Alpair 6.2P & 7.3M / Scan-Speak 10F DIY speakers

Discussion in 'Speakers' started by Jeb, Oct 6, 2016.

  1. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    I just finished building some small full-range bookshelf speakers using the 3.5" Markaudio Alpair 6.2p. 'P' is for paper.

    The cabinets are the mMar-Ken 6.2 and I used plans from Planet_10 hifi.

    Everything went quite well and I'm pretty pleased with the results. It really does sound better when you build it yourself.

    These were made from good grade 15mm birch ply from my local yard. The cabinets have a narrow lower-vent. The driver is braced tightly from behind with a holey brace and there is a second smaller lower cabinet brace. The inside is lined on all internal sides with 12mm wool-felt.

    Cutting flush-mount driver holes is not too difficult as long as you have a reliable jig, give the set-up adequate care and attention and just take it slow. The reverse of the driver-cut-out has a wide chamfer to give the driver some 'breathing' room. The only other thing of note is that - right at the end - the front of the cabinet receives a pretty hefty 45 degree chamfer, which also requires some care... so you don't screw up all your hard-work.

    Here's some other stuff I found was important along the way:

    • Just to echo what @Marvey has said before... you really can't have too many clamps. There were times when I was using 16 clamps on these small speakers. You can see the glue line and seam completely disappear with good clampage. Decent clamps don't cost a lot but the cheap ones suck badly.
    • For the small holes in the braces: inexpensive forstner bits are worthless. If you don't want to buy decent ones (a good 40mm forstner bit is £50+) use holesaws instead. Take it slowly and stop to clean out the hole/bit regularly to avoid wood burn.
    • Titebond 1 glue is good stuff. On the plus side it has a really good strong initial tack and sets quickly. But you need to work quickly and methodically and pre-plan your clamping because there's very little time for adjustments.
    • I found that the best thing for cutting the thick wool-felt was quality tailor's scissors. A stanley knife with a hooked carpet-fitter's blade also works but is not as accurate.
    • General Finishes High Performance Satin top-coat is nice stuff and the most fool-proof finish i've ever used. It flows and levels like a dream and doesn't yellow the wood like some oil based finishes.

    Anyway, here's a few pics I took along the way and of them in their temporary set-up with an integrated amp and a little portable Fostex DAC/ipad. So i'm clearly still in the 'just finished' honeymoon phase, but these sound really quite good given the size of the driver. I'm sure their limitations will become more apparent in due course.

    I hope to be able to put them on some stands and take some measurements and talk more about the sound soon.

    Including drivers I estimate overall cost of the speakers, all things considered, to be about £120.




    [​IMG]


    [​IMG]



    [​IMG]


    [​IMG]


    [​IMG]
     
    Last edited: Aug 12, 2017
  2. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    I've been enjoying the 6Ps but in terms of evaluating the sound I felt like i'd be talking out my ass unless there was something to compare it to. I bought some Alpair 7.3M ('M' for Metal). These are the more expensive 4" drivers. The boxes I built for these are the Slim Classic GR (Golden Ratio) dMar-Ken 7.3 – plans were from Planet 10 again.

    Box design is similar to the last one with a larger lower horizontal vent. These cabinets are a fair bit bit bigger (taller & deeper) and the internal bracing is a bit more hardcore. The brace makes these a bit harder to build, so the 6P cabinet would be a better first build.

    Here's the build..

    Here are the main pieces routed & ready to be glued. The brace does take time to get right – to allow for compression of the gasket and fit tightly, but not too tightly or it can stress the basket and distort (I did this first of all). It needs to be 35-40% holes and angled to not obstruct the internal vent opening. The small cut-out at the back of the brace is to accommodate the terminal cups/wires.

    [​IMG]


    The Alpair 7 needs a pretty deep rebate for flush mounting. 18mm ply would be ideal for the front baffle. Since I’m using 15mm I laminated a piece 3mm ply to the back then chamfered the rear of the opening, leaving plenty of ‘meat’ for the driver screws.

    [​IMG]


    As suggested in the plans you can fine-tune the fitment of the brace with shims. I cut the brace to within 1mm of the back of the magnet then added a couple of strips of thin card to get a nice positive contact. Don't use something lossy like rubber or a gasket - I believe that the idea is that vibrations be distributed through the brace and more evenly into the rest of the cabinet. Gluing the brace with the driver in situ helps to ensure correct alignment.

    [​IMG]


    By the time the back-panel goes on, half of the cabinet has become inaccessible so you need to have already lined most of the internal walls. I used approx 12mm wool-felt again. This is 54oz 'Envirolay'. I find that good clampage isn't about sheer force so much as frequency of clamps and ensuring a small amount of pressure is applied where all pieces meet

    [​IMG]


    Here is the cross-section before sealing her up. Because the brace is fitted just off-centre, one optional step is to line the central side of the holey-brace with a light material to kill any standing waves. I used a bit of 8oz polyester wadding.

    [​IMG]

    [​IMG]


    Final step before sanding/finishing is that 45 degree cut on the front. You can do this with a hand-held router but it’s much less daunting on a router-table. Whichever you choose, don’t go near the cabinet until you have mastered the cut on test-pieces first. It would be easy to mess up your many hours of work at this point. Multiple shallow passes is the way to go!

    [​IMG]


    After 3 light coats of clear finish, wire them up then give them a listen on my super new slightly too low stands available from your local Scandinavian furniture emporium. This pic gives an idea of the cabinet size difference for the 6P at the back and 7M. The 7M cabinets are really getting too deep to fit on a desk.

    [​IMG]

    [​IMG]


    Close-up. Good grade, clear-one-face birch ply has a nice pale lustre and some really cool fine details in the grain.

    [​IMG]

    Total cost of these speakers, not including my time. All things considered, inc drivers: ~£180

    Listening impressions coming next....
     
    Last edited: Aug 12, 2017
  3. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    Listening:

    I’ll make this as brief as possible. The more speakers I listen to, the more I realise I don’t know very much at all. I have no idea how the Alpairs stack up against other small full-range drivers. I’m much more comfortable talking about wood & building the things.

    I set myself up in a chair about 6 or 7’ from the speakers. Room is about 18 x 18' I also threw some other small 2-way bookshelf speakers – the Rega RS1 – into the mix.



    6P VS 7M

    • Both these drivers/cabinets produce really very surprising quantities of bass. It seems that the cabinets with the wide unobstructed horizontal vent might have been tuned to create a slight mid-bass emphasis. Both do better than expected even with electronic music. They can both resolve rapid drum beats on Chemical Brother’s ‘Star Guitar’ better than stock HD650/Crack. I have a 10” sub but don’t feel compelled to try and integrate it with either speaker at this point.
    • The 7M goes lower, hits harder and is significantly better controlled in the bass. The 6p seems a little sloppy and overly wet in comparison
    • The 7M holds its own much better when tasked with filling a room this size. The 6P suffers with some compression while the 7 seems more effortless and with improved separation & imaging. On ‘Carnival’ by Natalie Merchant, the 6P begins to sound a bit messy & congested in comparison.
    • The 7s also resolve more detail. On ‘Paris, Texas’ by Ry Cooder there is some fine detail and spatial cues that the 6Ps miss that makes the 7s more atmospheric.
    • One area where the paper drivers might have the edge is in the naturalness and richness of the midrange. Both drivers sound quite similar but the 6Ps perhaps have a touch more inner warmth/realism in the mids.
    • Both speakers are very smooth, glare-free easy listens with really sweet sounding vocals.
    • The 6Ps are perhaps better suited for nearfield listening. There’s a slight softness to the sound that makes them a bit friendlier for me in this position.



    Alpairs VS Rega RS1

    • So these 2-ways do do some things better than the single drivers. There seems to be better treble extension and sparkle – and I realise how I miss some of that with the Alpairs. Less bass presence and extension with the Regas but perhaps a little more tightly controlled.
    • Listening from different positions is so much better. I can get up and walk around and still enjoy these speakers. They disappear more. With the Alpairs, I start to feel like I’m missing too much information.
    • However, the Rega’s just sound overall noticeably flat in comparison, especially in the mids. There’s clearly less of a sense of music coming from REAL instruments such as the sound of a guitar coming from strings & a wooden cavity – this is much more tangible on the Alpairs.
    • Leaner, Less tonal richness and realism. Loss of clarity in the midrange.
     
    Last edited: Nov 1, 2016
  4. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    Superb writeup. Get a measurement mic, it will give you a good insight into what you can, cannot and are hearing.
     
  5. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    @Hrodulf thanks! I want to get a Mic. Can you recommend one? I should get something calibrated or with reliable calibration data right? I was looking at this one I saw recommended:

    http://www.isemcon.net/ashop/product_info.php?products_id=79

    It seems like a lot of money. Is there anything cheaper in Europe/UK that will give reliable results?
     
  6. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    This seems to be the go-to option for many enthusiasts - https://www.minidsp.com/products/acoustic-measurement/umik-1

    Under 100$ most measurement mics use the same capsule. What sets them apart is if they have an individual calibration profile. The only con for this mic is that it won't be capable to work with our software due to lack of synch between ADC and DAC clocks.
     
  7. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    I have a mic on the way :) In the meantime, I've gone in a slightly different direction than Bigger is Better.

    [​IMG]

    These are the Scan-Speak 10F-8424G00 ( http://www.scan-speak.dk/datasheet/pdf/10f-8424g00.pdf ) 4" fibre glass cone. The tiny cabinets are the nScan-Ken - free plans are from Frugal-phile - adapted for 15mm plywood instead of 12mm.

    These were really just a fun way to get maximum usage out of 1 sheet of ply, but i'm finding these drivers to be really excellent, even in these mini boxes. Based on what I've heard these seem deserving of a bigger and better application. I think they are most often used in 2 and 3-way designs.

    Compared with the Alpairs, the build quality is really nice - just simple and extremely solid with a smaller but very powerful magnet. It's not really a fair fight because in these cabinets they are very bass-limited but they still produce an enjoyable sound for a small system.

    In a fair few areas I think they are better than the Alpair 6 and 7 :
    • more extended in the treble
    • greater resolution in the midrange. They do a better job of capturing intonation, micro-detail and subtleties in vocals. More plankton.
    • A little cleaner and more precise.
    • Despite the lack of low-end the overall speed, transient response and attack in other areas is more compelling.
    • It seems like listening position is less critical than the Alpairs, although there's also a bit of boxiness in these cabinets.

    On the negative side: I don't think they are quite as smooth, warm and fluid as the Alpairs in the mids - and that does bother me a bit - but I would say they are technically more proficient in other ways and being able to fill a small room and enjoy high quality sound from such a small footprint gives me a lot of pleasure. I'm not sure about integrating my sub because I think I'd have to have the crossover set very high - but I'll give it a go.

    Not much to say about the design/internals. Similar vented design but these have a vent-spacer in the middle. I've seen it recommended in other plans to use a different kind of wood for the spacer so I used a bit of pine. Felt on all internal walls. I've been using these with a Topping 30w amp for a mini system for my workshop. The Topping amp is fine for £60 but somewhat dreadful once I compared it to my regular amp.

    PS - thanks to the SBAF Audio 101 thread for help with descriptors/terminology!


    [​IMG]
     
    Last edited: Aug 12, 2017
  8. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    I took some measurements this weekend. Took a while to find my way and still a huge amount to learn. I have no outside space at home so I transported everything over to my parents' house and used the garden there. I don't think the conditions are ideal by any stretch (concrete floor) and the space is a bit echoey due to fences/walls but it still gave the speakers much more space than I could obtain anywhere inside.

