SET and Fostex BLH Speakers

Discussion in 'DIY' started by purr1n, Feb 10, 2016.

  1. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,042
    Trophy Points:
    93
    BTW, Bottlehead has a passive baffle step corrector that goes between source and amp: http://bottlehead.com/?product=the-fix
    I have one built that I don't use. Happy to loan it out to people building their own speakers. I think the idea is it makes it easy to try out different EQ before soldering a fixed one together. Here's info from the "More" link on the site:

    The circuit’s designer, Paul Joppa, explains the rationale behind the kit:

    “The impetus for getting involved several years ago was the observation that many home-brew single driver systems needed some baffle step correction, which was (and still is) usually done at speaker level with a choke/resistor parallel combination. This combination reduced the effective midband efficiency to less than half of what it had been, meaning you needed more than twice as much amplifier power. This was a big problem when used with flea-power SET amps, and caused a lot of 45-powered speakers to suddenly need 300Bs.”

    By doing this bass and/or treble compensation ahead of the amplifier instead of inside the speaker one can use the volume control to adjust for the relative drop in the midband output level, thus retaining effective midband efficiency of your speaker.​
     
  2. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,042
    Trophy Points:
    93
    That's really interesting. I never knew the Nagas were "open top"! This design comes from a book by Nagaoka that I own (I posted about it earlier).
     
  3. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,957
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    @JK47 and @struggles FE168EZ Nagaoka project has inspired me to do a BiB (Bigger is Better) type horn for the FE126EN and FE108EZ 4" drivers I had lying around. I know @Hooncake tried out the Madisound desktop kit and said there was not enough bass. (My fault, I forgot to mention to not expect much bass), but I think the BiB might be fantastic for a 4". These bigger horns longer do seem to be the solution. Like the Nagas, the BiB's should be open up top.

    More details later.
     
  4. JK47

    JK47 Guest

    The Blumenstein Naga's are based of Nagaoka-san's F-81 enclosure. Blumenstein said he tweaked the tuning somewhat to account for the original Fostex driver intended for the F-81 no longer being available.

    I have heard the Naga's paired with the Blumenstein woofers at @shaizada's, and they are amazing.

    Here's a build and plans for an enclosure similar to the one we just made, but for a 4" Sigma or the new 4"FE-s driver that Madisound carries.

    http://spida.art.coocan.jp/pag9-0-27.htm
     
    Last edited by a moderator: Nov 9, 2016
  5. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,244
    Trophy Points:
    113
    Location:
    Northwest France
    I have acquired a couple of fe103-sol 16 ohm drivers! According to some forums the T/S parameters aren't suited for horns but they would work in a bass reflex or transmission line. Thinking about a Woden vampyr-v or frugal horn

    http://www.wodendesign.com/Vampyr-V.html

    But those are long term projects. In the meantime I'm probably making simple sealed boxes so they get some desktop use.
     
  6. JK47

    JK47 Guest

    IMG_0215.jpg
    Finally after months and months (since the LAX Show), I finally brought my speakers home. They were sitting in @struggles shop, I would stop by here and there, sanding and trying to get a uniform finish so I could get a nice coat of paint on them. Easier said than done, the massive weight of these enclosures put gouges and marks on whatever side was down. It was a constant battle filling in these blemishes. After flip flopping on choice of finish, I finally settled on the old faithful, semi-gloss Black finish.

    Under @struggles tutelage, I sprayed them myself in his spray booth. After lots of streaks, runs, and multiple sanding sessions knocking down the rough finish, I cried uncle and called it a day.

    Next up, building my own L-pad was in the cards to use with the Fostex T500mkII Super Tweeters @struggles lent me. I ordered a Hammond ABS plastic enclosure, Jantzen 1.5 uF caps (may try 1.0uF in the future), Fostex L-pad (can be bought for less without the Fostex branding, but I decided to carry on the Fostex theme), and a some binding post's.
    IMG_0209.jpg

    A set of Eden sound feet will eventually be put on, I just wanted to get these suckers playing music.

    IMG_0210.jpg

    IMG_0211 2.jpg
    IMG_0212.jpg

    IMG_0213.jpg

    IMG_0214.jpg

    4 Watts from my Elekit is enough juice to power these custom modded Nagaoka D-37 enclosures with Fostex Sigma 168eZ 6.5" drivers and Fostex T500mkII super tweets.

    Thanks @struggles for all the lessons on painting, finishing, and everything else.
     
  7. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,957
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Those tweeters sounded very nice. Very memorable.
     
