Without ever googling the TB W3-881SI, I bought 8 of them with a weird fantasy line array idea in my head. And a couple of buyout Nuance TW5-073LR tweeters in case there wasn't enough high end for my poor old decrepit ears. Then I found this thread...
Since I have lots of plywood in the garage and 8 speakers, I could make a pair (or two) to try tomorrow after work.
One question for Skeeter99: did you put two drives in the same size enclosure that cogitech did to start this thread? I like the two driver idea because it might extrapolate in to the 4 driver idea that I originally had.
I love this forum...
Since I have lots of plywood in the garage and 8 speakers, I could make a pair (or two) to try tomorrow after work.
One question for Skeeter99: did you put two drives in the same size enclosure that cogitech did to start this thread? I like the two driver idea because it might extrapolate in to the 4 driver idea that I originally had.
I love this forum...
I built some small closed box cabinets that would fit inside some 1925 cathedral speaker cabinets that I obtained, and after much research I used the $10 Peerless/Vifa TC9FD-18-08 paper cone 3 inch driver (Madisound or Parts Express has them). Not much bass, but amazing sound above 100HZ. Since I have tone controls, I've gotten back enough bass to be happy.
In a tower like you built, I would seriously consider using these (Xmax = 2.5mm) and instead of a port I might try a passive radiator near the bottom of the cabinet (near the floor). I have no idea of the proper way to design with Passive radiators, and I didn't want to like them since you'd think they would ring, thereby muddying up the bass, but when I used a regular woofer with a resistor across the coil, acting as a passive radiator (no feed from the amplifier), it sounded VERY good. The cabinet was able to stay small, and the bass was fantastic sounding. Since the mechanical damping in a ported design is very poor on either side of the tuned frequency, the ported approach may be more likely to have a ringing problem than even a regular passive radiator with out the electronic damping. The resistor (1-8 ohms) damps the "passive radiator" driver pretty well. Varying the resistor size may actually tune its function.
The thing about ultra cheap "buy-out" drivers is that they are often not very good. Most that I've seen have the foam suspension which deteriorates badly in about 10 years, or a meager Xmax, no shorting ring on the pole piece, or etc. For $10-$22 you can get very good drivers from Peerless and others. When you look at the total cost of the project and what your time may be worth, the extra $10 is a pretty small percentage. You'll be stuck with whatever you build for years maybe.
Roger Russel, former head of acoustic research at McIntosh ('68 - '92) is using the 3 inch Peerless TG9FD10-08 glass fiber cone version of the one I mentioned above. The cone material appears to be the only difference. At $22 each, they have a VERY good looking published frequency response graph, well ventilated spider, shorting ring on the pole piece (doesn't actually have to be copper to work - I think they use aluminum), and are arguably the best wideband driver out there for the 100HZ - 15kHZ range. The Dayton drivers with the extra large Xmax have such bulky surround suspension that they have "cavity effect" and it shows on their frequency response graph. The Peerless drivers do way better on that and off-axis response.
When it comes to measurements (pink noise and calibrated mic or equiv.), know that the room acoustics will usually cause +/- 10dB variations by the time you are in the listening position. So don't worry about 2 and 3 dB variations too much.
Roger Russel has a real fun website. He's retired now but still active as a speaker builder. Google his name and you'll find it. He loves the vertical line array approach; 25 three inch Peerless drivers per side run wideband, with active EQ (using opamps) ahead of the poweramp to get bass relatively flat down to below 30HZ. They go floor to ceiling, and thereby have a very intriguing way of interacting with room acoustics. He sells them for about $15,000 a pair with the active EQ box. Those $22 drivers are good enough for him and that...