    Measurements taken 1m from driver, 1m off the ground. I took readings at 0 degrees and 45 degrees. I wish I had done 15 and 30 degree too but it was a short break in the constant rain and I felt like I'd probably annoyed the neighbours enough at this point anyway.

    At this stage I was just hoping to perhaps correlate measurements with what I found from listening tests - starting with the basics: low and high end extension, overall smoothness of response and off-axis performance. Also, to compare vs my previous 2 way speakers.

    Set-up:

    - UMIK-1 Calibrated Mic - level with single driver or level with tweeter for the 2-way speakers
    - cheapest mic stand I could find on amazon (actually not bad)
    - IKEA Bosse stool

    (I'm not sure how much the rug helps?). iMac is not ideal field equipment.


    [​IMG]


    Measurements:


    dMar-Ken 6.2p (3.5" Alpair 6.2p) -

    0 degrees

    [​IMG]

    45 degrees:

    [​IMG]




    Slim Classic Golden Ratio 7.3 (4" Alpair 7.3M) -

    0 degrees

    [​IMG]

    45 degrees

    [​IMG]



    nScan-Ken 10F (4" Scan-Speak 10F) -

    0 degrees:

    [​IMG]

    45 degrees:

    [​IMG]



    Rega RS1 (5" woofer, 19mm tweeter)

    0 degrees:

    [​IMG]


    45 degrees:

    [​IMG]




    Thoughts:

    Some of the stuff I mentioned in my subjective impressions seems to bear out, some not so much:

    • relatively speaking, both the 6p and 7M speakers are more extended in the bass and treble than I expected them to be having read about the limitations of full range speakers. This is probably partly because my previous 2-way speakers the RS1s even with the larger woofer & tweeter really offer no advantages there - about on a par with the 6.2p. The 7.3Ms are even better.
    • obviously the mini scan-speak 10F speakers have no bass to speak of, but my feeling that they were better extended in the treble doesn't really come across in the graphs. Less so than the Alpair 7.3s again. I think some of it is that the Scan-Speaks are thinner and less body so that may give the illusion of better extension. I do still think they resolve more of the very finest details in the midrange.
    • the full rangers suffer quite badly when you step outside the sweet-spot. They really are best aimed right at you. The 2-way Regas are much more forgiving. Not much to choose from between the various full-rangers although in listening I felt the Scan-Speaks were less directional and in the frequency response look less ragged past 10Khz when at 45 degrees.
    • The 7.3M and 6.2P speakers look very similar in frequency response. The 7s have the edge in low and high end extension - I feel that this is noticeable in listening. They are also just a little smoother and more balanced throughout I don't really notice that peakiness past 10K
    [​IMG]

    • For each of the speakers i'd built I measured both left and right. From what I can see, the matching looks good. I find this quite pleasing because when you build from scratch you can begin to doubt yourself a bit. There's lots of little things that seem like they might make a huge difference in the sound I subjectively hear - e.g. how tightly to do up the mounting screws and compress the magnet against the brace.

    [​IMG]


    caveats and known unknowns

    I'm a total beginner. These measurements are not sanctioned by SBAF. I'm sure much of what I've said is very basic, but that's my limit until I play around more and can understand more. The process has been awesome. I'm not really a scientific type person at all but do find it extremely interesting to try and correlate my listening with what I might be able to extrapolate on the graph.

    Stuff I'm not sure about yet:

    • Being able to comment more on relative overall smoothness of response, minor peaks and troughs
    • Doing this outside, should I be gating these measurements? Looking at the impulse response I'm not sure if/where reflections and echoes are visible - 10ms?

    [​IMG]


    Of course any comments & pointers on the results and my methodology will be awesome. I can also post other measurements like FFT waterfalls and harmonic distortion but I don't have a good handle on those yet.

    I guess next step would be to do listening position measurements.




     
    Last edited: Aug 12, 2017
  9. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    For some reason I missed that you built Scanspeak fullrange speakers. Thanks for letting us know about this cabinet design. Why did you go for the 8424 driver instead of the 8414?

