1. OCAU Merchandise is available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion in this thread.
    Dismiss Notice

Subwoofer not very loud

Discussion in 'Audio Visual' started by neRok, Dec 16, 2021.

  1. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    They do, because they have AES/EBU inputs, which is 2chan digital signal (basically SPDIF) over XLR cable. So you daisy chain source to spkr_1 to spkr_2 to sub_n, and if you get subs with the same features, you can do DSP and DAC in the sub too.
     
  2. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,176
    If you're going to go into modding the analogue circuit probably best off having this moved to electronics. Save the 'whatever the fuck you're doing?' questions.
    The value of the cap is 10pf, it seems the schematic provides the value in the colour band notation (which is the printed number on some capacitors).

    Same goes for everything else: 154 is actually 150nF etc.

    The cap in the feedback of the final opamp stage is there to ensure stability of the opamp.

    ED- Just to be clear, I don't think there will be significant value to jumping into modding the existing circuit as it largely sounds as though it probably actually works, especially if you haven't done the basic checking and set up procedure that was intended for this type of equipment.
     
    Last edited: Jun 19, 2022
  3. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    Just looked into this, and you are right. I will have to review these simulations.
    Edit: Why is the one in the brown area 100P then? Shouldn't it be 101?
    Edit: Capital P, so it's peta? Regarding the light blue area, I was using 104n, but its 100000p, which is 100n. So, fluked it, its almost the same :lol:

    Was thinking of posting in electronics section too. Better do that.

    What "checking and procedures" are you referring to?
     
    Last edited: Jun 19, 2022
  4. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,176
    Checking that your source, and main amplifier is passing full range, then the subwoofer can be adequately loud by itself and is working as expected. Then performing the setup of this kind of crossover system by mains through the sub plate line out, setting approximate volumes, flipping phase and using a test tone at the crossover point to find the nullest phase for the listening position followed by finding the nullmost for the frequency setting, flipping the phase, setting levels and tweaking the frequency down or up depending on how much excess midbass you feel you're getting or not getting. The result won't be as state of the art as DSP - but it is as much enough to make justifications on what you should do next with what you already have.

    It might be the case that you have done this, it's just that with the divergence of obtaining other equipment and doing it a totally other way but then for unrelated reasons not doing that, it seems as though you've just jumped straight into diagnosing no apparent problem and planning to modify without a solid justification as to why.

    It could be a mistake but I think that's actually 100p - they might be trying to use it to decrease the gain stage's sensitivity to noise above audio frequency as well.

    It's possible the reason for how they present it is that it made it easier for unskilled assembly folk to look at the components and see that they're the same. Leaves to question how they would have done with the colour codes on the resistors but it's just a theory.
     
    Last edited: Jun 19, 2022
  5. Audionut

    Audionut Member

    Joined:
    Mar 7, 2006
    Messages:
    1,529
    Location:
    4610
    You're right, I totally missed the pass-through. It's questionable needing DSP in each and every speaker, but it's your money.

    This is a more appropriate response than, "whatever the fuck you are doing". It's not that I don't understand, or that this should be in electronics, simply that the OP is throwing spaghetti at a wall to see what sticks. All fun and games I guess, and I should check my attitude.
     
  6. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    I need to filter the low-bass from my bookshelfs with a high pass, and I only have 2 channels from DAC/preamp (left and right), so I can't DSP the signal coming out of PC, and so I'm wondering where I can do it. That's why I was looking at the "features" of this plate amp I already have, and it turns out it is high-passing the pass-through signals, so I can go DAC -> sub plate amp -> bookshelf amp, and get a "free" high pass along the way. It's also rolling off the treble though, but actually that will be a benefit for my particular bookshelf speakers.
    Edit: I also previously found an old review of my speakers from what appears to be a knowledgeable blogger. They recommended changing the internal crossover to give better results, and gave a few options. So I am contemplating making one of these crossovers, and thus the potential is there to build it with a low-frequency high pass, to remove the low bass.

    About modifying the bass boost, I've used WinISD to design a sub box with this plate amp in the past, and not knowing the true signal from this amp (slopes, boost point, etc), it made the design less than perfect (basically garbage data in, garbage design out). The actual boost level and point was proving a "problem" for that particular sub speaker too (when looking at excursion and port velocities and the like).

    Anyway, I've updated all my simulations, but I still don't know if 100P is 100p (and if so, why isn't it 101)? The pass-through signal is basically the same;
    full signal - pass through.png

    The low pass circuit doesn't put a hump in the signal;
    low pass knob circuit.png

    The volume knob now simulates as a low pass, but is flat where it matters;
    volume knob circuit.png

    And the simulated sub signal looks much the same (this is with both knobs at 50%);
    full signal - to sub - freq=050 vol=050.png

    And I've played around with different resistors and caps on the rumble filter (purple) and the variable low pass (lower value resistors gives less steep slope), and so that's all interesting.

    The Parts Express info I linked gave an option for 1dB bass boost at 36-40Hz by changing to 27k and 120k resistors in the rumble filter, and that gives this signal;
    Test Response (freq=050 vol=050 R23=27k R24=120k).png
     
    Last edited: Jun 19, 2022
  7. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    DSP in the speaker also implies digital crossover of the individual drivers. The amp in each of those JBL is 2x250W, which allows the same.