In a tower like you built, I would seriously consider using these (Xmax = 2.5mm) and instead of a port I might try a passive radiator near the bottom of the cabinet (near the floor). I have no idea of the proper way to design with Passive radiators, and I didn't want to like them since you'd think they would ring, thereby muddying up the bass, but when I used a regular woofer with a resistor across the coil, acting as a passive radiator (no feed from the amplifier), it sounded VERY good. The cabinet was able to stay small, and the bass was fantastic sounding. Since the mechanical damping in a ported design is very poor on either side of the tuned frequency, the ported approach may be more likely to have a ringing problem than even a regular passive radiator with out the electronic damping. The resistor (1-8 ohms) damps the "passive radiator" driver pretty well. Varying the resistor size may actually tune its function.
The thing about ultra cheap "buy-out" drivers is that they are often not very good. Most that I've seen have the foam suspension which deteriorates badly in about 10 years, or a meager Xmax, no shorting ring on the pole piece, or etc. For $10-$22 you can get very good drivers from Peerless and others. When you look at the total cost of the project and what your time may be worth, the extra $10 is a pretty small percentage. You'll be stuck with whatever you build for years maybe.
Roger Russel, former head of acoustic research at McIntosh ('68 - '92) is using the 3 inch Peerless TG9FD10-08 glass fiber cone version of the one I mentioned above. The cone material appears to be the only difference. At $22 each, they have a VERY good looking published frequency response graph, well ventilated spider, shorting ring on the pole piece (doesn't actually have to be copper to work - I think they use aluminum), and are arguably the best wideband driver out there for the 100HZ - 15kHZ range. The Dayton drivers with the extra large Xmax have such bulky surround suspension that they have "cavity effect" and it shows on their frequency response graph. The Peerless drivers do way better on that and off-axis response.
When it comes to measurements (pink noise and calibrated mic or equiv.), know that the room acoustics will usually cause +/- 10dB variations by the time you are in the listening position. So don't worry about 2 and 3 dB variations too much.
Roger Russel has a real fun website. He's retired now but still active as a speaker builder. Google his name and you'll find it. He loves the vertical line array approach; 25 three inch Peerless drivers per side run wideband, with active EQ (using opamps) ahead of the poweramp to get bass relatively flat down to below 30HZ. They go floor to ceiling, and thereby have a very intriguing way of interacting with room acoustics. He sells them for about $15,000 a pair with the active EQ box. Those $22 drivers are good enough for him and that...
The reasons I like these Peerless 3.5 inch drivers, besides the well ventilated spiders and vented pole piece, and of coarse the copper (aluminum) ring at the top of the pole piece, is that they have less cavity effect than the Dayton 3.5 inch drivers (for example) that have 2-3 times the Xmax. To get that huge Xmax, the surround suspension apparently needs to really stick up (which also cuts into the effective cone surface area), which on such a small cone creates significant cavity effect, as in frequency response anomolies (compare graphs). Plus the shallower suspension ring will stay out of the way of off axis treble response better. plus, if you had the Daytons doing excursions of 5mm pk, you might experience audible doppler effect modulation of any higher frequency content... So extra large Xmax isn't always necessarily ideal, given the tradeoffs. The Daytone is rated at 4 or 5 mm Xmax, and the Peerless is rated at 2.5.Cogitech,
So you like these $10 drivers? I have been thinking of trying them in the cornu.... hmmm. They look handy enough that one should buy them by the dozen
I've been living with the $12 black paper cone versions of these for a while now, and they amaze me at how good they sound run wideband driven by my computer and a 60 watt 3886 poweramp I built. The glass fiber cone versions ($22ea.) are in my stereo holographic soundbar cabinet. I'm still building the processor electronics and poweramps for that so I haven't really heard them yet. Those will only be asked to do 150HZ - 7kHZ, and I expect them to be quite good at that.
Cogitech,
So you like these $10 drivers? I have been thinking of trying them in the cornu.... hmmm. They look handy enough that one should buy them by the dozen
For cheap bedroom system I built, they are perfect. In Curnu I think the bass might surprise the crap out of you, and the high end is hot enough to balance, I would guess...
