Dont get the reason for the added resistor. The idea is to see if the amp is driving the speaker properly, the resistor gets rid of the amp load (unless its in parallel to the speaker). Measure the voltage vs freq with the speakers attached the way you listen to them. This measurement should be flat, or your amp is also an EQ.
From Toole: measure in the far-field and scale to 1m distance.
You really need to get these outside so that you can determine whether you have a problem with the speaker or with the room.
Get some beer and some friends -- don't let them drink too much beer before they have humped them outside & then back in,
dave
Yuch, that can change the system Q, but the driver is still the same. For a real difference the increased Re needs to be in the speaker itself. Plus wasteful, and lose damping, and on and on.
The original SPL of the port looked more centered on 35 Hz. But it would be interesting to see NEAR-field measurements, i.e.:
- Mic stuck right into the port
- Mic stuck so close to the middle of the woofer cone (between the cap and surround) that the SPL doesn't increase any more. So close the cone almost hits the microphone.
THAT would be interesting to me. Then we could know what the speaker is doing, and discuss the room.
Lest some kind of null be un-loading the woofer, good to do this for both speakers &/or move them to a different spot in the room (since outside seemed to be problematic) and repeat.
The plot of the port is with the microphone stuffed about 3" into the right port opening.
I can do a near field of the woofer as you request later.
I've not had a chance to use 3.x (no Leopard for me, and no machine capable of running snow Leopard). Execution of impedance in 1.x is very well done (even with the bug in the routine that bites on some drivers). Impedance in 2.x is problematic, but i belive 3.x does it similarily (in the plugins menu). You will still need to make the jig.
dave
dave
I don't see any inconsistancies here. The system seems to be tuned too low, the driver may well be stiffer than normal since the cone has been replaced. The 1 meter response looks about like it would in the case. The room is very large and descriobed as "open" to other parts of the space. Basically, no bass is what I would expect. Retuning the system (higher) will boost the 40 -60 Hz response, but in a room that large, leaky as he said, the bass is going to die below the tuning and that will sound thin. What you need is more subs - spaced arround the room of course. THEN you will get the bass that you are looking for. For the most part a measured rise in the bass of about 3 dB is required in a small room to "sound right". Flat bass will always sound thin unless its very loud.
Yes, Qe has gone up (motor strength is down) from spec., volume requirements have increased. i.e. you need a bigger box. With a larger enclosure volume "peaking" in the midbass will also reduce. (or.. driver is closer to spec. and your enclosure volume is lower than you think it is.) Hell, Qm could be a bit higher (less stiff), but this generally has far less impact on Vas.
Last edited:
After reconing, most PA drivers I know of do sound thin, it will get quite better with time. Also the Qms is pretty lowish, all the more reason to let it loosen up a bit.
I was put of balance by the plot from the port, there the tuning looks pretty spot one. If the frequency response at 1 mtr is room node free it indicates that tuning to 35 Hz will give a massively better response. A strong 40 Hz from a 15" (or larger) PA driver will give 99% of the audience the idea of 'deep' bass performance.
Actually I'm using two 18LW1400 in 1500 ltr tapped horns, going flat to 20 Hz. For 97,5% of music it doesn't do a thing, 40 Hz is more than enough. HT however....drool. It has inspired me though to try convince the local wood shop they should sponsor me to do 140 dB@20 Hz, half space. Anyway, waaay off topic. Good luck with your quest for bass.
Regards Johan
I was put of balance by the plot from the port, there the tuning looks pretty spot one. If the frequency response at 1 mtr is room node free it indicates that tuning to 35 Hz will give a massively better response. A strong 40 Hz from a 15" (or larger) PA driver will give 99% of the audience the idea of 'deep' bass performance.
Actually I'm using two 18LW1400 in 1500 ltr tapped horns, going flat to 20 Hz. For 97,5% of music it doesn't do a thing, 40 Hz is more than enough. HT however....drool. It has inspired me though to try convince the local wood shop they should sponsor me to do 140 dB@20 Hz, half space. Anyway, waaay off topic. Good luck with your quest for bass.