    I don't like the idea of converting analogue back to digital for DSP, like the basic miniDSP 2x4 does. Some active monitors seem to do this, because they only have analogue inputs with DSP ability, like the Genelec 8320A. And for digital crossover with analogue input, the conversion is necessary.

    I really like the idea of digital crossover, and active speakers with all the fruit in them make that rather easy. Passives speakers with a bunch of amps is an option, but the DAC/preamp options are limited. You can do PC 7.1 out like you said (for 8 channels, eg 2x3 for speakers, and 2 for subs), but then you're using unbalanced outputs with a DAC of questionable quality, and no volume knob. There's not a lot of options to buy either, Motu UltraLike-mk5 being one of the only, and it has 10 balanced outputs. miniDSP Flex and SHD only have 4 output, but there is the out of stock 4x10 and 10x10 (with "wiring" style output for the analogue balanced, not plugs), but if you are doing DSP on PC, why pay for DSP in the "DAC" if you don't need it*?!
    * The benefit of DSP on a device in the signal chain (eg the active speaker, or miniDSP) is that you change sources and not lose your DSP corrections.
     
    Last edited: Jun 19, 2022
  8. Audionut

    Audionut Member

    Joined:
    Mar 7, 2006
    Messages:
    1,529
    Location:
    4610
    Digital xovers are child's play. EAPO does that for free.

    Power distribution is not equal.

    So.........maybe don't use the cheapest version. Honestly, I'm a little flabbergasted in the context of this thread, that RRP $10,000 speakers suddenly get thrown in the mix, and concerns about digital>analog>digital conversion (valid if using low quality equipment), and then......you chose to rely on the cheapest minidsp for the narrative.:confused:

    This one's 8 in, 8 out. Heck, I threw the minidsp as a bit of an idea. You can do all sorts of things and whatever you want. This will give you a volume knob, for instance (review).

    Wrong. Any mid to high AVR will give you full digital chain, of known quality, with a volume knob. Researching particular needs will undoubtedly lead to other solutions.

    Fucks me. That's one reason I thought it very strange that $10,000 JBL's suddenly become relevant.

    IMO, a sound card would have been a better investment/solution to your current temporary?? needs, with respect to the thread OP. If the converse to that (basic) solution is excessively priced gear and a deep dive into audio theory, that's a different discussion.
     
  9. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    It's not free because you need a sufficient number of DAC and amps to do it.

    Any analogue->digital->analogue is a potential source of problems. It's a totally unnecessary step when starting with digital in the first place.

    I only mentioned it because its one of the few products in their lineup that is only analogue input.

    It doesn't have USB signal input, only 8xAES/EBU, so the problem of getting those digital signals out of the PC still exists. But yes, it would probably work, much like the 4x10 I mentioned. But it's $1370 AUD BTW. And now for my 2x bookshelfs I need 2x2ch DAC for the 4 AES/EBU signals (take for example 2x Topping E30 @ $240 might do it, so another $500), and now I also need 4x ~100W amps (something like Emotiva BasX A4 would do it for ~$1200 AUD). That's $3k before speakers, and only good for 2 way speakers.

    I'm not complaining about the price, I'm just comparing to what you can get from active monitor speakers for $6-8k.

    I came across it in my research. It's a nice bit of kit, but it's also as rare as unicorns (and not for sale new). Hopefully something like it comes along in the near future. It would also be better if it had a semi decent headphone amp within it, or maybe even a simple way to switch outputs. Still, it was ~$1500 AUD, so with amps is pushing $3k too.

    They're not good AFAIK, and they certainly aren't cheap, especially when considering the features I don't need but would be paying for (video features, built in amp, Dolby/DTS features, etc). As an example, Denon AVC-X3700H = $2700 at HN and had a lowly 98 on the SINAD score, which you will note at the time (2020) was one of the best AVR scores lol. The Okto 8 you mentioned had 118dB. Scroll down in that review, you will see Intermodulation Distortion with the notes "Good for an AVR, higher noise than desktop DACs", and you will clearly see in graph that it is worse than the 4200 Topping DX3 Pro+ I purchased for $300 the other day. Carry on to jitter = "Technically unacceptable". Also doesn't have balanced analogue outputs.

    The DX3 will give a better signal to the amps, and with the analogue "crossover" in my sub amp, the situation should be fine. But yes I could have done 3ch out of my PC with EqAPO doing the crossover, and fed the 2 amps that way, but the quality would have been worse. And yes, maybe its splitting hairs at this level, but better is better. I also don't listen to 320kb MP3, because why would I when FLAC is "free"?

    Also my UMIK-1 arrived on Friday, and it's apparent I need to do some room conditioning (have a massive node at 50Hz). So I will probably review that situation before shopping for better speakers.
     
    Last edited: Jun 19, 2022
  10. Audionut

    Audionut Member

    Joined:
    Mar 7, 2006
    Messages:
    1,529
    Location:
    4610
    Righto, meaningless numbers are your forte. Good luck with it all.
     