Bob,
You are preaching to the choir, the Vifa(Peerless) TC9FD is my favorite driver. It sounds amazing, and at $12, you can make lots of speakers
My favorite speaker cabinet for this Vifa is the Cornu spiral. Sounds like a much bigger driver and has great bass, mids, and fantastic highs.I used 4x of these drivers in a small sealed enclosure for a travel-sized karaoke box for a friend once. They sounded really good for the price and size of enclosure... I ended up order 4x more, just as a future project.
Fast forward nearly a year later, I saw them sitting on the shelf. Xmas was around the corner. Needed to do something with them... decided to build my Father some speakers for his small home. Needed something with a micro footprint, in a tower, that could give good sound. A bit of searching led me to these "nanoTowers". So... I build them! Well, almost the same... I used 9mm birch ply and did 5.25" x 5.25" x 45.5". Used the same port size/length of 1.5" diameter by 2.5" long. Used a bit of polyfill in each.
After hooking them up to the cheap T-amp Dayton audio makes and playing a variety of music through them... I can't say they are the best speakers I have ever heard... but they sure take the crown for the 3" and the $8/speaker category! The very top end wasn't all that brilliant, the midrange was very hypnotic, and the low end was... well, it was there! Which was pretty mind blowing. I played some Rap for a friend (his favorite genre), and he just shook his head. The low end was well placed in music!
I played Jazz, Electronic, Rap, Metal, 60's, 70's, Rock, Pop, Soundtrack... I gave them a good variety. As a metal-head, I can't say I could listen to these all day with some of the heavier stuff I like. Simply, didn't have the punch to keep up with the midrange heavy guitars and the top end of the cymbals weren't cutting through, though I'd still take these over any TV speaker for the same purpose! I found Jazz, Rock, Pop, and Rap to really shine on these, for whatever reason. I think Rap, simply because the dedicated low-end it offered just really showed the potential of these, where the midrange capabilites shined well onto the other few genres. Course, this is just my opinion on it.
I took some video of them doing their thing throughout those genre trials, but haven't had a chance to load it up. When I do, I'll do it as a response to existing nanoTower youtube vids.
So props to you, cogitech. Thanks for the inspiration! I'm sure my Dad will love them... sadly, it cost me ~$70+ to ship them (with amp, double boxed, etc), which means the shipping costs exceeded that of the materials being shipped... but I expected a nasty price tag on that end of things.
Fast forward nearly a year later, I saw them sitting on the shelf. Xmas was around the corner. Needed to do something with them... decided to build my Father some speakers for his small home. Needed something with a micro footprint, in a tower, that could give good sound. A bit of searching led me to these "nanoTowers". So... I build them! Well, almost the same... I used 9mm birch ply and did 5.25" x 5.25" x 45.5". Used the same port size/length of 1.5" diameter by 2.5" long. Used a bit of polyfill in each.
After hooking them up to the cheap T-amp Dayton audio makes and playing a variety of music through them... I can't say they are the best speakers I have ever heard... but they sure take the crown for the 3" and the $8/speaker category! The very top end wasn't all that brilliant, the midrange was very hypnotic, and the low end was... well, it was there! Which was pretty mind blowing. I played some Rap for a friend (his favorite genre), and he just shook his head. The low end was well placed in music!
I played Jazz, Electronic, Rap, Metal, 60's, 70's, Rock, Pop, Soundtrack... I gave them a good variety. As a metal-head, I can't say I could listen to these all day with some of the heavier stuff I like. Simply, didn't have the punch to keep up with the midrange heavy guitars and the top end of the cymbals weren't cutting through, though I'd still take these over any TV speaker for the same purpose! I found Jazz, Rock, Pop, and Rap to really shine on these, for whatever reason. I think Rap, simply because the dedicated low-end it offered just really showed the potential of these, where the midrange capabilites shined well onto the other few genres. Course, this is just my opinion on it.
I took some video of them doing their thing throughout those genre trials, but haven't had a chance to load it up. When I do, I'll do it as a response to existing nanoTower youtube vids.
