So I was browsing the PE buyout page, as I often do, and I came across these ultra-cheap TV full-range speakers:
EAB35995502 1-1/8" x 2-3/4" Paper Cone Full-Range Speaker 8 Ohm
As cheap as these are, it seems like it would be worth trying even if it didn't work out. Might go ahead and order a whole bunch of them, since you can get 50 of these suckers for $36 including PE's flat rate shipping.
I'm thinking doing these as OB or U-frame to make the whole project as cheap and easy as possible.
What do you guys think?
EAB35995502 1-1/8" x 2-3/4" Paper Cone Full-Range Speaker 8 Ohm
As cheap as these are, it seems like it would be worth trying even if it didn't work out. Might go ahead and order a whole bunch of them, since you can get 50 of these suckers for $36 including PE's flat rate shipping.
I'm thinking doing these as OB or U-frame to make the whole project as cheap and easy as possible.
What do you guys think?
Have fun, just remember it's going to take you a lot longer than you hope. Lots and lots of wiring to do. I was tempted to make a set of ob line arrays with something from this buyout but decided my time was better spent elsewhere. (Too many almost done projects in my plate)
It's going to make a very narrow array, mind you.
It's going to make a very narrow array, mind you.
Narrow in terms of dispersion or physical size? Do you think the dispersion characteristics would be different if they were arranged vertically vs. horizontally? I would think horizontal would be better, as it would allow for closer CTC placement.
Oh, I meant that with 25 a side you're looking at a surface area of roughly 27*2.75 inches. That's on the smaller side of a line array. Should have pretty decent horizontal dispersion across a 2.75" array width.
*Take all this in with the knowledge I've never built a line array.
*Take all this in with the knowledge I've never built a line array.
Well, with these drivers being as cheap as they are adding more wouldn't really be a big deal. I also haven't ever built a line array, but I don't have even a working theoretical knowledge of how they really work. But, I'm willing to give just about any project I can do for less than $50 a go.
Depends on how you want to use them, too. Ultimate near field computer speakers would be amazing potent with a 27" height and you could drive them from a headphone amplifier. 🙂 That's including a liberal amount of bass boost. Whether the drivers inherent distortion/frequency response even at extremely low drive level is good enough is another question however.
It's a great way to experiment and learn! Don't get me wrong, I'm encouraging you to try, just the amount of labor involved is quite high.
It's a great way to experiment and learn! Don't get me wrong, I'm encouraging you to try, just the amount of labor involved is quite high.
Don’t worry, you didn’t come across as discouraging. As far as labor goes, I’m not too worried about it. The way I think about, the longer a speaker project takes, the less money I spend starting new ones!
Remember that you need to get them in quantities of squares (4,9,16,25,36,49). This is so that you can wire them series parallel to maintain impedance that an amp likes.
A 49 driver array would be pretty amazing sensitivity. 1 watt would be 100dB+. Get good wire crimpers and lots of the correct Faston crimp terminals and at least 18ga to 16ga wire as the signal has to pass through the series part of it low losses. But on second thought, if xheapnis what you are after, soldering directly to tabs is least expensive. A hassle to remove though.
God luck.
If you got 16 of the TC6’s at $8 ea they would sound superior in every way. Not a super tall line array but enough to sound pretty good.
Peerless by Tymphany TC6FD00-04 2" Full Range Paper Cone Woofer 4 Ohm
A 49 driver array would be pretty amazing sensitivity. 1 watt would be 100dB+. Get good wire crimpers and lots of the correct Faston crimp terminals and at least 18ga to 16ga wire as the signal has to pass through the series part of it low losses. But on second thought, if xheapnis what you are after, soldering directly to tabs is least expensive. A hassle to remove though.
God luck.
If you got 16 of the TC6’s at $8 ea they would sound superior in every way. Not a super tall line array but enough to sound pretty good.
Peerless by Tymphany TC6FD00-04 2" Full Range Paper Cone Woofer 4 Ohm
Last edited:
Hmmm... 49 drivers per channel, that would really by something. As far as wire gauge goes, I do almost all of my projects with 16ga just because that's my go-to, so shouldn't have any problems with that. Heck, might order 150 of these things just to have extras, The form factor makes me think they'd work for octagonal omnidirectional speaker, too.