Regards Johan
I have what I think are impedance plots for this cabinet:
Here is the raw frequency plot data when plotted with a 829 Ohm resistor in series with the amp and the voltage measured across the speaker terminals:
The program I am using (Fuzzmeasure 3.2) is very fuzzy about how to actually generate impedance plots from the frequency sweep and I did my best:
My interpretation is that the cabinet appears to be tuned to 28 Hz, which is not too far off the predicted 30 to 32 Hz as per design.
I did more careful sweeps last night with the speaker in numerous orientations and the mic at various locations, including off-axis.
The real rolloff appears to be 45 Hz. I got this a few "angstroms"
from the front of the dust cap of the 2235H.
When I inserted the mic all the way into the port as far as I could go I got this:
The blue trace is the inside port mic. The magenta trace is at the dust cap. I did not normalize the SPL for the two plots, so the actual relationship is arbitrary for the levels, but accurate for the frequency.
I have a hypothesis that I will test later, but the ports look like they may not be augmenting the woofer enough. In other words, the actual acoustic output from the ports should reinforce the woofer's roll off at about 45 Hz and down.
As I stated before, there is a vertical 1" sheet of poly fill stretched across the interior of the cabinet over some braces. This subdivides the internal box behind the woofer and is about 4" from the tail end of the ports.
My hypothesis is that the this sheet is obstructing the ports slightly and prevents the ports from doing their job.
I could also be totally full of beans and nothing like that is actually happening, but I wanted to tear that sheet out and retest the speaker for any changes. Murphy says that the amount of improvement will be inversely proportional to the amount of cerebral and physical effort exerted toward any positive goal. So I expect to fall flat on my face.
Here is the raw frequency plot data when plotted with a 829 Ohm resistor in series with the amp and the voltage measured across the speaker terminals:
An externally hosted image should be here but it was not working when we last tested it.
The program I am using (Fuzzmeasure 3.2) is very fuzzy about how to actually generate impedance plots from the frequency sweep and I did my best:
An externally hosted image should be here but it was not working when we last tested it.
My interpretation is that the cabinet appears to be tuned to 28 Hz, which is not too far off the predicted 30 to 32 Hz as per design.
I did more careful sweeps last night with the speaker in numerous orientations and the mic at various locations, including off-axis.
The real rolloff appears to be 45 Hz. I got this a few "angstroms"
from the front of the dust cap of the 2235H.When I inserted the mic all the way into the port as far as I could go I got this:
An externally hosted image should be here but it was not working when we last tested it.
The blue trace is the inside port mic. The magenta trace is at the dust cap. I did not normalize the SPL for the two plots, so the actual relationship is arbitrary for the levels, but accurate for the frequency.
I have a hypothesis that I will test later, but the ports look like they may not be augmenting the woofer enough. In other words, the actual acoustic output from the ports should reinforce the woofer's roll off at about 45 Hz and down.
As I stated before, there is a vertical 1" sheet of poly fill stretched across the interior of the cabinet over some braces. This subdivides the internal box behind the woofer and is about 4" from the tail end of the ports.
My hypothesis is that the this sheet is obstructing the ports slightly and prevents the ports from doing their job.
I could also be totally full of beans and nothing like that is actually happening, but I wanted to tear that sheet out and retest the speaker for any changes. Murphy says that the amount of improvement will be inversely proportional to the amount of cerebral and physical effort exerted toward any positive goal. So I expect to fall flat on my face.
Last edited:
Someone asked for the plot and there it is.
The problem with measuring just the port at the mouth is that the port is adjacent to the woofer. So, it is not possible to isolate the port from the woofer.
The first chart shows the characteristic bass reflex impedance with the minima between the double humps to be an accurate measure of the box tuning ie 28 Hz.