So props to you, cogitech. Thanks for the inspiration! I'm sure my Dad will love them... sadly, it cost me ~$70+ to ship them (with amp, double boxed, etc), which means the shipping costs exceeded that of the materials being shipped... but I expected a nasty price tag on that end of things.
Attachments
Agreed. I don't think ANY DIY tower will beat it for the price, size, and/or simplicity! I re-read through the thread and you mentioned the imaging to be surprising, that's something I forgot to mention. Female vocals are usually my "go to" (such as Jazz) to see how centered the imaging is... these things passed the test like a champ! Definitely wide when needed... It's really hard not to over-judge these things, comparing them to higher-end speakers... they act like they should be classed a little higher than ~$8 each.
I'm actually planning out a 4x per tower build to see what they do in larger quantities. They got surprisingly loud off that amp (which is over-rated at 15w per channel at 4 ohm, so probably only ~6 watts each at 8 ohm, realistically), but with more of them comes more power handling (15x4 = 60 watts!). Not even sure what I'm going to do with a larger build, but should be interesting...
I'm actually planning out a 4x per tower build to see what they do in larger quantities. They got surprisingly loud off that amp (which is over-rated at 15w per channel at 4 ohm, so probably only ~6 watts each at 8 ohm, realistically), but with more of them comes more power handling (15x4 = 60 watts!). Not even sure what I'm going to do with a larger build, but should be interesting...
I've got mine in a FAST and I gotta agree, it's sparkly, but there is some breakup. Even though it measures flat, there is some nasty business up there. I can't believe how well it holds off axis though. WAY off axis I still have the entire top end.
And ya, I can't believe how good they are for the money. If they could make bass like a CHR70, I'd suggest these are better. But the CHR70 does make way more bass, which is worth the price tag difference or more imo.
Amazing feedback! I never would have imagined that my nanoTower would turn out to be such a success not just for me but for others. I really appreciate you taking the time to share your build story and listening notes.
Did you play with adjusting the damping at all? I found that optimal bass response was only achieved after adding far more damping than I had anticipated....
Looking forward to seeing your video response!
I'm putting together a budget frequency response analysis setup.
If it ends up working OK, I'll measure the in-room response of my nanoTowers and I suspect you will not believe your eyes. (which isn't to say they produce more bass than CHR70, of course...)
I am also putting together a budget measurement rig. Pano, the king of measurements suggested looking at the software called Holmimpulse. Check it out, it is perfect for what we are trying to do. http://www.diyaudio.com/forums/soft...lse-measuring-frequency-impulse-response.html
It's free and works very well. I'm using the mic from my zoom recorder now but it takes any analog mic.
The nanotowers look very cool and I am sure they sound great. They are definitely a mass loaded TL because a pure TL would have only A tuning freq of 77 Hz based on 44 inches. It is the constriction in the port that lowers the tuning and you can play with this or simply model it. In many ways, this resembles a TABAQ except longer and driver is at end. If you placed driver at 1/3 from closed end, it will sound even better.
Last edited:
I will check it out for sure. I have my eye on SynRTA, which is also free and seems very, very easy (important for me!
).http://www.libinst.com/SynRTA.htm
More on the Peerless/Vifa 3 inchers
I can't comment on using these in a nano tower (transmission line), but my concern would be trying to get bass out of an Xmax of 2.5mm. At low levels they might be quite nice.
I used the WinISD freeware beta program that I found on the web somewhere, to calculate the optimal internal cabinet volume for both the Peerless/Vifa TC9FD-18-08 (black paper cone $12ish) and the TG9FD 8 ohm ver (same driver but with glass fiber cone)(8 ohm ver. has slightly better top end response than 4 ohm ver.). I had to type in all the Q parameters from the published data sheets and save the driver info since this driver is not in the library list within the program. Bottom line; the glass fiber version ($22ish) doesn't need as much internal volume. When you shrink the volume, this software shows you the effect it has on the frequency response (which is why I love this software). Either driver gives you a slight rise in the 200HZ region when the box gets smaller. Up at a Q of about .9 there's only about a dB of rise on the TG, and 2-3dB rise with the paper cone. If it integrates well with a separate woofer, I doubt if a 2dB rise is a problem.