  11. flu!d

    flu!d Motoring and Intel forum admin

    Joined:
    Jun 27, 2001
    Messages:
    21,043
    NeRok,

    For what it's worth, I think what you're doing is very interesting and I look forward to seeing the results from someone thinking outside the box - Good job.
     
    neRok likes this.
  12. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    New Topping DAC arrived, so I'm going to start mucking around that soon. I did some measurements with my UMIK-1 the other day and noticed a massive peak at 50Hz, even with my bookshelfs high passed, ports stuffed, and sub off. I found a post online last night that said this can be because of 50Hz grid frequency. I wonder if this better DAC and Emotiva power amp will be better?! I will update about that in my REW thread once I know.

    But back onto this topic, I've had a go at replicating the plate amps EQ into WinISD. For this rumble filter (high pass + boost) I can enter a custom high pass (specify frequency and Q factor), or the better option seems to be the "Peaking 2nd order highpass" that needs boost dB and peak frequency Hz. So I've attempted to calculate the boost dB-SPL value (the frequency can be simply measured in the simulation). Using the graph below as an example;

    The green line is the rumble filter only, and you can see it flat lines at ~200Hz and ~7.66dBV gain (from 1V AC source, which is what line level is). The peak is ~31.6Hz and 11.86dBV, which makes for 4.2dBV difference. So then I found a formula to turn dBV to Volts, then another to turn Volts to Power (Watt), and another to turn Power to dB-SPL. And so if this is all correct, it turns out the answer is the same 4.2dB level difference (this is the SPL boost).
    calcs.png

    So I plugged that into WinISD, and for the variable low pass I just eye balled it, which gave this curve;
    winisd-eq.png
    And then I combined it with a 10" sub I had already configured in WinISD (not the same one as OP), with a 15L sealed box, and at 1 watt it makes a lowly 81dB lol.
    winisd-sql-1watt.png
    If I crank it to its max 125W, it makes 102dB.
    winisd-sql-125watt.png
    You might think a bigger box would help, but actually the cone excursion increases with a bigger box, and its over its max at this volume vs box size already.
    winisd-excursion-125watt.png

    Anyway, lots to muck around with. This 10" sub is evidently better suited to a ported box, and I haven't looked into the 12" from OP yet.

    AIso found an interesting nugget of info last night that I need to look into;
     
  13. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    A bit more info on this point...

    Here is a subwoofer design, and I have the same driver available (so could make the same design). That site has/had a WinISD config that I have a copy of. Below is its response curve, using custom spec EQ filters. It shows a nice flat response from 25-100Hz.
    before.jpg

    And here is the same subwoofer design, but now using the plate amp I have been talking about, with its built in filters. The result is dreadful.
    after.jpg

    The rumble filter peaks too strongly and could do with being a little lower, and the low pass doesn't really go high enough.
     
  14. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,176
    Yep that's reasonable enough - the plate amp's boost certainly doesn't really do the ported option any good here. Room gain at LF makes things worse too.

    With the sealed boxes, Linkwitz's biquad compensation method (ala. his Thor design) works quite well, but it's quite intensive on amplifier power.

    Sealed boxes definitely need cone area (and stroke). I run 6.5" quad side firers and could do with a little bit more cone.
     
    neRok likes this.
  15. OP
    OP
    neRok

    neRok Member

    Joined:
    Aug 19, 2006
    Messages:
    4,315
    Location:
    Perth NOR
    Just thought I would put an end to this thread, because I sold the amp. I spent some time in WinISD doing designs with the 10" and 12" subs I have, and the bass-boost/rumble-filter was a pain in the ass. It was possible to "fight" it with DSP, but the low pass was a different story. I double checked my circuit simulations were correct, and they seemed to be, which meant the low pass was always kicking in too low, even on max. I really wanted it to be flat until at least 100Hz, then drop, but instead as you can see in my simulations, it was dropping basically straight after the boosted area. Whether this was what it was actually doing is another question, but I decided it was too much effort to answer...

    Re sub design - putting my 12" in any sized sealed box wasn't going to be good enough. I fiddled around with a vented design and got a decent result, except the cone excursion was getting high. The problem is, from all the literature I have on this sub, none list the X-max (the working limit for cone excursion). So I was shooting blind, and coupled with "fighting" the plate amps characteristics, I decided fuck it, and put them up for sale. It was at the point of too hard, and if I really want to build a sub, I think I could get better results buying more appropriate and modern equipment.

    But on the sub building front, I actually came up with a nice design for the 10" I have. I've sized it up so that I can make it out of 1 sheet of 1200*1200 17/18mm thick plywood, with a single vent from 90mm storm water pipe, both from Bunnings. So it should be a nice cheap beginner project, to see if I like building speakers. I certainly like the idea of building speakers, but the research and execution is another matter (much like engine building - I'm quite interested in the technicality of air flow, cam profiles, etc, but I really can't be fucked doing the work). I already have a more suitable plate amp for this build too - just a little 100W with simple filters and no boost. If I ever get around to it (I've had the 10" sub for years), I'll post a thread.
     
    Last edited: Jul 6, 2022

Share This Page

Advertisement: