I am always experimenting with open baffle designs / configurations. My latest iteration is a small horn loaded open baffle line array. I love the Matsushita 8PW1 full range drivers but have been able to hoard only two pairs so far, with two more to go. Instead of waiting to get my hands on two more pairs, I put together an Open Baffle Line Array with the 8” different FRs I had. They are all paper cone and similar sensitivity and impedance. The Qts and Fs decreases from top to down. They are driven by a DIY pair of F2a SE monoblocks with a very special handwound OPTs with secondary as 0-1-2-4-8 ohms impedance. Currently all FRs are wired in parallel and connected at 2 ohms secondary of the OPT. I love how they sound and are little scary for my modest room. The calculated sensitivity stands at around 105 dB. Amp is driven by ECDesigns Power DAC R.
Tyimo, thanks, the profile is slightly different. Measures 7.75” wide at the mouth, 44” wide at the flare, is 24” deep and curve measures 32”.
The vocals are smoother, detailed, soundstage is another level I have never heard before, overall sensitivity is way higher than just 105dB.
The horns are removable and I can even play with different profiles within the same curve length.
New OPTs give me the freedom to play with both wiring configuration of the drivers from 8 ohm to 2 ohm impedance.
The vocals are smoother, detailed, soundstage is another level I have never heard before, overall sensitivity is way higher than just 105dB.
The horns are removable and I can even play with different profiles within the same curve length.
New OPTs give me the freedom to play with both wiring configuration of the drivers from 8 ohm to 2 ohm impedance.
I have changed mid range from 6FE100 to PRV 5MR450NDY. Distortion went way down and sound is much cleaner. Can play cleanly very loud now. Finalized new crossover and soldered into P2P boards. Finally, brought speakers into larger room and it still sounds great. Amazing dynamics and bass slam/whack/punch. 
By crossing rather high at 4500Hz and using the mid axis as the reference axis, the delays the tweeter enough so that they are time aligned if using asymmetric slopes (4th on low pass of mid and 2nd on high pass of tweeter). I have done this many times and the timing is very crisp with a good step response. Sometimes though, the reference axis can be the tweeter and it still works. For example, on my LS3/5A homage, the crossover slopes look like this:
And the step response looks like this:
Which incidentally, looks very similar to the step response of the LS3/5A (from Sterophile review), which uses a much more complicated XO at a lower frequency:
Just commenting as symmetric xo fanboy... Neither the summed spl response nor step response don't look t´pretty to me. One full octave below graphical xo point are out of phase...
I have actually tried these Harsch asymmetric xo's with minidsp, but I just didn't get any rational progress with delay settings vs. response linearity. And the sound of eg. piano with well timing-matched LR/Duelund xo simply sounds natural. The beauty of dsp is that it is so easy to change settings when listening, to compare different configs.
I have actually tried these Harsch asymmetric xo's with minidsp, but I just didn't get any rational progress with delay settings vs. response linearity. And the sound of eg. piano with well timing-matched LR/Duelund xo simply sounds natural. The beauty of dsp is that it is so easy to change settings when listening, to compare different configs.
Hi Juhazi,
It’s a compact passive bookshelf speaker with a simple 6 component XO that works with any amp. DSP active will of course, let one dial in any crossover desired but isn’t very transportable. In the end, the sound quality has been very well received by all who built it or listened to it as natural and engaging. The step response and the sound is extremely close to the well regarded LS3/5A (a review from an owner of one who compared them side by side). Open baffle speakers on the other hand, tend to be large and stationary so more suited for DSP active. But passive crossovers have the advantage of allowing one to try out different amps and not needing three amps.
It’s a compact passive bookshelf speaker with a simple 6 component XO that works with any amp. DSP active will of course, let one dial in any crossover desired but isn’t very transportable. In the end, the sound quality has been very well received by all who built it or listened to it as natural and engaging. The step response and the sound is extremely close to the well regarded LS3/5A (a review from an owner of one who compared them side by side). Open baffle speakers on the other hand, tend to be large and stationary so more suited for DSP active. But passive crossovers have the advantage of allowing one to try out different amps and not needing three amps.
Xrk, isn't this thread is supposed to be about open baffle speakers, which are not small or transportable? You didn't show or link measurements of your latest OB multiway, but instead used measurements of LS3/5 clone and said that they are practically similar.
I use dsp xo, because I can't design passives and don't want to make small/portable diy-speakers. But, crossover including delays we are supposed to evaluate and what really matters, is the acoustic outcome (verified by measurements). If it's done with passive or active, analog or dsp means is irrelevant, and not at all the point of what I said in post #3812. And to evaluate a speaker at internet forum, we need more measurements than on-axis spl and designers enthusiastic impressions. For example criticism and questions...
I use dsp xo, because I can't design passives and don't want to make small/portable diy-speakers. But, crossover including delays we are supposed to evaluate and what really matters, is the acoustic outcome (verified by measurements). If it's done with passive or active, analog or dsp means is irrelevant, and not at all the point of what I said in post #3812. And to evaluate a speaker at internet forum, we need more measurements than on-axis spl and designers enthusiastic impressions. For example criticism and questions...