Another cool thing about this software is that you can tell it 2 of the dimensions of the box internal, and it instantly calculates the third. It starts with a recommendation with a ratio that may be best (no harmonic resonances doubling up?). But sometimes you need certain dimensions to be different for some reason (I did).
If you want to run these wideband, there won't be any bass below 100HZ in a closed box. I added 2 Peerless 6 1/2 inch woofers in my soundbar, but the same program said they'd be rolling off below about 90HZ with a Q of about .8, so I decided to go with Q=.9 (slightly overdamp) and use active electronic EQ ahead of the poweramp to get it flat to around 35HZ.
A ported box will extend the bass of the 3 inch drivers a bit (1/2 octave?), but low bass can be very nice. Used as a midrange driver from 150HZ to 7kHZ, either of these 3 inch drivers would be a home run IMO. Above 7kHZ I'm using some Dayton 3/4 inch domes that Zaph claims are ruler flat from 4kHZ to 20k+. I'm not sure that paying more for a 3 inch will get you a better driver.
Note; these drivers are difficult to front mount on a baffle that is more than about 1/4 inch thick, because of restricted air flow out the back. I chose to rear mount them on 1/2 inch MDF, and router the 3 inch holes with a 1/2 inch radius bit so they flare open on the front side (to minimize cavity effect). If I was doing a vertical line array (like Roger Russel did), I'd have a custom 1/4 inch brass mounting plate made (probably expensive) that would mount on the box, so rear radiation would be easy (low reactance).
I can't comment on using these in a nano tower (transmission line), but my concern would be trying to get bass out of an Xmax of 2.5mm. At low levels they might be quite nice.
I used the WinISD freeware beta program that I found on the web somewhere, to calculate the optimal internal cabinet volume for both the Peerless/Vifa TC9FD-18-08 (black paper cone $12ish) and the TG9FD 8 ohm ver (same driver but with glass fiber cone)(8 ohm ver. has slightly better top end response than 4 ohm ver.). I had to type in all the Q parameters from the published data sheets and save the driver info since this driver is not in the library list within the program. Bottom line; the glass fiber version ($22ish) doesn't need as much internal volume. When you shrink the volume, this software shows you the effect it has on the frequency response (which is why I love this software). Either driver gives you a slight rise in the 200HZ region when the box gets smaller. Up at a Q of about .9 there's only about a dB of rise on the TG, and 2-3dB rise with the paper cone. If it integrates well with a separate woofer, I doubt if a 2dB rise is a problem.
Another cool thing about this software is that you can tell it 2 of the dimensions of the box internal, and it instantly calculates the third. It starts with a recommendation with a ratio that may be best (no harmonic resonances doubling up?). But sometimes you need certain dimensions to be different for some reason (I did).
If you want to run these wideband, there won't be any bass below 100HZ in a closed box. I added 2 Peerless 6 1/2 inch woofers in my soundbar, but the same program said they'd be rolling off below about 90HZ with a Q of about .8, so I decided to go with Q=.9 (slightly overdamp) and use active electronic EQ ahead of the poweramp to get it flat to around 35HZ.
A ported box will extend the bass of the 3 inch drivers a bit (1/2 octave?), but low bass can be very nice. Used as a midrange driver from 150HZ to 7kHZ, either of these 3 inch drivers would be a home run IMO. Above 7kHZ I'm using some Dayton 3/4 inch domes that Zaph claims are ruler flat from 4kHZ to 20k+. I'm not sure that paying more for a 3 inch will get you a better driver.
Note; these drivers are difficult to front mount on a baffle that is more than about 1/4 inch thick, because of restricted air flow out the back. I chose to rear mount them on 1/2 inch MDF, and router the 3 inch holes with a 1/2 inch radius bit so they flare open on the front side (to minimize cavity effect). If I was doing a vertical line array (like Roger Russel did), I'd have a custom 1/4 inch brass mounting plate made (probably expensive) that would mount on the box, so rear radiation would be easy (low reactance).
