My real OB speaker project

Discussion in 'Speakers' started by Serious, Mar 13, 2017.

Tags:
  1. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Figured I'd post a pic of what my OBs look like at the moment. They pretty much look the same as they did this time a year ago, but I haven't really posted a proper pic since. I think the only other pics I posted are here and here. I still want to improve the baffles, but I also want to do it right and with measurements. I don't have an accelerometer yet, so I'll have to buy one first.
    I tried to add some felt on the edges of the woofer baffle and I found out that you can easily overdo it. This 5mm piece of felt did smoothen the FR a tiny bit (0.5db less at 4kHz), but adding too much makes it worse and rolls off the mid-treble off-axis. I slightly prefer the sound this way, but it's not a big difference.

    Amon.jpg

    I also haven't really posted a FR in a long time and it's quite a bit better now. Here's what the FR looks like now:
    VoxOB FR at LP.jpg

    I think I can improve on both the FR and the sound in general by adding some damping on the back of the baffles, some braces, etc, but I'll wait for now. It's good as is and I don't really feel the need to improve it.
     
    • Like Like x 14
    • Epic Epic x 1
    • List
  2. sphinxvc

    sphinxvc Gear Master (retired)

    Staff Member Pyrate BWC
    Joined:
    Sep 28, 2015
    Likes Received:
    3,320
    Trophy Points:
    113
    By far the coolest speakers on SBAF. And I don't just mean aesthetically, but from a design and execution perspective.
     
  3. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Voxativ OB Phase 3 (finished speaker) - Amon

    Amon Triptych small edit.jpg


    I mentioned multiple times how I wasn't quite contempt with my speakers the way they were, but that it would take some time before I wanted to continue to improve them. To be honest I always knew it would be a matter of years. I could've done it last year, but didn't really want to fully commit to the project. Well, this year I sacrificed some two weeks to properly finish the speakers. At least to the degree that I had envisioned it. Here's a quick rundown of what that meant to me.
    Quick posts about each part to follow:
    As with these kinds of DIY projects there's probably always something that could be optimized. But I consider them finished now. Which doesn't mean I will move on to the next project any time soon. My proposed 3 way coaxial desktop speaker project turned out to be a doomed from the start when the only driver I liked turned out to be a colossal POS (high distortion) - I bought LS50 Metas instead. Apart from that I'm out of speaker ideas and a good SET amplifier for the Voxativs would be expensive as hell.
    Sort of jokingly I will refer to the speakers as "Amon" sometimes in reference to the egyptian god Amun-Ra. Even after all this time I think they're giant killers in the most important ways.
     
    • Like Like x 9
    • Epic Epic x 9
    • List
    Last edited: Mar 13, 2022
  4. Thad E Ginathom

    Thad E Ginathom Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    14,135
    Trophy Points:
    113
    Location:
    India
    Stunning.

    :bow: :bow: :bow:
     
  5. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    The cabinet

    Baffle Damping - Bitumen + Lead.jpg

    I glued* (with special damping glue) a piece of 3.6mm bitumen and on top of that a plate of 1.2mm lead in a strategic location to the back of the front panel. It looks small, but it does the job. I chose lead because I wanted something heavy with high self damping. Another option would've been to use a thicker plate of stainless steel, which would've been a lot more expensive but also be stiffer which would make the bitumen act more in a constrained layer damping scheme, I'm not sure which one would've been better.

    When knocking on the front panel the sound is reduced by 15 dB!
    I also added a brace between top and bottom (which you can see to the right) and another one between the ends of the U-frames (which you can't see just yet).

    So what did the accelerometer measurement look like?

    Accelerometer reading on center of woofer baffle (subtracted filter response).jpg

    This is honestly all I could've hoped for. The peak at 700-800 Hz is reduced by roughly 30 dB!

    I also added Dynamat to the woofer basket with similarly spectacular results. Tapping the woofer basket it now sounds solid instead of the ringy sound it created before.


    *I secured the lead plate with four screws because the last thing I want is a large plate of lead to come flying down more than a meter. I don't think this will affect the results much, if anything it might add a bit more stiffness.
     

    Attached Files:

    • Like Like x 4
    • Epic Epic x 4
    • List
  6. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    The damping

    wool damping.jpg

    I went for 8cm of sheep's wool covering the whole backside. I later realized I had to slightly widen the opening around the woofers (effectively chamfering the wool), but otherwise it still looks like it did in that image. Glued on with textile spray glue. I tried covering the Voxativ baffle, but it always sounded worse to me. Covering the U-frames however makes it sound cleaner to me. The crossover is normally also covered in wool, which is why there's a gap on the lower part of the wool.

