I'm going to post this here. This is my amplifier that I will hopefully put into shops as a new product. I'm proud of this. I might put these into the amplifier thread where I talk about the actual amplifier. I'm posting this here because it's mostly a showcase of the woodworking (which is relevant to speaker-building 
).
).Attachments
I took my Celestion apart. The enclosure is about 16.5L. The port is about 7cm diameter by 14cm long. Crossover frequency is 2.5 kHz. I have attached photos.
Attachments
Thanks.
I've decided to go with something like the Gibbon 88 clone, but I'm not sure of a few things.
If I use a 100 cm high speaker which has a top chamber of 55x32x19 sealed off, giving a front baffle of 100x22 and a depth of 55cm including wood width, I get 32 L. Is that right?
Then, if I use the Gibbon 88 layout, the simulated frequency response isn't how yours comes out, Steve. From about 150 Hz to 1 kHz, the FR is down about 2 dB. If I put the tweeter at the top and the woofer below (like the Celestion F20, but quite a way down), the frequency response is far flatter.
I have attached my project file if anyone is keen to look (rename to BPJ).
I've decided to go with something like the Gibbon 88 clone, but I'm not sure of a few things.
If I use a 100 cm high speaker which has a top chamber of 55x32x19 sealed off, giving a front baffle of 100x22 and a depth of 55cm including wood width, I get 32 L. Is that right?
Then, if I use the Gibbon 88 layout, the simulated frequency response isn't how yours comes out, Steve. From about 150 Hz to 1 kHz, the FR is down about 2 dB. If I put the tweeter at the top and the woofer below (like the Celestion F20, but quite a way down), the frequency response is far flatter.
I have attached my project file if anyone is keen to look (rename to BPJ).
I don't actually think on-axis frequency response is the overall guiding factor in how a speaker sounds:
http://www.diyaudio.com/forums/multi-way/305201-visaton-2-way-tower-monitor-audio-bronze-5-a-6.html#post5023831
Any mug can use target curves like LR2, BW3 and LR4 to design a speaker. These days I think that what you tend to hear is low distortion, a flat power response and an easy to drive impedance that your amp will like.
TBH, I have seen so many badly designed speakers that claim to be a perfect 24dB/octave LR4 or some such, and therefore perfect. But just take the drivers outside their comfort zone:
http://www.diyaudio.com/forums/multi-way/190663-can-you-have-sparkling-treble-but-without-sibilance-2.html#post2604196
A two way is a seriously hard speaker to get right. So many compromises. but those old engineers got things greatly right just using their ears:
Rogers Loudspeakers › LS5/9
What you do, is find a good idea. And then fiddle with it. Dispersion is difficult. So many theories. But what I like about John DeVore's upside down placement with a 6" is that it seems to solve the 500Hz peak that most small speakers have.
IMO, the theory isn't perfect. So I prefer trying a fun idea. I've had as many disasters as the next man.
http://www.diyaudio.com/forums/multi-way/305201-visaton-2-way-tower-monitor-audio-bronze-5-a-6.html#post5023831
Any mug can use target curves like LR2, BW3 and LR4 to design a speaker. These days I think that what you tend to hear is low distortion, a flat power response and an easy to drive impedance that your amp will like.
TBH, I have seen so many badly designed speakers that claim to be a perfect 24dB/octave LR4 or some such, and therefore perfect. But just take the drivers outside their comfort zone:
http://www.diyaudio.com/forums/multi-way/190663-can-you-have-sparkling-treble-but-without-sibilance-2.html#post2604196
A two way is a seriously hard speaker to get right. So many compromises. but those old engineers got things greatly right just using their ears:
Rogers Loudspeakers › LS5/9
What you do, is find a good idea. And then fiddle with it. Dispersion is difficult. So many theories. But what I like about John DeVore's upside down placement with a 6" is that it seems to solve the 500Hz peak that most small speakers have.
IMO, the theory isn't perfect. So I prefer trying a fun idea. I've had as many disasters as the next man.
Thank you, and yes, I agree, but what I mean about the frequency response is that I can't get what you get using the Gibbon 88 layout. But that's just what I'm finding.
