Interested. I have couple 2251j and found them the most natural low-midrange drivers that I ever had for 40+ years. I A/B test them and compared with PHLs, JBL M-209, 2123H, CMCD-81H, Isophons, Grundic's, Telefunkens, Goodmans, Altecs, B&C, 18Sound and lot's of others. Not even mention the low efficiency/ highly marketing stuff like ScanSpeak Revelators, Dynaudio etc. There is something "magical" in 2251j in 250-800Hz range. Very realistic sound, correct tone and overtones... I also wanted to married them with GRS PT6825 planars from 800Hz up to 4000Hz))) Can you provide exact baffle size and xover details? Your impulse responce looks exellent.
The 2251 has more Xmax than the 2123H 5.7mm apposed to 2.5mm so Vd is much better, in fact Vd id the same as the 8" driver used by SL in the LX521.
I am using a MiniDSP to XO and EQ the drivers. I am still waiting for the NEO's to arrive so I have not worked on XO yet. I have only an approximation of 160Hz and 800-900Hz for XO frequencies. I will XO to the NEO8 at the lowest F the NEO8 will work at with low distortion. I want to attenuate the 830Hz dip in the FR in the 22151 plot.
The baffle is 310mm at the widest point. I kind of just drew an outline and cut the baffle with a jigsaw. I was lucky this one worked well straight away. I used a 400mm drum to trace the curve on either side of the 2251. The top of the baffle is not finished as it's 110mm wide and will be narrowed to about 90-100mm for the NEO drivers.
I am using a MiniDSP to XO and EQ the drivers. I am still waiting for the NEO's to arrive so I have not worked on XO yet. I have only an approximation of 160Hz and 800-900Hz for XO frequencies. I will XO to the NEO8 at the lowest F the NEO8 will work at with low distortion. I want to attenuate the 830Hz dip in the FR in the 22151 plot.
The baffle is 310mm at the widest point. I kind of just drew an outline and cut the baffle with a jigsaw. I was lucky this one worked well straight away. I used a 400mm drum to trace the curve on either side of the 2251. The top of the baffle is not finished as it's 110mm wide and will be narrowed to about 90-100mm for the NEO drivers.
Nice project design. The only thing I wonder is if having the Neo on the baffle with the mid-range might vibrate the Neo? Is it worth measuring the JBL and Neo with say a single 700hz tone, then a single 1200Hz tone and then with a 700Hz and 1200Hz together and look at the FFT's?
Yes the 2251 with no baffle needed a fair amount of EQ to flatten the response. The big dip at 280Hz and ripples was a floor reflection that smoothed with floor treatment.
The base line is the 2251 in the baffle with no EQ
That's a great Idea to check for IMD products in the midrange. I have 2x signal sources so I can do this test. I would be more inclined to hit the JBL with 200Hz and the NEO8 with 4kHz as our hearing is very sensitive to 4kHz being modulated.
I re-tested the 2251 this morning and slightly altered the EQ from the plots I posted and while doing this I put my hand on the baffle to feel the vibration so you are spot on. Then I started to think about separating the baffles and hanging them from a frame.
2 less PEQ's that the previous EQ and still flat within 3dB over the range I plan to use it. The red trace is an identical measurement to the plot in my previous post with the floor treatment.
Basic mechanics would suggest to isolate the mid-range form the mid-bass excitation its best to entirely separate the midbass axial vibration. Even suspending the drivers from one frame will couple them in the axial plane.
How about mounting the planar on it own frame? The frame can be minimal so the planar is essentially naked. Two aluminum extrusions would do. Aluminum can bent to fit around the woofer. Bolt the Neo to them from the sides. Mass load the frame just below the Neo eg with a brick to prevent swaying. Mount the frame on two brick-sorbothane-brick bases. There will be minimal vibration going up the frame. What vibration there is will be in the vertical plane with minimal axial blurring action on the membrane. The Neo position can then be adjusted behind the JBL so the impulses align.
Looks good. 200Hz should really get the baffle rocking.
Yes the dipole peak is clear at about 720Hz. ripple above the dipole peak is significantly reduced by baffle shape and is about 3dB. ka=2 of the 2251 at about 835Hz so XO needs to lower than this. The plan is to test XO to the NEO8 from 650Hz. 2 things will govern where the 2251/NEO8 XO is set and these will be smooth off axis response and NEO8 distortion.
The goal of this build is to have low distortion. smooth on and off axis frequency response and good in room power response. This is relatively easy to accomplish with good drivers and a MiniDSP. The limiting factor is the 5 PEQ's in the MiniDSP 4x10HD.
The goal of this build is to have low distortion. smooth on and off axis frequency response and good in room power response. This is relatively easy to accomplish with good drivers and a MiniDSP. The limiting factor is the 5 PEQ's in the MiniDSP 4x10HD.
If you need to push the JBL dipole peak higher to meet the Neo8 can you trim the baffle? Is the JBL model the 2251J? I tired an Edge sim but Im not sure of the dimensions of the cone vs the driver mount. I estimated the cone D in mm from the Sd and the driver mounting baffle as 10'. Nude the peak is about 1100Hz. It may be close as when I made the baffle bigger the dipole peak was just above 700Hz.
Last edited:
It's a 2251JPL. I found another thread where the NEO8 was XO at 600Hz so my 650Hz XO should be no problem. This keeps the 2251 operating below the baffle step.
https://www.diyaudio.com/community/...aborative-speaker-project.231353/post-4231049
I have also given some thought to baffle vibration. A driver is also a generator, so to test isolation of the NEO all I need to do is excite the 2251 with a signal and measure the voltage generated on the NEO8 and NEO3 (with a scope or spec anal) this will tell me to what extent the diaphragm is excited by baffle vibration with less uncertainty than measuring 2 signals with a microphone.
https://www.diyaudio.com/community/...aborative-speaker-project.231353/post-4231049
I have also given some thought to baffle vibration. A driver is also a generator, so to test isolation of the NEO all I need to do is excite the 2251 with a signal and measure the voltage generated on the NEO8 and NEO3 (with a scope or spec anal) this will tell me to what extent the diaphragm is excited by baffle vibration with less uncertainty than measuring 2 signals with a microphone.
Brilliant. It would be great to measure the planar output too and correlate the two methods for proof of concept.
excite the 2251 with a signal and measure the voltage generated on the NEO8 and NEO3
That's exactly what I plan to do. I have both a scope and HP spectrum ananlyser. The SA will show any diaphragm vibration in the frequency domain. I'll post the results when I have done the testing. Still waiting for the NEO's to get through Aus Post and arrive.
Would be nice to see more close-ups and to learn more about the design (drivers, crossovers, etc)
I ended up changing the mid/woofer baffle shape to the Linkwitz to smooth off axis response, it works and it's easier to make than the curved one. I also found turning the NEO3 on its side provided much better off axis response. On axis response is flat within 3dB 200-20,000Hz. I still have to integrate the subwoofer.
XO all Linkwitz
4kHz 12dB
700Hz 24dB
Sub not integrated yet somewhere between 120-160Hz
The test baffle is 2 piece to facilitate changing upper baffle to test response. I tried a few iterations including wider, wings etc,etc and this was the best baffle
XO all Linkwitz
4kHz 12dB
700Hz 24dB
Sub not integrated yet somewhere between 120-160Hz
The test baffle is 2 piece to facilitate changing upper baffle to test response. I tried a few iterations including wider, wings etc,etc and this was the best baffle