Hi Juhazi,
My apologieds for posting related but not specific info on my OB XO - did not mean to confuse folks. The current XO in the mid to tweeter is probably closer to symmetric. Here is the predicted XO response curves (black is predicted system, red is woofer, blue is mid, green is tweeter, yellow is actual measured response). This is at the midrange driver axis, 0.5m away at 2.0Vrms. The physical nature of the SLOB with a chamber etc. has too much delay (about 2ms) and probably impossible to make time aligned passively without a massive physical offset between woofer and tops. The XO frequency is about 500Hz, and delay is 2ms, so somehow it doesn't seem to be an issue when listening. The time alignment on the mid and tweeter worked out very well though as you can see the phase is relatively flat until we get to the mid to woofer XO frequency. But this seems not to be noticeable as the percusive time alignment between the mid and tweeter is good. My mic has inverted polarity so all the step responses are upside down (but positive amp polarity corresponds to driver membrane pushing out).
Here is the predicted and measured XO response curves:
Here is the predicted step response:
Here is the measured step response - and the tweeter and mid are well-aligned at the leading edge, the SLOB woofer peak comes about 2ms later:
I hope that clarifies the XO I am using a bit more. All of this is done with a 12 component passive XO.
My apologieds for posting related but not specific info on my OB XO - did not mean to confuse folks. The current XO in the mid to tweeter is probably closer to symmetric. Here is the predicted XO response curves (black is predicted system, red is woofer, blue is mid, green is tweeter, yellow is actual measured response). This is at the midrange driver axis, 0.5m away at 2.0Vrms. The physical nature of the SLOB with a chamber etc. has too much delay (about 2ms) and probably impossible to make time aligned passively without a massive physical offset between woofer and tops. The XO frequency is about 500Hz, and delay is 2ms, so somehow it doesn't seem to be an issue when listening. The time alignment on the mid and tweeter worked out very well though as you can see the phase is relatively flat until we get to the mid to woofer XO frequency. But this seems not to be noticeable as the percusive time alignment between the mid and tweeter is good. My mic has inverted polarity so all the step responses are upside down (but positive amp polarity corresponds to driver membrane pushing out).
Here is the predicted and measured XO response curves:
Here is the predicted step response:
Here is the measured step response - and the tweeter and mid are well-aligned at the leading edge, the SLOB woofer peak comes about 2ms later:
I hope that clarifies the XO I am using a bit more. All of this is done with a 12 component passive XO.
Last edited:
Thank you Xrk!
Relevant info there, naturally with response peaks and dips. Some of those come from edge or cone resonances. Only you know and can test changes with eq of a single driver or different delay. Vertical off-axis changes tell more about interdriver interferences. Horizontal off-axis should remain more constant with vertically aligned drivers, but shows important directivity smoothness. Evaluating and testing woofer-mid response and xo is very difficult indoors, and I do now how difficult it is to bring these colossuses and measuring gear outdoors on a sunny day...
To all OB fans here, please show us backside response and directivity too! It is not so important in details, but predicts room (power) response. Typical case is to have a monopole tweeter, which leaves backside sound power to practically zero level, like here
https://www.princeton.edu/3D3A/Directivity/Gainphile R16/index.html
Relevant info there, naturally with response peaks and dips. Some of those come from edge or cone resonances. Only you know and can test changes with eq of a single driver or different delay. Vertical off-axis changes tell more about interdriver interferences. Horizontal off-axis should remain more constant with vertically aligned drivers, but shows important directivity smoothness. Evaluating and testing woofer-mid response and xo is very difficult indoors, and I do now how difficult it is to bring these colossuses and measuring gear outdoors on a sunny day...
To all OB fans here, please show us backside response and directivity too! It is not so important in details, but predicts room (power) response. Typical case is to have a monopole tweeter, which leaves backside sound power to practically zero level, like here
https://www.princeton.edu/3D3A/Directivity/Gainphile R16/index.html
A good measure of a speakers phase coherence is the Group Delay, it tells more than the step response.
I think posting speaker measurements is valid as this is a showcase for what people have built and the correct measurements give a good indication of how a speaker will sound in a room.
@Juhazi is correct in that the off axis response if done out doors or in a large area to minimise reflections will tell a lot about how a speaker sounds in room. Floyd Toole used this method to measure speakers in an anechoic chamber and relate it back to how the speaker will sound in room in double blind tests when he was with Harmon.
Keep up the posts.
I think posting speaker measurements is valid as this is a showcase for what people have built and the correct measurements give a good indication of how a speaker will sound in a room.
@Juhazi is correct in that the off axis response if done out doors or in a large area to minimise reflections will tell a lot about how a speaker sounds in room. Floyd Toole used this method to measure speakers in an anechoic chamber and relate it back to how the speaker will sound in room in double blind tests when he was with Harmon.
Keep up the posts.
Hi,
Is a SLOB the same thqn the open W from Linkwitz ?
What about no side walls? Too much bass loss and efficienty ? If the tile to travel is longer with the side walls, the U frame behind should drop also a little the efficienty ?
I ask as some seems to like minimum baffle loading above in the thread (baffle less dipole), so a quasy SLOB (no side walls instead), did you try that X ?
Is a SLOB the same thqn the open W from Linkwitz ?
What about no side walls? Too much bass loss and efficienty ? If the tile to travel is longer with the side walls, the U frame behind should drop also a little the efficienty ?
I ask as some seems to like minimum baffle loading above in the thread (baffle less dipole), so a quasy SLOB (no side walls instead), did you try that X ?