|
|
Post by svart on Jul 12, 2016 10:02:48 GMT -6
So this has been on my mind for a while, since before I even started (and completed) the two-way monitors.. But now I have a much better understanding of how the modeling software works, and how it equates to the physical result.. So much so that I'm pretty confident that I can tackle a 3 way design now, which is NOT as simple as adding another driver to the mix.
So I have a few options to consider..
Price. I can attempt a "best for the money" type of design, which aims to use the most cost-effective drivers.
Price/performance. I can mix the drivers for best options of performance in certain places, like getting a better midrange for critical imaging, etc.
There is also the tradeoff of distortion/linearity and euphonics. In my two-way monitor I went with super low distortion drivers, but they are a bit clinical to listen to, although they aren't fatiguing at all. In this design I have the chance to re-use some of what I learned about aluminum coned drivers, or move back to paper cones. Yet again, a non-trivial design choice to make.
Then comes the driver selection..
Dome tweeter, ring radiator, etc. I'm leaning towards a traditional dome with better off-axis dispersion on this one, although the better the off-axis, the more they cost.
Midrange.. I'm trying to figure out what to do here. I've always wanted to use a dome midrange, but reading about dispersion problems, and the lack of choices kinda makes me want to go the route of a standard cone. Then, the choice becomes, do I go with a smaller cone, or possibly a larger one?
I've been toying with the idea of using a larger midrange, like 5" so that the excursion is minimal for the low cut-off point. That way distortion is lessened, at the expense of possibly lower sensitivity at the high end rolloff.
For the woofer, I had planned on trying one of the 8" offerings from Dayton, but now I'm considering a 10" speaker, although there is no option for 4 ohm 10" speaker, which leaves me with either going with another choice, or doing some fancy impedance matching.
Another issue is that I found the aluminum tweeter to have a drastic dip in output beyond 10K. I can tilt the response, but it dips efficiency down into the low 80db range, which is bordering on too low for some amps. For this I might need to stick with a silk dome for better efficiency at the cost of some possible distortion. As long as I keep the cutoff frequency above 2x the Fs of the driver, I should be OK.
So right now it's looking like:
Dayton drivers, RS28F-4, RS125-4, RS225-4.
|
|
|
|
Post by kcatthedog on Jul 12, 2016 11:15:15 GMT -6
Ported ? Front or back ?
|
|
|
|
Post by svart on Jul 12, 2016 12:18:45 GMT -6
Ported. Back most likely. I don't particularly care for front ports. It's hard to control the resonances and the front-of-speaker mixing of the port and the woofer, and there's the port chuffing that happens.
|
|
|
|
Post by NoFilterChuck on Jul 13, 2016 9:18:05 GMT -6
Lol turn down the volume if you hear air moving in the ports
|
|
|
|
Post by svart on Jul 13, 2016 9:41:33 GMT -6
Lol turn down the volume if you hear air moving in the ports Nope, port chuffing is a design flaw. Either your ports are designed incorrectly for the air flow, or the port is too small and the box size/air volume needs to be changed. I generally use flared ports, flared on both sides, and try to use the largest port I can for the box size, adjusting the box dimensions to work with the large ports. Typically what happens is that commercial designers *stop* designing once their primary criteria are met, namely performance at a price point. They rarely go the extra mile to tune the remaining parts of the system, unless the price point is very high. This is why you never hear of these problems on high-end monitors and speakers, but you hear them all the time on cheap/value/budget systems. |
|
|
|
Post by NoFilterChuck on Jul 13, 2016 11:41:41 GMT -6
Do you really hear port chuffing when you're listening at volumes lower than speaking level?
|
|
|
|
Post by svart on Jul 13, 2016 11:58:17 GMT -6
Do you really hear port chuffing when you're listening at volumes lower than speaking level? If the ports are bad enough. Listen to any KRK monitor long enough and you'll start to hear it.. Besides, assuming that your customer only listens at one level is asking for problems. If you design something, you have to design it for all listening levels, not just those you assume people should be using. |
|
|
|
Post by kcatthedog on Jul 13, 2016 15:03:55 GMT -6
The original adam a-7 had chuffing problems and as svart points out they then re-desinged it, at some cost to them and got it right.
|
|
|
|
Post by svart on Jul 13, 2016 15:22:25 GMT -6
So I messed with this over the last few days, and I have something I think I want to investigate further. I sim'd a large number of tweeters and midranges, and settled on the silk version of the dayton tweeter I used in the other speakers, the RS28F. The midranges were a LOT harder to find a good candidate. The smaller ones seemed to have worse break-up at the high end, right where I wanted to put the upper crossover. Some of the paper ones seemed to have other dips or peaks right in the middle of the midrange. The metal coned ones had crazy peaks in the breakup region, and would need a lot of crossover work to get under control so that they didn't bleed through the tweeter region. I also wanted to move to a larger midrange, since I believe that the speaker moving less at lower frequencies, means better midrange control overall, with less beaming at higher frequencies due to the larger cone. Think of it as using the cone as a sort of waveguide on it's own. With a LOT of work, trying different speakers, crossover setups and such, I finally worked out a crossover that uses 2nd order crossovers with minimal *fixing* of the driver peaking. Only the tweeter really needed some help, a *tilt* filter and a small band-stop filter to fix a peak and flatten the frequency on the top end. I had to do something similar on the metal-coned monitors as well, but I think this one might be a little more elegant. The midrange I ended up working through was the RS125-8, which is the 5" metal cone midwoofer.  I don't have box sims yet but I suspect it'll be around 1.5cu ft. IF this is the design I decide to go forward with, the drivers and crossover components come to roughly 490$. Add another 50-100$ for *good* wood and supplies and these would sit around 600$. |
|
|
|
Post by kcatthedog on Jul 13, 2016 15:48:20 GMT -6
So on paper you are within 4-5 db from about 50 to plus 10k cycles: impressive starting point !
