10 Hz (Red). But the input to the op-amp stage inherently implements a 1st order HP filter due to inter-stage capacitor isolation and resistance. The FR response of that stage suffers from that filter (blue). The op-stage output sees 1st order HP due to capacitor isolation and the inherent input resistance of the TDA3116 class D chip. That results in a 2nd order HP from BT signal to voltage out (green).

    Click image for larger version

Name:	2 x 120 W BT AMP FR Graph.png
Views:	91
Size:	33.6 KB
ID:	1454157

    Comment


    • #17
      Originally posted by Millstonemike View Post

      +1. Look at this graph of a 2.0 BT amp. The BT receiver chip output is flat down to 10 Hz (Red). But the input to the op-amp stage inherently implements a 1st order HP filter due to inter-stage capacitor isolation and resistance. The FR response of that stage suffers from that filter (blue). The op-stage output sees 1st order HP due to capacitor isolation and the inherent input resistance of the TDA3116 class D chip. That results in a 2nd order HP from BT signal to voltage out (green).

      Click image for larger version

Name:	2 x 120 W BT AMP FR Graph.png
Views:	91
Size:	33.6 KB
ID:	1454157
      With the OP's original request of something that would roll off to avoid X-Max wouldn't this scenario be almost what he is looking for?

      Comment


      • #18
        Thank you for your replies,

        I used sine wave for frequency tests, only the first test is the actual song, other than that its individual frequencies. I will write things in numbers to make it more clear.

        1) i decided to build another identical speaker but this time change 1 port to 2 ports, see how it affects things.
        2) I have some spare amps laying around and will test them, see if anything changes. I am getting about 50-55W per channel power draw. Some goes to heat but not much, amp comes without heatsink and barely heats which amazes me. Its a low quality amp but quality is ok for me. I would like some more dB though.
        3) What is even stranger is that bump at 30Hz, it makes no sense ... could it be that the speaker enclosure is making the noise, vibrating, instead of the actual bass ?

        I am including pic and link of the tiny amplifier. Like i said, till now it has been perfect for me. No idle noise, no noise in silent parts of songs, better bass then other amps i tried, button, etc.

        https://www.aliexpress.com/item/4000...476d3c51kNlTrI



        Attached Files

        Comment


        • #19
          Originally posted by SentinelAeon View Post
          Thank you for your replies,

          I used sine wave for frequency tests, only the first test is the actual song, other than that its individual frequencies. I will write things in numbers to make it more clear.

          1) i decided to build another identical speaker but this time change 1 port to 2 ports, see how it affects things.
          2) I have some spare amps laying around and will test them, see if anything changes. I am getting about 50-55W per channel power draw. Some goes to heat but not much, amp comes without heatsink and barely heats which amazes me. Its a low quality amp but quality is ok for me. I would like some more dB though.
          3) What is even stranger is that bump at 30Hz, it makes no sense ... could it be that the speaker enclosure is making the noise, vibrating, instead of the actual bass ?

          I am including pic and link of the tiny amplifier. Like i said, till now it has been perfect for me. No idle noise, no noise in silent parts of songs, better bass then other amps i tried, button, etc.

          https://www.aliexpress.com/item/4000...476d3c51kNlTrI


          What power supply are you using (Volts, Amp rating)? 50 W per channel is pretty good for for this class of amps. The chip is likely designed to use the ground plane as a heatsink.

          Run the frequency test with a sine wave and measure the volts out of the amp (into the speaker) with a DVM. Say from 10 Hz to 100+ Hz in 10 Hz increments. Set the volume/signal so that your getting somewhere in the neighbor of 10+ V AC rms at 100 Hz (that will be the graph's baseline). If you post those values I can create a DB chart. I suspect you'll find a roll off at the lower end. It's an artifact of the specific implementation of these Class D amps the 1st order HP filters created by isolation capacitance in conjunction with input resistance (impedance) through the signal path and power stage of the amp.

          Lastly, does you amp actual look exactly like the picture in Ali Express (the amp chip itself)? What you see is what you get isn't guaranteed on Ali.

          Comment


          • #20
            Yes, the amp looks like this. Usualy i am running the amp on 25.2V 6S 18650 batteries (the one in question actualy is 6S2P for extra juice since its normal laptop batteries, the future speaker will use 6S 10A batteries). When i test it with DC power, i use 24V 5A external power suply.

            With this type of amp, the most i got out of a single channel was actualy about 65W when using a 3 ohm speaker.

            I will try to do what you said. Though i only have the normal cheap multimeter, as far as i know they are 60Hz ones so values above 60Hz, well, i will just do a longer test and try to detect the min/max values.