  8. Jh4db536

    Jh4db536 Friend

    Pyrate
    Joined:
    May 25, 2016
    Likes Received:
    931
    Trophy Points:
    93
    Location:
    Los Angeles
    Home Page:
    Haven't seen much data on BK20 cabinets and F208ez, here's my first crack at this. Forgive my noobness with REW and microphones (USB UMIK). Since i have the whole house to myself for now room dimensions are 12.5 x 18 with a large opening to a larger living room on the right side. I marked the floor so i could push it back to get the bass right.

    I used the rule of thirds to position the cabinets making basically a equilateral triangle of 5'9" in the middle of the room. Toe in is minimal because i thought the center lost focus. Also reduced the upward tilt because it made the image too high. In this position, the focus is right above the fireplace or beyond. There's depth and width is a little limited to just the wall about 3 feet from the cabinets.

    Norah Jones and John Mayer sound like theyre here performing in this room for me. Imaging is great.

    [​IMG]
    [​IMG]
    [​IMG]
    BK20 on axis mic at 24" horizontal parallel to the floor
    No subs no tweeters, only 1.25 oz of stuffing in each cab so far (about a ziplock bag amount which is very little)
    DG300B amp, Gamma3 DAC
    These seem to roll off earlier than the F168ez.
    Super tweeters and lpad parts are in the mail.
     
    Last edited: Jun 15, 2017
  9. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,957
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Very nice. Looks like a slightly rising response until top end rolloff. Did you find it shout at all?
     
  10. Jh4db536

    Jh4db536 Friend

    Pyrate
    Joined:
    May 25, 2016
    Likes Received:
    931
    Trophy Points:
    93
    Location:
    Los Angeles
    Home Page:
    Yes there's a little shout and rasp. Sometimes it sounds thin depending on the track. I thought it would be more severe.

    Do you think that could be fixed with additional stuffing in the blue area? Or do I need PEQ to fix this?

    Have a recommended PEQ calculator or software?

    Would a paralell RLC : 10ohm, 1.5uH, 1.8uF work?
     
    Last edited: Jun 16, 2017
  11. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,957
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    PEQ, or an RLC circuit (with appropriate values) I used (a few posts back) should work.

    I suspect there might be a 7kHz peak if you take a nearfield measurement.

    The thinness... move the speakers closer to the wall. They might need more 2pi reinforcement.
     
  12. Jh4db536

    Jh4db536 Friend

    Pyrate
    Joined:
    May 25, 2016
    Likes Received:
    931
    Trophy Points:
    93
    Location:
    Los Angeles
    Home Page:
    I received the super tweeters yesterday and Lpad parts. Built that right up with a 1.2uF (range is 1 to 1.5uF). Based on what i read the FE208EZ is happy with a 1.2 through 1.5uF, which i think would put me in the correct range needed to dial in the Fc point to be where it needs to fill in the FE208ez roll off. The 1.2 might work better than a 1.5uF with the 168EZ since it rolls off later and could probably work better with a higher Fc.

    I think i bought the parts from the same place as JK47 lol...they only sell 1 box that is capable of fitting everything although i couldve went expensive and bought something like this
    www.reallygoodstuffs.com/L-PAD Enclosure.htm

    The cap was bigger than i thought. i didnt even bother to look at the dimensions. Almost didn't fit and i had to improvise a little but it worked out. I'll have this whole thing working by end of day or tomorrow.


    [​IMG]
    [​IMG]
    [​IMG]
    [​IMG]

    With and without super tweeter at listening position. T900 unplugged and overlayed.

    [​IMG]

    I spent today playing with stuffing. I ended up with a total of 1.5oz in the blue area and 0.75oz in the green area. It is my belief that anything more than 1.5oz in the blue area makes the horn sound muffled/congested and kills clarity. There should just be enough to kill the rasp and fill out the sound. 0.75oz is pretty much filling up the entire green cavity and this cleans up the bass slightly.

    These measurements are done on the 300B amp and the Gamma3 DAC which is warmer than Yggdrasil and the 2a3. As you can see the midrange hump between 700hz to like 8khz is NOT Shout or Honk - to my ears it is actually smooth midrange emphasis like what you might expect from stereotypical tube (300b) sound signature. It's not "n" shaped exactly, because the highs are sweet, just sparkly enough without being offensive. I might've just measured the tube amp effect. This will be interesting to measure the EC amp later because it sounds a lot different than the 300b amp. To get to the Bruel & Kjaer target, i might even need to extend the subwoofers response through 350hz.

    Again at listening position - this is where i ended up in optimizing the stuffing. Except for the bass performance I could probably quit now right here and be content. I removed the tilt (completely level now) and the bass hump came down.

    [​IMG]

    Going to try messing with the height and maybe pull the cabs out another 3" from the wall to attack that 200hz bump. I'm pretty sure a Digital MiniDSP would fix this easily, an will allow me to keep it minimally invasive - to stick with the philosophy of nothing between the amp and drivers, avoiding PC based source/software and full retard room treatment.