So to tune higher would mean shaving length from the vent. Correct tuning would tailor the quality but not substantially changing the amount of bass. For that the crossover network would need adjustment (bigger inductor) and/or padding the top end.
Madisound ( in the past) is notorious for going light on BSC in their quest for flat LEAP analysis. Perhaps they would offer free support to modify their "mix-up". LOL
So to tune higher would mean shaving length from the vent. Correct tuning would tailor the quality but not substantially changing the amount of bass. For that the crossover network would need adjustment (bigger inductor) and/or padding the top end.
Madisound ( in the past) is notorious for going light on BSC in their quest for flat LEAP analysis. Perhaps they would offer free support to modify their "mix-up". LOL
Last edited:
Looking at the 1 m response it looks like the tweeter level is nicely down. so perhaps tuning the box 10 Hz higher to peak the low end for more non-dry bass and padding the mid down closer to the tweeter and possibly placed more against the wall could be a short term fix. IMO
FWIW Try some/another 1 m measurement w much more smoothing and move things (mic and box) around abit to see how much the LF changes.
"As I stated before, there is a vertical 1" sheet of poly fill stretched across the interior of the cabinet over some braces. This subdivides the internal box behind the woofer and is about 4" from the tail end of the ports.
My hypothesis is that the this sheet is obstructing the ports slightly and prevents the ports from doing their job."
nah this is a NOT bad thing, anyhow not at 4" away and won't effect the ports primary contribution at 28Hz maybe helpful at the pipes upper resonances tho.
FWIW Try some/another 1 m measurement w much more smoothing and move things (mic and box) around abit to see how much the LF changes.
"As I stated before, there is a vertical 1" sheet of poly fill stretched across the interior of the cabinet over some braces. This subdivides the internal box behind the woofer and is about 4" from the tail end of the ports.
My hypothesis is that the this sheet is obstructing the ports slightly and prevents the ports from doing their job."
nah this is a NOT bad thing, anyhow not at 4" away and won't effect the ports primary contribution at 28Hz maybe helpful at the pipes upper resonances tho.
Here are some 3-D renderings of the room, which combines a kitchen and family room. Total room floor size is 15 feet by 29 feet. Lowest ceiling is 8 feet and it rises to 12 feet at the one side.
Some walls are invisible for clarity in the drawing and the roof is obviously not shown:
Combined far-field frequency response tests were swept in this configuration with the mic at the sofa.
A future plan is to rotate everything 90° so both speakers are against the end wall with the pull-down projector screen between the speakers. That box suspended in the middle of the air is the projector box. It pulls out of the upper wall as a sliding beam that pockets into the ceiling.
I ran a combined sweep in this configuration and the overall frequency response tended to flatten out at the bass and mid, but had a weird dip at the high end.
Obviously, some room treatments are needed.
Some walls are invisible for clarity in the drawing and the roof is obviously not shown:
An externally hosted image should be here but it was not working when we last tested it.
Combined far-field frequency response tests were swept in this configuration with the mic at the sofa.
A future plan is to rotate everything 90° so both speakers are against the end wall with the pull-down projector screen between the speakers. That box suspended in the middle of the air is the projector box. It pulls out of the upper wall as a sliding beam that pockets into the ceiling.
An externally hosted image should be here but it was not working when we last tested it.
I ran a combined sweep in this configuration and the overall frequency response tended to flatten out at the bass and mid, but had a weird dip at the high end.
Obviously, some room treatments are needed.
Err... the smoothed response was when I moved the sofa and speakers to the second layout with both cabinets tight against each corner, but angled so that they faced the center of the sofa.
What was interesting was when I pulled the cabinets out from the corner 12" in each direction the frequency response was closer to the that of layout 1. So, in this scenario the best response was with both cabinets tight into the corner with their faces angled toward the center of the sofa.
Also, the maximum height of the ceiling was 12 feet, not 20. 20 feet would be a huge loft, no?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.