            Comment


            • #21
              Don't really know much about your mic, your room, or your method.
              Even "half-way-decent" SPL meters (like an old Radio Shack) don't measure well (flat) below the 50-100Hz range. (Then there's "weighting".)
              Are you measuring a single speaker, or a (stereo) pair? Measuring more than 1 (at a time) is problematic.
              Are you measuring "near field" (like about 0.5cm away from the voice coil), or farther away (like 1m)?
              If you're not taking "gated measurements" you're likely susceptible to all kinds of reflections.
              There's also "room gain", which is the source of MOST problems.

              WinISD can generate test tones (sc). My vers. (0.50a7) has a "Signal Gen." tab.

              Comment


              • #22
                Originally posted by SentinelAeon View Post
                Yes, the amp looks like this. Usualy i am running the amp on 25.2V 6S 18650 batteries (the one in question actualy is 6S2P for extra juice since its normal laptop batteries, the future speaker will use 6S 10A batteries). When i test it with DC power, i use 24V 5A external power suply.

                With this type of amp, the most i got out of a single channel was actualy about 65W when using a 3 ohm speaker.

                I will try to do what you said. Though i only have the normal cheap multimeter, as far as i know they are 60Hz ones so values above 60Hz, well, i will just do a longer test and try to detect the min/max values.
                A modern, cheap multi meter ... most are based on a chip that does several functions. I have found their AC measurements to be quite reliable below 10K Hz (verified against a scope). Even the Harbor Freight $5 DVM measured accurately down low. The AC measurements will tell you the frequency response of the amp's output.

                And do one additional measurement with the volume at full circa 100 Hz, That will tell you the nominal power out given a 4 ohm driver. So far, everything lines up with your PS and power readings.

                As Chris says, dB measurements down low are problematic.

                Comment


                • #23
                  I will do the measurments with multimeter asap.

                  I have a question about ports. I only have small 3cm port right now. So i was thinking of using 2 ports instead of 1. So that is 2x 3cm port. Space is important to me since i am making small speakers. So my question is:

                  1) Using 1 bigger port vs 2 smaller ones, if i tune both to same frequency, they will both same roughly the same amount of space/volume ?

                  To clarify, i made a little example in winISD. As you can see, using 2 ports takes a tiny bit more volume, but not much. Is this result and my calculations in the attachment ok ? So if i take something from this:

                  1) If possible, use 1 port of larger diameter for very small reduction in volume taken by port
                  2) If you dont have 1 large port available, u can use 2 smaller ones and will only lose tiny amount of volume this way
                  Attached Files

                  Comment


                  • #24
                    Originally posted by Millstonemike View Post

                    A modern, cheap multi meter ... most are based on a chip that does several functions. I have found their AC measurements to be quite reliable below 10K Hz (verified against a scope). Even the Harbor Freight $5 DVM measured accurately down low. The AC measurements will tell you the frequency response of the amp's output.

                    And do one additional measurement with the volume at full circa 100 Hz, That will tell you the nominal power out given a 4 ohm driver. So far, everything lines up with your PS and power readings.

                    As Chris says, dB measurements down low are problematic.
                    I did the test and am posting my findings. I used an android app called frequency generator to produce frequencies, i was testing 1 channel, i connected my TCP115 4 ohm to 1 of the channels and i measured the voltage at the speaker side. You told me to get a 10V baseline at 100Hz ... sadly i wasn't able to do that, as with that frequency generator the highest value i was able to record at 100Hz was 7.1V. I will write some points of what i gathered and in the attachment u can see voltage at various frequencies:

                    1) There was no difference between L and R channel, each measurment was about 15-20 seconds long, values kept jumping so i noted the highest value i saw
                    2) The highest voltage i managed to get during frequency measurment was at 40Hz and that was 7.6V
                    3) When testing a certain prodigy song, i managed to get 10V with it
                    4) there is some extremely bass boosted song and with it, i managed to see 13.6V
                    5) what was really wierd was ... this amp is supposed to deliver 2x55W at 24V and only 2x20W at 12V. Well, at 100Hz 12V i actualy managed to get 7.6V, same frequency at 16V also 7.6V, while at 21V i managed to get only 7.3V and at 24V 7.1V. This makes no sense

                    I will be glad if you help me make some sense of this

                    Attached Files

                    Comment


                    • #25
                      Originally posted by SentinelAeon View Post

                      I did the test and am posting my findings. I used an android app called frequency generator to produce frequencies, i was testing 1 channel, i connected my TCP115 4 ohm to 1 of the channels and i measured the voltage at the speaker side. You told me to get a 10V baseline at 100Hz ... sadly i wasn't able to do that, as with that frequency generator the highest value i was able to record at 100Hz was 7.1V. I will write some points of what i gathered and in the attachment u can see voltage at various frequencies:

                      1) There was no difference between L and R channel, each measurment was about 15-20 seconds long, values kept jumping so i noted the highest value i saw
                      2) The highest voltage i managed to get during frequency measurment was at 40Hz and that was 7.6V
                      3) When testing a certain prodigy song, i managed to get 10V with it
                      4) there is some extremely bass boosted song and with it, i managed to see 13.6V
                      5) what was really wierd was ... this amp is supposed to deliver 2x55W at 24V and only 2x20W at 12V. Well, at 100Hz 12V i actualy managed to get 7.6V, same frequency at 16V also 7.6V, while at 21V i managed to get only 7.3V and at 24V 7.1V. This makes no sense

                      I will be glad if you help me make some sense of this
                      Not sure exactly what you measured. So my comments ...