Trying to wrap my head around how to wire a line array... Is it important to make sure that every individual driver is contributing equally to the impedance? For example, should I make sure that I don't create 8ohms by both wiring 4P-2S and 2P? I feel like the 16ohm impedance gives a bit more leeway in this department, but I want to make sure I understand the way the signal needs to flow. My understanding is that more impedance for a woofer equates to less power and thus less sensitivity to the woofer, but I could be wrong.
Oh, absolutely it would sound better. But, 150 of the cheap TV speakers would cost $91 shipped, where 16 of the TC6's would cost $127 shipped, and would only be enough for one channel. The idea of this is more of an exploration of what can be done with these ultra budget drivers than a quest for high fidelity. Plus, it should be just plain fun either way. Who knows, may never be more than a mental excercise, but for under a hundred bucks, why not buy a gross of full range drivers?
Think in terms of blocks: let's say a 36-driver array/side.
Wire blocks of 6 drivers in parallel down the array, now you have 6 nominally 1.25 ohm sub-arrays (not subwoofer arrays, mind you 😀). Now wire these blocks in series and you end up back at your original impedance of 8 ohms. You also gain ~6x the power handling ability of one driver as result of this scheme.
Pretty sure you want to parallel then series to minimize driver-to-driver variations, but I might be doing the math in my head wrong.
The overall impedance of the driver doesn't necessarily change it's efficiency, which is a power in to power out relationship. If we look at it from the basics of V=IR; and P = V^2/R, then for 16 ohms vs 8 ohm wiring, we need sqrt(2) times the voltage, albeit at sqrt(2) less current. So our voltage sensitivity has dropped, but the power remains the same.
Given the efficiency of the system will increase (roughly, less in reality due to interference/other effects) by 3 dB per doubling of drivers, having 36 drivers a side is just over 15 dB more efficient than one driver. 78 db nominal jumps up to closer to 93 dB. This gives you a bit of room to eq the bottom before it falls apart, but surface area is king for LF performance because these guys have laughably small xmax and stressing with displacement will make them fall apart really_fast.
Thanks DPH, that was more or less my understanding of the system. The number I came up with, though, was 32 drivers per side, with four sub-groups (not subwoofer groups) in series each with eight drivers in parallel, for a nominal impedance of 8ohm from the 16ohm drivers. Should give better voltage sensitivity, as you said, with increased current. I'll be running these off of solid state amps exclusively, and for my amps at least I've found that they really do best with 8ohm nominal loads.
Sounds like a plan! I wanted to be thorough with an explanation of wiring, in case any of this was uncharted territory for you.
I really appreciate that. I've done plenty of electronics before, along with crossovers etc, but I've never done a line array. As far as LF goes, though, I'm planning on supporting with some sort of small sub up to about 200 Hz, so it shouldn't be a problem. I've never even auditioned a line array before, so I'm mostly just curious what it's going to sound like.
If you decide to go for an infinite array, the theory presented in this thread is a very good start: Infinite Line Source: analysis
It also explains why you'd have to factor in the use of EQ to shape the output. Of course there are other options, such as a finite array using shading etc. Figure out what way you want to go so the theory matches your attempt. Just remember, whatever driver you choose, you'll get it back on steroids within an array.
Well, I just ordered 100 of these, so we'll see how things turn out... Probably gonna do the baffles out of foam core, keep XRK interested...
I'd go for series wiring, and then parallel the groups.
The reason is that any connections or bits of wire will have a non-trivial impedance compared to the 1.25ohm groups, whereas if you series a bunch first, you've got a higher impedance group where small resistances in wire and connections won't make a difference.
Chris
Thanks for that advice, Chris. I'm planning on doing slide on connectors, too, so you're right that it would definitely add a statistically significant resistance to small loads. I'll try to redraw my diagram later to try to get somewhere in the ballpark of 6-8ohms nominal, let you guys know what I decide on. I also realized yesterday that I was thinking these drivers were 16oh, but they're 8ohm. The ones next to them in the list are the ones that are 16ohm...
- Status
- Not open for further replies.