|
|
|
|
Post by svart on Jul 13, 2016 18:17:34 GMT -6
So on paper you are within 4-5 db from about 50 to plus 10k cycles: impressive starting point ! I could make it ruler flat, but I actually prefer a little tilt as it is in the picture. This helps keep it from being too piercing and helps keep the audio in perspective without a sub. It's only a couple db tilt. The pic makes it seem more drastic than it is. |
|
|
|
Post by kcatthedog on Jul 13, 2016 18:21:45 GMT -6
Understood, plus you have your previous experience to know how the lines on paper translate to real world sound:go with your gut !
|
|
|
|
Post by svart on Jul 13, 2016 19:16:29 GMT -6
Understood, plus you have your previous experience to know how the lines on paper translate to real world sound:go with your gut ! I'm actually surprised that nobody really seemed interested in the speakers I built last time. I really feel like they are worth the build, and definitely surpassed my expectations, and I believe with the last tweak, surpass the Amphions I meant for them to compete with. |
|
ericn
Temp
Balance Engineer
Posts: 16,107 
Member is Online
|
Post by ericn on Jul 13, 2016 20:18:04 GMT -6
Try the Skanspeak ( Quested and PMC lower end) or the Tang-Band
dynaudio copy ( real Dyn in Dyn and K&H ) Tang Band better QC than Dynaudio! either mates well with Morrel Cat 308!
Woofer to taste!
|
|
|
|
Post by kcatthedog on Jul 14, 2016 6:10:45 GMT -6
Fair enuff: try b4 you buy ?
In my case, the fact the I could try the amphions in my home (two sets) at very low expense: (just ship back to Montreal) was a big factor for me. I wasn't going to buy monitors I had not auditioned. No criticism of you, but I think we all hear differently, so a monitor one person might love another might not. |
|
|
|
Post by svart on Jul 14, 2016 7:22:45 GMT -6
Fair enuff: try b4 you buy ?
In my case, the fact the I could try the amphions in my home (two sets) at very low expense: (just ship back to Montreal) was a big factor for me. I wasn't going to buy monitors I had not auditioned. No criticism of you, but I think we all hear differently, so a monitor one person might love another might not. That's why I took them to John's place and had multiple folks take a listen comparing to amphions. The general consensus was that they were close, so I took the thoughts/critiques home and applied them to the design. Now I'm sure they slay the Amphions! |
|
|
|
Post by mdmitch2 on Jul 14, 2016 8:18:05 GMT -6
Fair enuff: try b4 you buy ?
In my case, the fact the I could try the amphions in my home (two sets) at very low expense: (just ship back to Montreal) was a big factor for me. I wasn't going to buy monitors I had not auditioned. No criticism of you, but I think we all hear differently, so a monitor one person might love another might not. That's why I took them to John's place and had multiple folks take a listen comparing to amphions. The general consensus was that they were close, so I took the thoughts/critiques home and applied them to the design. Now I'm sure they slay the Amphions! I must have missed that whole thread. I'd love to hear your speakers against my psi a21's. |
|
|
|
Post by svart on Jul 14, 2016 8:28:29 GMT -6
That's why I took them to John's place and had multiple folks take a listen comparing to amphions. The general consensus was that they were close, so I took the thoughts/critiques home and applied them to the design. Now I'm sure they slay the Amphions! I must have missed that whole thread. I'd love to hear your speakers against my psi a21's. |
|
|
|
Post by svart on Jul 14, 2016 13:49:29 GMT -6
Did some cost reducing and further streamlining of the crossover. I was able to work it so that I could use smaller(cheaper) inductors and caps of the same values for bulk purchase, etc.
I got the costs down to around 420$, and with some searching I found some surplus places for some of the crossover components, hopefully to get the cost down even more.
If someone were to build these, there are a few caps that could be sub'd with electrolytics to really drop the price, but I'm not going that route.
Still not sure if I'm ready to pull the trigger on purchasing the parts yet or not. I might let this sit for a week or so to think about the crossovers some more.
|
|
|
|
Post by kcatthedog on Jul 15, 2016 6:28:59 GMT -6
No, only half  |
|
|
|
Post by svart on Jul 15, 2016 9:17:56 GMT -6
So here's the next iteration of the crossover. You have to be very careful to use inductors that have the DCR listed under the inductance value, or you'll be way off. All of the inductors are air-core 18 or 20 gauge crossover coils from Parts Express.  3way.pdf (21 KB) |
|
|
|
Post by guitfiddler on Jul 15, 2016 19:48:32 GMT -6
No, only half She said half, I said stereo dammit! Looking nice, I have a buddy(carpenter) that wants to build a set of monitors, these would be nice for the studio. |
|
|
|
Post by svart on Jul 17, 2016 14:00:33 GMT -6
No, only half She said half, I said stereo dammit! Looking nice, I have a buddy(carpenter) that wants to build a set of monitors, these would be nice for the studio. These aren't built yet.. So I don't know how it'll turn out in real life. I'm still deciding whether I want to put the cash down to build a set of these.. I do have a set of the metal coned monitors I might part with if you're interested. |
|
|
|
Post by guitfiddler on Jul 17, 2016 21:36:14 GMT -6
I'm actually more interested in the building of the newer 3-way design, but thanks for asking. I already have some projects on hold because of the not enough money to go around problem.
|
|
|
|
Post by svart on Aug 4, 2016 21:41:16 GMT -6
I think maybe next month I might bite the bullet and build these, unless I can move some stuff I've been trying to sell.
|
|
