A(nother) reason NOT to DIY

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
There was a thread a while back that suggested this.
At the price point of those speakers, these was my thoughts:
http://www.diyaudio.com/forums/showthread.php?postid=882222#post882222
IMHO, for that type money, there is a great reason TO DIY.
At a certain price point however, there may well be a reason not to (other than just to have fun).
A member of another forum http://www.audioasylum.com/audio/speakers/messages/226294.html pointed these out http://www.bestbuy.com/site/olspage...t&productCategoryId=cat03043&id=1138085354138
on sale up until 6/10/06 @ $40 - per pair (someone from Best Buy must have seen the thread;) ).
For that type of money, I may now agree with ScottG.
I of course went out and bought (the last) pair at the closest BB (no, I don't work for them LOL) to see what $20/speaker got the non-DIY consumer (the vast majority of music listeners). The fact that it was a coaxial made me feel a little better in case I was squandering $40 that could have been used to buy beer.
I'll be honest with you. I wasn't expecting much. Several yrs back, I pulled apart a similarly priced KLH bookshelf (that BB sold also) given to me by a friend. The only suprise there was the thickness of the (chamfered) front baffle and maybe the decent size motor on the foam surround, poly cone woofer. Very little damping inside the enclosure, cap on the 99 cent tweeter, 50 cent clip binding posts ,etc,etc. They sounded a shade better than the price - which means not very good at all. Pretty bad actually.
These are quite a step forward. Are they going to replace my DSP dipoles? No.
But what I found left me quite astonished at the manufacturing capability of cheapo chinese speakers. I've seen some pretty decent bargains of course, but at the $20 (profit included!) price point, I wasn't expecting quite this much.
The enclosure alone is probably worth the $20 to avoid the headaches of enclosure building. Yes, its probably only about 5/8" (15mm) MDF, but the sides (the largest surface) are curved, making them fairly rigid. The top/bottom/back are small enough to make the MDF thickness ok. The front baffle is actually 2pc and over an inch (30mm) thick! The inner walls have plenty of acoustic foam. There is a real XO, not just a tweeter cap.
The drive unit itself might be worth the $20. Its a real CF cone (not poly made to look that way), VC is about 1.5" and a fairly large motor structure. The fabric dome sits in a small horn/phase plug ala KEF/Tannoy etc.
Decent quality banana plugs on rear.
Now the inevitable question. What does it sound like?
I hate those blathering subjective "reviews", so let me just say this. It sounds pretty decent. Remember, I'm used to listening to full sized active OB speakers. All boxes now have that familiar "sound" that some will like. Once I (try to) get past that, these sound far better than any $20 speaker have the right to. Great? No. Good....sure.
Rather than pick apart the treble/mid/bass, I suggest going down to BB and have a listen for yourself. Even if you don't need them for yourself, they would probably make a nice gift for the non-audiophile or kid going off to college, etc.
Could they use a sub? Yes! ALL bookshelfs do. No driver producing your mids should have anything to do with deeper bass. Especially one which horn loads your tweeter!

Now, being a DIYer, the next step (of course) is removing them from the box and putting them on an OB, crossing above 300hz, reworking the T/M XO, applying a felt ring around the horn and woofer surround......etc,etc :) . Probably after the World Cup :D

cheers,

AJ


And as always, the pics

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
 
That's funny, I've got a pair of those sitting in my listening room right now. They were dropped off by a friend who asked "How much do these suck?" This was a guy who buys speakers based on dollars per pound. On that basis, I guess they do pretty well. If anyone's interested, I'll post some measurements.
 
With the advent and acceleration of Chinese manufacturing, many very competent designers of drivers and systems are able to make quantity purchases of quality products at dirt cheap prices. $5 drivers and crossovers and $20 cabinets for large quantity production runs of respectable quality were unheard of until recently.
There is speculation that this will not endure forever due to the historical precedence of industrialized societies producing a middle class that expects an ever increasing standard of living and quality of life.
 
That's funny, I've got a pair of those sitting in my listening room right now. They were dropped off by a friend who asked "How much do these suck?" This was a guy who buys speakers based on dollars per pound. On that basis, I guess they do pretty well. If anyone's interested, I'll post some measurements.

John, I actually think there would be plenty interest, given the price and (relative) availability. I would be interested in seeing a distortion sweep of the finished system (rather than individual drive units).
FR on and off axis. FR @ say 80, 90 and 100db, 1m to see the change in the tweeters response as woofer excursion level increases.
I think to wring the most out of these boxes (other than felt mods) would be to rework the xo and then use them with stand (sub)woofers that crossed no lower than 200hz, better yet closer to 300hz for higher output systems.
I'm not giving up my W22's, but I can't believe how decent these sound for 20 stinking $ ea.
The DA175 sounds better but is almost $20 by itself.
Keep us posted. Thanks.

cheers,

AJ
 
Interesting tip - thanks AJ. And I concur that seeing Zaphs measurements would be very interesting indeed - see whether there is enough inherent overlap in the drivers workable areas to merit some experimentation.
I've always been meaning to pull the drivers from my Tannoy system 600's and play around with some DSP xovers to see how phase manipulation alters the sound. I've never gotten around to it, but at $40/pr these would be ideal as long as the breakup of the CF cone is far enough away from the lower end of the tweeter's usable range.

Intriguing to be sure.
 
Not being near a Best Buy (I live in India) I dont think I can ever lay my hands on these. Wish I could though. :-(

There seems to be a lot going for these speakers
a) non parallel side walls
b) decent looking driver simialr to Tannor or KEF "fullrange 2way" drivers
c) Solid looking cabinet with adequate internal damping.

wish the driver did not have a stamped frame though.
 