    Speaking of 2/3-way, I like the idea of having each driver in its own cabinet like Wilson Audio. Ever think about adding a tweeter and maybe a midbass module later on?
     
  10. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    The 8424 were the ones that Dave at planet_10 tested when he designed those little cabinets, so I just followed his lead really. However, he also said that the characteristics of the 8414 would make them more suitable as a full-range than the other ones. http://www.diyaudio.com/forums/full-range/285856-scanspeak-discovery-10f-8414g-10-a.html

    I haven't taken the 8424 any further. They're currently being used as little surround speakers for movies. I'm not yet sold on the ScanSpeak fibreglass as a longterm choice (based also on the other speakers I built). People say they're very resolving yet also very smooth but I don't totally hear it that way in my system. That being said, I'm reserving judgement until I get some better quality amplification. Of the small speakers, there was something about the paper cones that I found more natural sounding, so I might go in that direction next time.

    I'm looking forward to seeing what you decide for your new cabinets or whatever else you have in the works.
     
  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
    Just caught up with this thread. It's awesome to see that you got to compare the mini Scan Speak widebanders to the Markaudio ones.

    I think it mainly has to do with dispersion. I think you can see in your graphs that the Scan Speak has more energy between 5 and 10kHz 45° off-axis and comparing the spec sheets the Scan Speak definitely has better dispersion, despite the Markaudio graphs looking better at first glance, at least until you realize the scaling is totally different. The K+T, Troels' graphs and the audioexpress ones also look impressive up to 20° or even 30° off-axis. And in your graphs at 45° the 10F seems to be able to almost hang in there with your 2 way. Very impressive

    If my FE83En experiments have taught me one thing it's that in a widebander dispersion is probably more important than on-axis FR. The power response is what matters, not the on-axis FR. What's interesting to me is that the bigger metal driver seems to have wider dispersion in the top octave than the smaller paper Markaudio driver. Do the resonances in this region cause the metal to flex so that the effective radiating area is much smaller than even the paper?

    Technically yes, although your measurements look decent as is. You can get rid of the ripples in the midrange by gating them. The one at the 10ms marker (or about 3.3ms after the IR peak) is the floor reflection. With 1m distance to the ground and 1m distance to the speaker there should be a reflection at 6.5ms after the speaker emits sound or about 3.6ms after the initial impulse hits the microphone if my math doesn't fail me. Maybe the smaller speaker was closer to the ground? Either way it's close enough.
    The rug won't really do anything except for very high frequencies which are directional anyway. Windowing the response after 3.4ms will leave you with 300Hz frequency resolution - hardly ideal. You could try lifting the speakers even higher off the ground and then getting the nearfield response of the driver and port to merge the two, which is what nearly everyone does to get quasi-anechoic results without an anechoic chamber or lifting the speakers meters off the ground.

    One thing that would be very interesting - at least to me - is impedance and FR adjusted to 2.83Vrms input - this way you can calculate efficiency and compare the sensitivities of the different speakers. Although as I understand it you didn't adjust the amp output level during the measurements anyway, so they're already comparable between eachother.
     
    Last edited: Apr 16, 2018
  12. Jeb

    Jeb Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    391
    Trophy Points:
    63
    Location:
    Bristol, UK
    Hey man thanks a lot for taking the time to address those questions and help interpret the measurements. I learn a lot from your posts. Going higher off the ground might be possible next time. During casual listening I sometimes wish that the Troels' 2-way off-axis was better.

    One little thing to add is that I was able to briefly compare (but not measure) the 10F to those little Anthony Gallo orb speakers (A'Diva SE). The dispersion on those seemed good, but otherwise I thought the ScanSpeaks were better in every respect. Noticeable was the greater body to instruments and voices using the 10F. Not that these tiny cabinets are that much use for anything on their own, it was just interesting given how expensive those orbs are.
     
  13. dogears

    dogears New

    Joined:
    Mar 5, 2021
    Likes Received:
    0
    Trophy Points:
    1
    Location:
    Philippines
    great post!
     

Share This Page