Last edited:
I use HolmImpulse. Tried making the switch to the more advanced REW, but HolmImpulse is just so easy to use and so good and keep going back and only using REW when I need certain features it offers.
Cogitech, I can measure 20hz out of my EL70. But the real question is SPL. When it comes to bass and measurements, they need to be couched with the applications. Room measurements? For EQing? To determine F3, F10? Etc. This is a 3" driver with a small Sd and small xmax. It only has so much capability. Just like the EL70 can whoop this TB in the Bass department, but the only time I ask it to do 20hz is in the nearfield at easy listening levels.
I nice way to measure real bass performance, is this:
Measure nearfield (2inch).
Set the software to -24db and increase your preamp volume until you have 2.83V at your speaker terminals.
Measure in room at -24db
Measure in room at -21db
-18db
-15db
etc.
Do this until the frequency response overlays start to bunch together or you hear audible distortion. My guess is you'll have to use the high pass feature in HolmImpulse right around your tuning. Watch you don't pop the driver either.
BTW, I'm not knocking this driver. I'm using it and liking it. I'm just realistic about it's performance. It's $9.99 and I'd prefer it to a CHR70 in a FAST. That's what I was saying. That's impressive. Full Range, ya, I'll take the CHR70 without a second thought.
EDIT - You'll need a decent condensor mic to measure reliably below 100hz. If just starting out and on a budget, google Panasonic WM61A. I have a couple and they're pretty decent. Also check out Cross Spectrum Labs for a calibrated Behringer ECM8000. I have the premium + or what ever it is. It's a really nice mic and feels good knowing it's been calibrated from 5hz to 25khz. (I don't actually use that range, cause I'd want a better mic to reliably measure those extremes).
Cogitech, I can measure 20hz out of my EL70. But the real question is SPL. When it comes to bass and measurements, they need to be couched with the applications. Room measurements? For EQing? To determine F3, F10? Etc. This is a 3" driver with a small Sd and small xmax. It only has so much capability. Just like the EL70 can whoop this TB in the Bass department, but the only time I ask it to do 20hz is in the nearfield at easy listening levels.
I nice way to measure real bass performance, is this:
Measure nearfield (2inch).
Set the software to -24db and increase your preamp volume until you have 2.83V at your speaker terminals.
Measure in room at -24db
Measure in room at -21db
-18db
-15db
etc.
Do this until the frequency response overlays start to bunch together or you hear audible distortion. My guess is you'll have to use the high pass feature in HolmImpulse right around your tuning. Watch you don't pop the driver either.
BTW, I'm not knocking this driver. I'm using it and liking it. I'm just realistic about it's performance. It's $9.99 and I'd prefer it to a CHR70 in a FAST. That's what I was saying. That's impressive. Full Range, ya, I'll take the CHR70 without a second thought.
EDIT - You'll need a decent condensor mic to measure reliably below 100hz. If just starting out and on a budget, google Panasonic WM61A. I have a couple and they're pretty decent. Also check out Cross Spectrum Labs for a calibrated Behringer ECM8000. I have the premium + or what ever it is. It's a really nice mic and feels good knowing it's been calibrated from 5hz to 25khz. (I don't actually use that range, cause I'd want a better mic to reliably measure those extremes).
Last edited:
The real question is SPL for some people, indeed. The real question for me, is overall frequency response at the intended/desired SPL. To test this, I plan to simply measure from the listening position. Doing a realtime frequency analysis of pink noise from 2" away from the driver would skew the results tremendously and give no sense as to what the speakers "sound like". Just my opinion, and I could be totally off-base, but I prefer to keep tests as "real world" as possible. If this makes them less scientific, then I am willing to accept the resulting criticism and/or cynicism.
EDIT: Thanks for the tip on the mics! (but what do I do with a capsule? - other than the obvious, uncomfortable option?)
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.