    Long-term solution is to treat the room with bass traps on the wall behind the cabinets and call it a day...if i can find some on craigslist.

    [​IMG][​IMG]

    According to the Schroeder Freq it's around 211hz for a room of this dimension and would be similar to the other room i would put my setup in. I ran a sinewave at 200hz and started walking around with the microphone.

    If i were ever to treat a small room, the most effective spot to position a low freq tuned bass absorbing panel would be the wall directly behind the cabinets (+10dB). Second place would be in the front corners at tweeter height (+7dB). Third would be on the side walls in front of the cabinet about 4 feet (+5dB).

    Conclusion:

    Everything below 250hz is not really existent even though i measured something there. Really need a sub or a woofer to help even on BK20. I thought these would fare a little bit better than BK16 in this regard it's probably about the same. Imaging, staging, female vocals, guitars is like A+. The depth goes 5-10 feet past the cabinets and the width is about 5 feet past the cabinets depending on which DAC and amp i use. It puts up an image so real it's like there's actually a live performance in the room. It's highly resolving of all the nuances of vocals and guitar string plucks. It's throwing loads of details at you (especially the more aggressive EC amp). EDM jazz where there's a lot of bass is not gonna do well without a sub/woofer...it flat out sucks, but this is what you signed up for when choosing relatively small horns.

    The bass that is down there is described as having no attack, but only decay or sustain. Is it possible to make the horn roll the "bass" off steeper by putting dampening in the mouth? Maybe it's better off not being there at all.

    I feel good about the T900's, i need @JK47 or @struggles to bring over the T500's so we can compare.
     
    Last edited: Jul 1, 2017
  13. JK47

    JK47 Guest

    I finally got around to taking some measurements. The Elekit powered the speakers and tweeters, REW and Umik provided the data at listening position, using only the right channel.

    IMG_0286.JPG
    lstening position distortion.jpg listening position various tweeter settings.jpg

    Distortion measurment taken with tweeter set at -8 on Lpad
    Blue is tweeter off
    Green is tweeter at -15 on the Lpad
    Orange is tweeter at -8
    Red is tweeter at -4

    I generally set the tweeter at -10 on the Lpad, which has a 1.5uf Jantzen cap.
     
    Last edited by a moderator: Aug 28, 2017
  14. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,957
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I need those tweeters.
     
  15. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,244
    Trophy Points:
    113
    Location:
    Northwest France
    Are you running the fe168ez full range without crossover components?
     
  16. JK47

    JK47 Guest

    yes
     
  17. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Brace yourselves! Looks like 2kHz Mt.Doom is about to erupt.
     
  18. JK47

    JK47 Guest

    Thanks, for adding noise to this thread with your witty comment.

    Go back to post #27... It seems to be a trait of the Sigma driver.

    Sorry cupcake, it's the 150hZ eruption that's gonna knock you out, BAAMMMMM... KO!!!
     
    Last edited by a moderator: Aug 29, 2017
  19. JK47

    JK47 Guest

    After almost 6 months of waiting for my comrades at Viawave to get the 8ohm version of their GRT-145 ribbon tweeter out into the wild it arrived this morning.

    IMG_1363.jpg

    http://www.viawaveusa.com/grt-145.html

    I came across the 4 ohm version last year when reading up on a link posted here to the DIY loudspeaker maestro Troels Gravesen at http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm recommending it for one of his latest projects.

    His review http://www.troelsgravesen.dk/Viawave-GRT-145-4.htm

    It's been so long since I sold the Fostex T500 mkII super tweeters back to the original owner so he could sell his system in a package deal that the buyer insisted on including the T500's (they were out of stock at Madisound until July. This all happened in April).

    So far so good, they add that icing on the proverbially cake in the top end that I was so sorely missing these last months. I doubt they're as smooth as the T500 mkII's, but I would like to try them out against @Jh4db536 's Fostex T900's, and I think they would compete really well. They are $225 each, which I find reasonably priced in this day and age.

    IMG_1364.jpg
    Sitting on top of the Oreo™ packing foam stands
    IMG_1365.jpg

    They are suppose to be roughly 95dB efficient, so they don't really need the Lpad to attenuate them any where near as much as the T500 mkII's (106dB) did, mostly just the capacitor in the circuit to cross them over. My Fostex Sigma drivers are 95 or 96dB, so it's all good and they match fairly well.

    Here's a quick and dirty measurement from the listening position with both channels playing.

    Blue is no tweeter, and Red is tweeter at max on the Lpad.

    GRT.jpg
     
  20. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,244
    Trophy Points:
    113
    Location:
    Northwest France
    Nice measurements even with all the diffraction from the edge of the tweeter. Foam core baffle would be cheap and effective until you get something more permanent.
     

Share This Page