Member
Joined 2005
Paid Member
Hi AJ,

This doesn't surprise me at all.

The most significant cost in the production of the loudspeaker parts and completed speaker has been the labour. But now you why so much manufacturing has been moved to China.

When we individuals buy DIY parts from the main distributors for building our toys, remember that we're paying FULL RETAIL too. Even though they may come in brown cardboxes and shipped from a warehouse, whenever we buy single units, or a low number of units (generally <4 of the same unit) at retail level (ie. overheads such as end-user customer support, returns etc) we pay 100-500% of the unit's cost (generally 100+) price.

Anyway, it's a bit hard to explain unless you work in import/distribution/wholesale/retail, but the fact that we can pay retail, and build out speakers that are competitive with other commercial offerings that cost 3-5 times as much only means 1 thing- that the pricing of luxurious items like hi-fi are purely based on "What the market will stand"
It is completely arbitrary, and only set to maintain profitabiltiy.

To be honest- the market can have more speakers like this out there, and less of the Bo$e rubbish that sticks 2 dollar drivers into 50 cent plastic cubes with a cap and 5 cents of wadding and sells to the end consumer for $200 ea.
 
Well said Navin the speaker industry needs more cheap speakers like this. Imagine if mini-systems and "ghetto" blasters had these speakers instead of the horrible speakers that currently come with them.

For $50 in DIY parts you can't come close, or atleast its performance would be more or less same if you were lucky.

But offcourse half the fun is building it and the pleasure from the moment you flick the switch and everything is just making music.
 
Insignia NS-B2111 test results and modifications.

Frequency Response - Combined Anechoic
This is merged from several plots to get an accurate representation of the full range. Basically, it's far field merged with a nearfield plus diffraction sim, then summed with the port response. Pain in the *** to do but generally pretty accurate. The response is similar to many cheap Asian drivers, with a cliff in the midrange. For cheaper drivers, less R&D is invested to find cone geometry and materials that give smooth response through the upper midrange. I would call the stock response curve "not very good". The hole in the lower treble is clearly audible as missing music information, just as the peak at 800hz makes female vocals edgy. Bass tuning is on the warm side, as this is a higher Qts driver that would probably be better used in a sealed enclosure than vented. Still, this is a $20 speaker, and for that it's not bad at all. I've seen worse at a few times the price. Sensitivity is relatively low for this system.

Frequency Response - @ 70, 80, 90 and 100dB
The is the frequency response from an MLS impulse at multiple levels. Things don't really change until around 100dB, when there's a few minor differences here and there.

Frequency Response - Off axis: 0, 15, 30, 45, 60 deg
These off axis plots are horizontal, not that it matters much with a coaxial. I would declare the crossover point to be higher than optimum for the woofer, considering a broad null begins to show up centered around 3kHz where the woofer's power response rolls off. Of course, crossing the tweeter much lower would probably result in gross distortion, so I'd generally take poor power response over a distorted lower treble.

Harmonic Distortion - Sample 1
Harmonic Distortion - Sample 2
Harmonic Distortion - Enclosure Rattle Problem
These harmonic distortion plots are taken under the same conditions as my 5.5" driver test group although in this case it's a system plot including the crossover and the tweeter. It looks relatively decent for a driver in this price range. The flat 3rd order through the midbass makes this speaker sound a little thicker and muddier than many better designs, but it's not too bad. Note the drop in HD as the tuning frequency nears. Obviously, this speaker's real isssues are in it's frequency response. A note about the enclosure rattle problem - I thought this was interesting so I posted it. I had taken the binding posts off and set them on top of the enclosure. They seriously rattled at 800 hz and this image shows what that looks like on the HD plot. It appears this enclosure's weak spot is the top and bottom.

Cumulative Spectrum Decay
Nothing unexpected here based on the frequency response, but I grabbed one anyway because I like all the pretty colors. :)

I've got an add-on filter design I will be posting later. It's just a couple of notches. I was going to post it now, but I saw a typo in the image.

Additional stuff that could be done
There could be some stiffening added to the top and the bottom. Better damping could be used also, and given the thin walls, preferably something with a mass loaded layer such as whispermat or sonic barrier. I could toss the driver, peel the plastic veneer off, add a 1/4" layer of MDF all the way around, veneer it and use the enclosure as a great base for something else. The entire crossover could be yanked and redone rather than the add-on notches I have shown. I'll leave these modification ideas to other people.
 
Insignia NS-B2111 add-on dual notch filter

Schematic, transfer function, frequency response and impedance
This is just a set of notches I threw together. This can be hooked up externally in series with the positive terminal, or it can be mounted inside on the bottom, spliced into the positive wire.

In general, the notches favor dips over peaks because dips are less noticable, and power response does play a factor. A mild rise in the 3kHz area and a peak at 15kHz flatten out slightly off axis. This puts the design axis around 5 degrees or so.

In case it's not obvious, tube amps should stay away from this due to it's wild swinging impedance.

Maybe in the future I'll get individual response curves and individual distortion plots so I can completely rework the crossover, but for now this set of notches might help someone who doesn't want to even open the box up. There was some talk on another forum about using this speaker as the system for a DIY crossover design contest. That might be fun.

This box was sealed to help reduce a little boom due to the tuning of the port. This comes at the expense of a little power handling and extension in the 60hz area. The system is best used with a sub anyway. Oh yeah, black socks work best for stuffing the port - you can push them in a ways so you can't see them. If you need your socks, you might want to just glue a disc onto the back of the port.

If anyone else has measurements for this system, please post them. Obviously, with something this cheap, consistency is an issue. The more measurements we compare, the better we can design a filter that works best for most of the production variations of this model.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.