    You can also see the brace I mentioned above, plus the finished crossover board in its position with the new woofer cables. This also gives you a glimpse of how the Voxativ driver is mounted.

    Of course I had to measure the effect it has on the frequency response:
    Amon - effect of wool in back - average from 15 to 180 degrees.jpg
    The massive boost in nelow 50Hz is likely an artifact from the wind that day. It wasn't as windy when I measured the other speaker the other day.

    There was a bit of a comb filter action going on with a dip at 500Hz, 750Hz and 1kHz before that is now gone. I also think it's possible that the upper midrange is improved. The CSDs certainly look like it.
    Note that this measurement completely negates what I thought back in 2017 that the Voxativ 2kHz peak in anechoic conditions is from the driver. Most of it was the frame all along. There's also a remainder of a 4kHz peak from the baffle, which can be seen in the CSDs, but it averages out in-room.

    I specifically bought Monacor MDM-3 for trying Voxativ baffle damping schemes, but none helped with the 4kHz peak and no matter where I put it (on the back of the baffle or directly behind the driver) - it always hurt the openness.

    In-room the frequency response looks largely similar, except for less of a dip around 200Hz and some more bass extension. Maybe the midrange response is also somewhat improved. As mentioned, I think it mainly just sounds cleaner.
    I attached an in-room RTA around where I normally sit in the hammock. The hammock generally gives a better lower midrange response, but is a bit harder to measure in the upper midrange.
     

    Attached Files:

    • Like Like x 4
    • Epic Epic x 4
    • List
    Last edited: Mar 3, 2022
  7. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    The crossover

    XO boards small.jpg


    The values all stayed the same, but I ended up switching my DIY graphite resistors for pathaudio resistors. Sadly the Pathaudios do sound noticeably better with a more free flowing, less grainy presentation. Smoothness and clarity could be the theme of the whole speaker upgrade and it's mirrored in the crossover upgrade.

    I went for an XO board with multiple layers. If I had to do it again I probably would've used Bitumen as the damping layer since the board was still somewhat resonant with the inner layer of Polystyrol and outer layers of thin plywood, with an extra plate of HDF on the bottom. The whole thing rests on a soft 20 PPI PU filter foam for decoupling. I use beam clamps and threaded wooden inserts to both secure the crossover board to the frame and connect the wires. This works well for directly connecting the flat speaker cable I have to the flat coil and I use M8 crimped ring terminals for the other drivers. There are even hollow screws with a 4mm hole which would allow banana plugs to be used. I quite like the unconventional beam clamps since they allow direct cable-to-cable contact with plenty of mounting pressure.

    beam clamp connectors small.jpg

    I chose to build the whole board with an additional ground connection, which interfaces with a wooden ground box, which is then connected to the GND of the power strip. Currently my wooden box is empty and the ground connection is connected directly to the power strip. All the driver baskets and the pathaudio resistors and CDE caps are connected to ground. I use multistrand tinned copper for the ground connection.
    You will often read that a ground connection subjectively lowers the noise floor. I find it also improved the imaging in my system.

    The woofer cables are 6mm² tinned copper for the lower woofer and two of the same 6mm² tinned copper for the upper woofer to compensate for the twice as long cables. The cables are PVC insulated and only sleeved with cotton. I find the tinned copper insulated with PVC to actually sound quite good.
    The ground wires are thinner and are unsleeved PVC, except for the Voxativ ground wire which is the Duelund DCA20GA. You can see the ground wire to the right and in the picture I attached.
    I attached two images which give you a better idea of how I secured the ground wire to the baskets. It was a no brainer with the Voxativs and for the woofers I just had to screw in an M4 screw into a hole which was already there, secured with a nut from the inside.
     

    Attached Files:

    • Like Like x 3
    • Epic Epic x 3
    • List
  8. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    The cables

    Both cables small.jpg

    Based on the cable comparison I did a while ago: https://www.superbestaudiofriends.org/index.php?threads/cable-building.86/page-38#post-296533
    For the woofer cables I went looking for what is essentially a flat, braided, tinned copper cable with decent area while not breaking the bank. Turns out tinned copper shielding braid was prety much just what I was looking for and it can be had for decent prices. I ordered 12mm^2 braid and sleeved it with cotton. The whole wire (including the sleeve) is roughly 18mm x 2mm in size per phase, so roughly 4mm thick in total.

