Hey there, this is a dedicated thread on my open baffle speakers. The speakers are quite easy to build, easy to tweak or upgrade, and they are pretty decent especially for the cost.
In the following posts, I'll provide details on the speakers, so you can replicate them and/or discuss with me about how to improve the speakers.
Disclaimer: I'm a recreational speaker builder with no commercial intensions. The present speakers are considered by me as being in beta phase. They are definitely ready to rock and I'm listening exclusively on them for almost a half year now. I will provide my wishlist of improvements along the journey here.
In the following posts, I'll provide details on the speakers, so you can replicate them and/or discuss with me about how to improve the speakers.
Disclaimer: I'm a recreational speaker builder with no commercial intensions. The present speakers are considered by me as being in beta phase. They are definitely ready to rock and I'm listening exclusively on them for almost a half year now. I will provide my wishlist of improvements along the journey here.
First things first: here are some images of the speakers. I'll follow up with some specs on the drivers and the crossover. I'll provide some dimensional measurements and materials of the speaker "cabinet". Later, I'll make some response measurements. With impedance graphs and sensitivity measurement, I'll need your help, since I never made these before, I guess.
I'll also provide some sample videos on the way. Feel free to ask or discuss at any time.
I'll also provide some sample videos on the way. Feel free to ask or discuss at any time.
Here are the speakers. The cable management and the crossover housing is not existing at the moment. I plan to somehow guide the cables through some kind of cotton tunnel and just let one cable go away from the speakers. The crossover is planned to be located externally near the amplivier, rather than somewhere near the speaker. Whether this is a good idea I don't know.
It's a good idea to support the weight of the cable on the chassis, and to provide support there in case the cable is pulled.
This is the current crossover. I'd consider it as a budget release candidat. It is keept quite simple and therefore somehow affordable. In the near future, I'll order a bunch of parts and work on a more refined version. The final version will than be with more premium parts, GR Research style.
Initially I simulated a version of the crossover using vituixCAD2. Instead of the response curve provided by the manufacturer, I measured the drivers (approx. 10 cm away) and used this response curve for simulation. I used the simulated crossover as a starting point to build the actual crossover. Than, I started to measure the speaker with the crossover applied, tried to address some problem areas and measured again. I'm not sure if this is the orthodox routine to tune a crossover, though, it seems quite intuitive to me. Finally, I came up with the crossover you see in the pictures.
Initially I simulated a version of the crossover using vituixCAD2. Instead of the response curve provided by the manufacturer, I measured the drivers (approx. 10 cm away) and used this response curve for simulation. I used the simulated crossover as a starting point to build the actual crossover. Than, I started to measure the speaker with the crossover applied, tried to address some problem areas and measured again. I'm not sure if this is the orthodox routine to tune a crossover, though, it seems quite intuitive to me. Finally, I came up with the crossover you see in the pictures.
It isn't you, there are no pictures.. but either Mooly or I would be happy to put them into their respective posts for Irkiosan as they come through.
hi there. having some problems with uploading pictures via laptop. I'll do it via smartphone. This worked in the past.
Very cool design. You should get a frugal-award for it.
If had seen this example I would have decided for the WS25E although I´m probably just as happy with a (more expensive but still kinda frugal) 830669.
If had seen this example I would have decided for the WS25E although I´m probably just as happy with a (more expensive but still kinda frugal) 830669.
But not on an open baffle. I´m not a simulation expert but Basta is a pretty good SW for OB and it indicates that the two WS25E with those XO would run out of xmax=6mm at 85dB SPL. But then I know that this is plenty loud for most of us.
When I remember correctly, hornresp makes rather similarly conservative predictions. In real life these woofers perform better. Howerver, I never actually measured it.
I had similar positive experiences with other Visaton chassis, like the WS 13 E for example. For the price, the WS 25 E are a great choice.
One woofer rolls off with a second order low pass filter quite low at about 120Hz (top on the chart = bottom woofer on the baffle)
The other woofer crosses over with the tweeter. It has to roll off rather steap, therfore I applied a third order low pass filter. I had to apply a notch filter to ground to compensate for a unpleasant hump. (Middle part on the char).
The tweeter is crossed over as low as possible to provide a rather smooth transition since the woofer doesn't perform pretty good in the higher range. Luckily, a first order high pass was sufficient to provide a good result. The resistor is there to pull the outputof the tweeter down to match the woofer output.
What you show is the simulation for a driver in a CLOSED BOX of infinite volume.
When you mount a woofer in an open baffle, the near-resonance behavior is indeed like what you get for an infinite closed box. But this does not at all take into account the losses from the open baffle itself. You need to calculate these "baffle response losses" for your drivers and combine it with your infinite box model response.
One program that makes it easy to calculate the open baffle response curve is "The Edge" by Tolvan Data. I used it to simulate the contribution of the open baffle (size, number of woofers, and mounting position) on the woofers' output and have attached the response. This is for a position that is approximately equal to the top of the baffle and at 3m distant from it. You can see that the curve slopes down and this will mean some loss in bass output level compared to the tweeter level.
Have you compensated for the response I show in the attachment as part of the crossover, apart from using one woofer only at low frequencies?
Also, if you measure the drivers but only 10cm away, you will only get part of the open baffle losses in the measurements because you are partly in the nearfield at that distance. It is better to do a true nearfield measurement on a woofer and then combine that with a model of the open baffle loss response. Alternately you can do a measurement outdoors at 1m distance or more and with the speaker elevated off of the ground.
For a first pass and given the drivers used, I think this is a good job so far.
Attachments
Hey there. As already posted I post some short impressions of the speakers here. The video clips are recorded on a Galaxy Tab S6, so the sound quality isn't that great. However, it is good enough to get an impression of the capabilities of the speakers.
What you obviously cannot perceive in the videos is the soundstage of the speakers. If set up in a equilateral triangle, the soundstage is quite deep. Especially in good live recordings you have the impression the singer's voice comes from behind the televion. You find yourself searching for some kind of center speaker 🙂 Unfortunately, I don't have the space to test how wide the soundstage can go (when putting them in a larger equilateral triangle). When I put the speakers in another triangular setup the depth of the soundstage collapses into the typical stereo impression.
Sample 1:
DIY Open Baffle speakers | demo 1 - LinuxRocks PeerTube
Sample 2:
DIY Open Baffle speakers | demo 2 - LinuxRocks PeerTube
Sample 3:
DIY Open Baffle speakers | demo 3 - LinuxRocks PeerTube
If you wish, I can do more samples with songs you suggest.
What you obviously cannot perceive in the videos is the soundstage of the speakers. If set up in a equilateral triangle, the soundstage is quite deep. Especially in good live recordings you have the impression the singer's voice comes from behind the televion. You find yourself searching for some kind of center speaker 🙂 Unfortunately, I don't have the space to test how wide the soundstage can go (when putting them in a larger equilateral triangle). When I put the speakers in another triangular setup the depth of the soundstage collapses into the typical stereo impression.
Sample 1:
DIY Open Baffle speakers | demo 1 - LinuxRocks PeerTube
Sample 2:
DIY Open Baffle speakers | demo 2 - LinuxRocks PeerTube
Sample 3:
DIY Open Baffle speakers | demo 3 - LinuxRocks PeerTube
If you wish, I can do more samples with songs you suggest.