Either way, I do see the nice roll-off of the tweeter with that layout. And I agree it's important because my Celestions struggle with a few songs, in particular, May it Be from LOTR OST. When Enya sings the Elvish chorus, the tweeters crack up something terrible.
So then, you mention a double reflex. I'm just going with 32L closed. So it should be something like this:
Is that right (with the top chamber being the 32L)?
Either way, I do see the nice roll-off of the tweeter with that layout. And I agree it's important because my Celestions struggle with a few songs, in particular, May it Be from LOTR OST. When Enya sings the Elvish chorus, the tweeters crack up something terrible.
So then, you mention a double reflex. I'm just going with 32L closed. So it should be something like this:
An externally hosted image should be here but it was not working when we last tested it.
Is that right (with the top chamber being the 32L)?
I had the chance to take my amplifier to the shop here where you get the best of the best audiophile equipment. I hooked up a pair of Monitor Audio Bronze 6's and listened to plenty music on them. I also listened to B&W CM10s. Both speakers sound great, the CM10s are more balanced, but don't have as much bass extension. I would say they need a sub. I absolutely love those Bronze 6s. I'll listen to the 683s soon for a comparison, but I'm definitely going to try get my hands on a pair of Bronze 6s.
Either way, I'm going to build the Gibbon 88 clone.
Either way, I'm going to build the Gibbon 88 clone.
I've already started my final planning. Capacitors aren't the easiest thing to come by for this application, so I'm probably going to use 1uF, 2.2uF and 4.4uF capacitors in various combinations. I think i need about 90m of enamelled wire, and I can get 500g rolls, which are 44m each, so that should be fine. I'm very keen to do a peg board for the crossovers.
Then, I'm going to make huge bookshelf speakers, or rather stand mount speakers. I'm going to make a 30L vented box with the vent in front. The port will be 5cm diameter and about 15cm long. I'll make the box about 25cm wide, 35cm deep and 46cm high. I'll put the drivers on in a similar layout to the gibbon 88s. I'm going to try the box with a plug in the port and not and see which I prefer. It's not long before I start buying my stuff, and I'm really looking forward to it.
I'm also going to buy monitor audio bronze 6's when I can. They're so good! And they look amazing! I guess when I do get the MA's, I'll pretty much be able to answer my own question.
Then, I'm going to make huge bookshelf speakers, or rather stand mount speakers. I'm going to make a 30L vented box with the vent in front. The port will be 5cm diameter and about 15cm long. I'll make the box about 25cm wide, 35cm deep and 46cm high. I'll put the drivers on in a similar layout to the gibbon 88s. I'm going to try the box with a plug in the port and not and see which I prefer. It's not long before I start buying my stuff, and I'm really looking forward to it.
I'm also going to buy monitor audio bronze 6's when I can. They're so good! And they look amazing! I guess when I do get the MA's, I'll pretty much be able to answer my own question.
I've had huge success with very cheap drivers. Case in point; I bought a 10" "300W" driver for a friend. Cost next to nothing, but the weight of the magnet looked promising. I measured the driver, decided a 45 L sealed box would suit it, built in a 100 W amplifier, and that sub shakes the windows at 25 Hz! I also did some compensating filters. I knew that with a cheap driver, distortion may be a problem, so I just made it a down-firing sub. Sure enough, it did have distortion, but it was an unexpected single fundamental frequency. Still, the higher frequencies vanish when you face it downward.
I have nothing against cheap drivers, provided they can measure well. The Visatons are my first "expensive" set of drivers I've bought for a build. This is why I've put much more thought into them. I also decided on a 2-way because I understand the electronics better. I can move to a 3-way when I've learned more. Then, I'm going with bookshelf type speakers because that's what I need in my dining area.
I'm going to build and listen. I'll post what I do, and hopefully these work well. When I'm done, I'll post a properly recorded video of the Celestions vs the new ones.
In the future, I can always use the drivers I buy now in a different enclosure, and support them with other drivers. But that's not a current problem. There's still a long way to go.