    For the Voxativs the choice was clear based on the cable comparison I did: Mundorf silver/gold foil. The problem is that the 11mm one was both somewhat large in area and too expensive for me. So I ended up ordering half the 11mm I would've needed and cut it into two 5.5mm wide strips myself.
    divided silver gold foil small.jpg

    It turns out dividing the cable was not as hard as sleeving it. In the end both jobs were manageable, however.
    So what does it sound like?
    It gave me the slight boost in upper treble I was looking for. Otherwise it just sounds super smooth/liquid and resolving as hell. Imaging has more depth. Let's just say I'm not disappointed and it sounds like what I was looking for in a cable for the Voxativs.

    For a better perspective on size, here are the two cables next to a string of relatively thick AV and power strip cables.
    cables on floor small.jpg
    (Added another picture which puts all three speaker cables into perspective. 2.5mm² ground, 12mm² woofers and 0.385mm² Voxativ.)

    I'm currently using these antistatic foam packing peanuts as cable lifts. I'll replace these eventually with conductive foam cut to size. I think they're a good sounding option, but not really necessary with cotton insulation and not really pretty.
     

    Attached Files:

    • Epic Epic x 5
    • Like Like x 2
    • List
    Last edited: Mar 3, 2022
  9. Walderstorn

    Walderstorn Friend

    Pyrate
    Joined:
    Mar 20, 2016
    Likes Received:
    1,904
    Trophy Points:
    113
    Location:
    Europe
    Thanks for sharing your project. It look very unique and interesting!
     
  10. Thad E Ginathom

    Thad E Ginathom Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    14,135
    Trophy Points:
    113
    Location:
    India
    Astonishing.

    The only thing I feel able to comment on, rather than just admire/drool over, is the difference between steel and lead. Steel rings, lead is dead: I'm sure you made the right decision!

    But, btw, how easy is it to get lead these days?
     
  11. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I just ordered it from eBay. Rolled lead looks to be available from many places.

    The feet are still going to take a while to get here. Three things in the meantime:

    The finish

    close up far.jpg

    I used the same Auro white hard oil 126-90 I used before, but instead of applying it as intended I basically used it as paint. I used wood stain for the back which ended up looking very similar while not taking nearly as long to apply. I basically spent a week sanding and then painting the speakers. And the finish is still far from perfect.
    My dad used the same hard oil in a similar fashion for his DIY turntable and it got somewhat darker over time. It took me some time to get used to the color, but now I think it looks great, especially in the right light.


    Updated frequency response

    Amon - in hammock with wool under bed RTA and sweep average.jpg

    I put the large roll of sheep's wool under my bed again (the one I use for measurements). It acts as a bass trap and in this case helped quite a bit with bass linearity. I haven't heard the bass of my OB speakers in such good quality before. It even sounded somewhat sluggish before which is mainly because I was used to the sound with the wool. Due to the additional 5kg of wool in the back of the OBs I think I'm even getting a bit more damping of the 40Hz room mode. Coupled with the improved extension it now sounds more controlled than it ever did.


    DIY ground box


    Outreq.jpg

    I had experimented with a cable connecting the Voxativ basket to the electrical outlet's ground connection before I built my crossover with three connectors. I kind of wanted a ground box, at least in part because I thought it was funny. Which is also why I filled it with literal dirt. I built it so it can be used in three ways: A direct high conductivity connection with a wire connecting input to output, a low conductivity connection connecting input to output and simply not connecting the output at all like the Entreq boxes are built.
    I initially aimed for 1 MOhm for the high impedance connection since that is what ESD safety equipment seems to be built with. However over time it became more conductive. After two days or so I'm now at 1/4 that.
    Subjectively it sounds better, the lower impedance to ground I have. The direct cable connection sounds the cleanest, least grainy. The connection through the dirt seems to have become better sounding over time.
     

    Attached Files:

    • Like Like x 3
    • Epic Epic x 3
    • List
    Last edited: Mar 6, 2022
  12. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,470
    Trophy Points:
    113
    Location:
    Winnipeg
    I would recommend replacing the dirt with some dry sand or rocks. Anything with moisture content can lead to long term issues (random things growing, rot, mould, etc).
     
  13. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Heh yeah, I think I'll have to do that. I thought the soil was reasonably dry when I bought it, but after it not closing properly the first night it already attracted some tiny flies that I can't quite get rid of completely at the moment. For now I sprayed it with 99% isopropyl alcohol. Even baking it probably doesn't get rid of all of the humidity and I bet that'll make it much less conductive.