                      If your frequency app is generating a sine wave (e.g., pure tone), your measurements should not jump around. You should have been measuring with the DVM set to AC volts (I assumed you knew that). You should measure with the 24 V PS. If you can't get a steady AC reading at 100 Hz in the 16 V range (AC), your app may not have enough signal amplitude to drive the amp to full volume with a 24 V PS. Any measurements above 16 V AC with the sinewave generator and the amp is clipping. Is your 24 V PS rated at 3+ amps?

                      For this test, forget music, you want a sinewave generator. If you not sure what the app is generating, use the SineGen app on a PC/laptop and drive the amp with the PC's headphone/line output or it's BT signal.

                      It should be easy (and quick): set the generator to 100 hz, adjust the volume to ~ 10 V, measure the AC output and record the value. Leave the volume setting fixed for the remainder of the measurements. Step down to 90 Hz, measure the output and record. Repeat down to 10 Hz. Here's an example of an amp I measured: You should get a list of values as in the grey boxes.

                      Click image for larger version  Name:	Amp Measurment.png Views:	0 Size:	256.5 KB ID:	1454528

                      Comment


                      • #26
                        My power suply is rated 5+ amps. I did measure AC voltage, you can see the multimeter in the picture, set to ACV. Sadly my amp has only bluetooth input so i cant use my computer. The program i am using is using sinewave. As for the program not being able to deliver enough signal, i dont know what to do about it. The only other idea i have is to generate individual sine wave signals and save them as .mp3 and individualy play them from my phone. I will try that and see if it makes a difference.

                        Comment


                        • #27
                          Ok, it was the programs fault. What i did was make individual sine waves with audacity, at 0.8 amplitude, frequencies from 100 to 10, saved as individual files and played on my BT speaker, power suply was set to 16V, volume was set so i got about 10V at 100Hz. Again, the values are jumping about 1V up and down. And at lower frequencies they tend to go higher ... up to 10.5V. And here is the values i read (again, setup is same as in the upper picture, multimeter set to ACV). A while ago i read somewhere that this amplifier is probably so cheap because it doesn't have any filters.





                          Attached Files

                          Comment


                          • #28
                            Originally posted by SentinelAeon View Post
                            Ok, it was the programs fault. What i did was make individual sine waves with audacity, at 0.8 amplitude, frequencies from 100 to 10, saved as individual files and played on my BT speaker, power suply was set to 16V, volume was set so i got about 10V at 100Hz. Again, the values are jumping about 1V up and down. And at lower frequencies they tend to go higher ... up to 10.5V. And here is the values i read (again, setup is same as in the upper picture, multimeter set to ACV). A while ago i read somewhere that this amplifier is probably so cheap because it doesn't have any filters.




                            Ok, the test parameters (PS @ 16 V, Volume set to get 10 V) and data looks good. Hopefully it's accurate. Your amp does not lose much of anything down low. I've seen that on similar cheap amps, the decoupling caps on the amp's input creating 1st order HP at 3 Hz. I can't explain the jumping signal. You may want to test higher frequencies just to make sure you're not losing anything there. But the HF meter will not measure correctly into the higher K Hz frequencies.

                            Back to the original issue: It works, works well and the woofer doesn't bottom out down low under heavy use. The models are not perfect as per my original response.

                            Your not engineering a critical function for a moon landing. Don't over think it

                            Comment


                            • #29
                              Ok, thank you for your reply.

                              I just rechecked my winisd graph and things are more OK than i thought. As u can see on a picture, most of the graph is between 99dB and 104dB, only dropping at 70Hz under 100dB. Well, the prodigy song did produce 102dB which all in all is perfectly fine.

                              The reason it bothered me was the sine wave app i was using, like i said, it gave only 7 volts out at full volume .. but when i used sine wave made in audacity i got up to 16 or 17V. I decided i will retest my speaker with the correct sine wave recordings and dB meter. But all in all, it appears ok, graph shows me drop of 8dB between 100Hz and 60Hz, while dB meter reading showed 11dB which is still ok. What is more ... i am amazed at that performance under 60Hz. dB meter shows same 70dB at both 50Hz and 15Hz. While the graph shows me a difference of a whooping 37dB between 50Hz and 15Hz. So i am complaining that speaker will perform better than the graph is showing and i would love to know why.
                              Attached Files

                              Comment

                              Working...
                              X