I have nothing against cheap drivers, provided they can measure well. The Visatons are my first "expensive" set of drivers I've bought for a build. This is why I've put much more thought into them. I also decided on a 2-way because I understand the electronics better. I can move to a 3-way when I've learned more. Then, I'm going with bookshelf type speakers because that's what I need in my dining area.
I'm going to build and listen. I'll post what I do, and hopefully these work well. When I'm done, I'll post a properly recorded video of the Celestions vs the new ones.
In the future, I can always use the drivers I buy now in a different enclosure, and support them with other drivers. But that's not a current problem. There's still a long way to go.
diyaudio.com is one of my most frustrating forums. And, TBH, I make multiple contributions to the theory, all around the internet.
Every day, we listen to these old theories of a loudspeaker sounding like reality. Well, actually, it's kinda built into the Universe that loudspeakers DON'T sound like the original sound. And thank goodness we can distinguish illusion from reality.
I have built loudspeakers with spherical metal domes:
I have built loudspeakers with cone tweeters:
TBH, I would go for the cone tweeters every time. Simply musical, even if a bit poor on dispersion.
But an easy and flat amplifier load is such a no-brainer that I would also incorporate it. The bass circuit's impedance correction might as well use NP electrolytic capacitors as more expensive devices. After all, they are in series with resistance.
Every day, we listen to these old theories of a loudspeaker sounding like reality. Well, actually, it's kinda built into the Universe that loudspeakers DON'T sound like the original sound. And thank goodness we can distinguish illusion from reality.
I have built loudspeakers with spherical metal domes:
I have built loudspeakers with cone tweeters:
TBH, I would go for the cone tweeters every time. Simply musical, even if a bit poor on dispersion.
But an easy and flat amplifier load is such a no-brainer that I would also incorporate it. The bass circuit's impedance correction might as well use NP electrolytic capacitors as more expensive devices. After all, they are in series with resistance.
I've found polypropylene capacitors for all the values. I've had to use a parallel combination to get 15 uF and 6.8 uF, but it's worked out quite well. The caps are 275 V. Then I'm going to use two 33 uF (polarized) capacitors in series to get my ~18 uF. These are 100 V each. I'm using 16 AWG enameled wire. I ordered a kilogram, which is hopefully 80 - 90 m. I'm winding air core inductors.
I did some tests. The W170S does sound nice, and it has excellent bass extension. At 20 Hz, it's silent, so that's a good point for lack of distortion. I bought another 6.5" which cost about 8 times less, and that driver sounds very good too, but the Visaton is definitely in another class. So that's good.
Steve, I've followed your advice very closely. But I'm making one deviation with my crossover - rather than a 1.6 mH inductor in the bass part, I'm using another 0.82 mH. It's going to be much easier to wind 4 identical inductors, and the 80-odd extra turns are going to put me over the wire length budget. The impedance still looks great. Let's see how it works out. Components should arrive next week, which is well in time; I won't have the drivers in boxes by then.
I did some tests. The W170S does sound nice, and it has excellent bass extension. At 20 Hz, it's silent, so that's a good point for lack of distortion. I bought another 6.5" which cost about 8 times less, and that driver sounds very good too, but the Visaton is definitely in another class. So that's good.
Steve, I've followed your advice very closely. But I'm making one deviation with my crossover - rather than a 1.6 mH inductor in the bass part, I'm using another 0.82 mH. It's going to be much easier to wind 4 identical inductors, and the 80-odd extra turns are going to put me over the wire length budget. The impedance still looks great. Let's see how it works out. Components should arrive next week, which is well in time; I won't have the drivers in boxes by then.
Steve, we are now on the same level, tweeter wise. From now on I'm working primarily with cone tweeters for these sound nice and are half as expensive as budget range domes. I have designed a filter for the project I was talking about recently. Very satisfied with results. 
Marc, you should get 86 m out of 1 kg enamelled 16awg wire which is exactly enough to make 4x 0,82mH inductors on a 20 mm diameter bobbin. Good luck!
Marc, you should get 86 m out of 1 kg enamelled 16awg wire which is exactly enough to make 4x 0,82mH inductors on a 20 mm diameter bobbin. Good luck!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.