    There are these tiny steel pellets available for a reasonable price: https://ballast-produkte.de/lochkappen-aus-ballaststahl-vorverpackt-im-polypropylenbeutel/
    Entreq seems to use iron chips, but that somehow seems to be more expensive. Sand is also quite cheap and coal is even cheaper. I'm pretty sure the cable is going to sound better anyway, so I could also just use the gravel I have. It just won't be conductive.

    Before I built the box I tried to see if I can measure something with my multimeter. Set to the uA range I could see a miniscule increase in AC current. Something like 0.7uA vs 0.04uA noise. I think I remember that connecting the driver baskets also increased the noise so I thought they also act as an antenna somehow. I want to provide a low resistance path to ground and hope I'm not just injecting noise that way.
     
    Last edited: Mar 6, 2022
  14. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,470
    Trophy Points:
    113
    Location:
    Winnipeg
    Fruit flies lay their eggs in soil where the larvae eat decaying matter (which nice soil is full of). They aren't super difficult to get rid of, but they are incredibly annoying and multiply very fast.
     
  15. Thad E Ginathom

    Thad E Ginathom Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    14,135
    Trophy Points:
    113
    Location:
    India
    Ground box? I'm inclined to laugh, but hey, I do not know or understand its purpose, so I'd better keep quiet!

    As for the soil: Wouldn't an hour or two at high oven temperatures leave it pretty-much sterile?And completely dry.
     
  16. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    The feet

    Carbon Fiber + Sorbothane disks speaker feet.jpg
    (Yes, they are small. I wanted to keep a certain ratio between width and height while keeping the speakers close to the floor. Plus these didn't break the bank.)

    I finally received the missing piece for my speaker feet: The carbon fiber disks. I replaced the 8mm alu disks with 5mm carbon fiber and 3mm sorbothane disks. I wanted a thin Sorbothane to give me rigid coupling for the subbass and decoupling from upper bass on. My speakers weigh only about 100lbs. According to the Sorbothane vibration calculator the relatively soft 50 Duro is just about right, giving me a resonance frequency of 65-ish and 95-ish Hz for the front and rear feet respectively. It looks kind of bad in the right picture, but the measured deflection is roughly 17% - lower than the maximum recommended 20%.
    Another option would've been to use thicker disks for a resonance frequency closer to 30Hz, but there's not a lot of content in that region. For a subwoofer where you don't want to annoy your neighbors that would be the better choice.

    So how does it sound? The spikes and alu disks were already an improvement over coins I was using before (which were an improvement over the felt I used for easy positioning). Well, this combination is better still. I haven't tried the carbon fiber by itself, so I can't say how much is due to the carbon fiber and how of it is due to the Sorbothane, but I'm willing to bet most of the improvement is from the Sorbothane.
    The midrange is somewhat clearer with somewhat better imaging. The upper bass is also a bit clearer and subjectively sounds at least a bit more linear. The lower bass and subbass sounds nearly unchanged, which was my goal. If anything it sounds even more solid and thunderous, possibly because the Sorbothane really sticks to the floor.
    One thing that's definitely noticeable is that the floor vibrates noticeably less now, especially in the upper bass and lower midrange.

    I tried to measure a difference and while I can't be sure that what I measured is due to the feet vs just random variance (they were done a couple days apart), I did see a slight trend for a more linear upper bass in the measurements. It could be wishful thinking, but I think at least part of it could really be from the feet:

    Amon - on couch - rough difference of speaker feet LR1.jpg

    I averaged left and right for below 450Hz and then overlaid the earlier result from 100Hz to 300Hz. Of course it's entirely possible that this is either cherry-picking or that I moved the speakers slightly. Both measurements were spatially averaged RTAs. I also attached what just the left and right channel looked like for the new configuration, merged with the sweep at the listening position for sub 100Hz.

    That's pretty much it as far as my speakers go. I'm also not planning any major upgrades to my setup in the near future. The last upgrade was a Pi2AES, actually. It's now my main source. Otherwise I'm still on the Gungnir MB Gen 1 + Ragnarok. I'll upgrade the XLR cable from the inakustik copper DIY cable I'm using to a DIY silver in oversized PTFE jacket soon. A good sounding combination IMO, possibly somewhat more neutral than cotton, but not quite as natural. That's after comparing the two. I may experiment with different cable lifts or cheap aliexpress power cables and/or power strips and that's pretty much it. Sure, a better amp and DAC is going to make a difference, but for now I feel I've arrived.

    EDIT: XLR + RCA cables built: https://superbestaudiofriends.org/index.php?threads/cable-building.86/page-49#post-378977
     

    Attached Files:

    Last edited: Apr 1, 2022
  17. Locman

    Locman New

    Joined:
    May 26, 2021
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Canada
    Hey guys,

    I am a OB newbie and not a builder but interested in PAP Duets with the Voxativ mid driver. Do they measure decent as compared to your custom designs? Thanks!
     
  18. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    @Locman I don't have experience with the PAP Duets directly, but since I started out with one woofer and had my PAP clone configured in TMM configuration a while back I think I can at least help you somewhat. I also compared my Beta 15A woofers to the PAP woofers. I will try to be as unbiased as possible, though I'll admit I love these speakers (my Voxativ OB).

    What you aren't getting is the most precise, deep and impactful bass from a single 15" woofer. The second woofer really helps mitigate some wooliness. I also find the PAP woofers to be a little less clean, especially at higher volume levels compared to the dirt cheap Beta 15A. I know others think otherwise. As such I would expect there to be an even larger difference with the PAP woofers when running just one woofer vs two per side. IME the PAP woofers really sort of fall apart at higher volume levels.

    In this case what you need to ask yourself is this:
    Am I listening at higher SPLs frequently? Do I want the sound to stay as clean as possible while doing so?
    If you don't frequently listen at 80 to 85 dBA or more the difference will likely be minimal.


    As for your question about the measurements, there are really three answers. Note this applies to the Trio and I can't say how much of it applies to the Duet:

    1. Anechoically they're not close to my design. See page 3 of this thread. See this post in page 3 of this thread. The shallow crossover combined with the acoustical delay the woofers have make a decent frequency response practically impossible to achieve anechoically. However:
    2. In-room the measurements are totally fine. As far as I know the design was voiced by ear from a legend that is sadly gone now. They are carefuilly balanced to be as linear in-room as possible. With one caveat: They are unapologetically bassy. There's a definite boost in the mid to lower bass region to make placing them closer to a rear wall possible.
    3. By ear they still sound neutral, though I would say not as smooth as my design and not as extended. Mine seemingly go both lower and also higher in frequency.
    I've seen designs with one woofer and a widebander on a piece of wood with very good linearity in the midrange anechoically (which the PAP Trio lacked) by using a different crossover philosophy. So perhaps these are different. The downside would likely be more crossover parts, which raises cost dramatically if you want the most transparent sound. Good sounding crossover components (and these speakers would make good use of them) - say Duelund or TOTL Jantzen and Mundorf components, etc) - can cost a fortune.

    In any case what you're getting should be a very transparent and honest speaker, as long as your room allows for an OB speaker. My speakers did not sound that great in the living room, which is a less damped room and as such emphasizes the upper mids a lot more. They sound too thin in that room.
     
  19. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    2023 Updated measurements (brutally honest):


    Frequency response:

    Amon FR in room.jpg


    Distortion:
    Amon distortion.gif


    Now distortion doesn't look super impressive, but you have to consider that while D2 is kinda high in the midrange, D3 is very low and D4 and D5 aren't anywhere to be seen.
    The treble distortion plateau can likely be ignored since the distortion products are in the top octave and above.

    Overall not too bad a result, I think. But it's not as perfect as what I had shown previously, especially the frequency response, which does have a couple wiggles in the upper midrange. Compared to Hathor it can be a bit too much, a bit overbearing at times, where Hathor is smoother or more laid-back.
    Plus Hathor has more top octave energy and a totally different presentation.

    Compare to Hathor here:
    https://www.superbestaudiofriends.org/index.php?threads/diy-cardioid-monitors.13292/#post-416191
     
    Last edited: Sep 18, 2023
  20. EagleWings

    EagleWings Friend

    Pyrate Contributor
    Joined:
    Jul 20, 2017
    Likes Received:
    1,709
    Trophy Points:
    93
    Location:
    India
    For a widebander, those upper mid wiggles look so much better. With large widebanders, 99% of the time, I typically see worst response in that region than what you got.

    I’ve been curious about what contributes majorly to the ‘overbearing nature’ of many widebanders and in this case, there is not much in the FR and Distortion measurements of both speakers that suggest the Amon to be more overbearing than the Hathor.
     

Share This Page