.. Being transmitted via photon emission from either flat or slightly curved surfaces, and makes direct and clinically significant changes in the behaviour of our species, but only small fraction of individuals are affected. Genetic and environmental factors are suspected..
That is certainly one advantage of using the active EQ, you can more easily fix any response slopes or dips with a few button clicks rather than having to modify some of the resistor values.
Have you determined the resonance frequency?
BTW Congrats on completing another very nice build
That is certainly one advantage of using the active EQ, you can more easily fix any response slopes or dips with a few button clicks rather than having to modify some of the resistor values.
Have you determined the resonance frequency?
BTW Congrats on completing another very nice build
The diaphragm is tensioned pretty high. To find the resonance I downloaded a tone generator App to my iPad and used its tiny speaker to excite the diaphragm while I listened with my ear on the panel. The Ipad's speaker is so small I couldn't even hear it until the tone sweep got above 80 Hz and it rapidly got louder going up. This made the task a bit difficulty buy I'm pretty sure I heard the diaphragm's resonance peaking around 115 Hz.
Last edited:
Did you have the bias supply powered up at the time you performed the test? Remember Fs (resonance frequency) will be reduced as bias voltage is increased. To make it easier to hear, you could run a patch cord from your iPad headphone output to the the input of the amplifier driving the panels. Just be sure you have the volume on the iPad turned down before you start the test. I often do this with my cell phone.
If your diaphragm resonance is 115Hz, that is just an octave away from your crossover point and my guestimation is that it would have a Q of about 8.
Attachment #1: shows the response without and without your 24dB LR(Linkwitz-Riley) HPfilter set at 218Hz. You can see that the diaphragm resonance puts 20dB more output at resonance than the target filter response.
Attachment #1: One way to fix this is to add a notch filter to the signal path in your active crossover. You have the DCX-2496, so you should be able to add a 115Hz EQ filter with Gain = -20dB and Q=8. This will cancel out the primary diaphragm resonance and leave you with the desired response. I plotted a couple other curves to show that even if you miss the exact frequency by +/-10Hz you should still have a very workable crossover response.
NOTE: You should probably test each panel as they may have slightly different Fs. Also, as Bazukas mentioned, since you used mechanical tension method with no heat annealing the Fs will likely decrease over the course of 6 months as you play the speakers. You may want to check the Fs later in the year to see if they have drifted.
Alternatively, you could attack the problem acoustically by adding some damping fabric to the rear stator as golfnut suggested. In your case, I would think a layer of thin crafting felt would work great. You can get it cheaply at most fabric stores. You might consider leaving a 2" gap running down the middle to leave the segment handling the higher frequencies free from potential filtering affects. One nice thing about acoustic damping is that it does not require tuning for a particular frequency. It works just the same even if Fs drifts.
Attachments
Thanks Steve,
You can sim just about everything!
Glad to see I was kinda on the same page with you on a fix.... last night I upped the X-over frequency and slope to 250Hz/48db and then applied a narrow -12db parametric EQ at 112 Hz and it tamped down the resonance quite nicely.
And this morning I did a time alignment with the DCX and some fine tuning with the DEQ and now these panels are sounding exquisitely good.
You can sim just about everything!
Glad to see I was kinda on the same page with you on a fix.... last night I upped the X-over frequency and slope to 250Hz/48db and then applied a narrow -12db parametric EQ at 112 Hz and it tamped down the resonance quite nicely.
And this morning I did a time alignment with the DCX and some fine tuning with the DEQ and now these panels are sounding exquisitely good.
Last edited:
Good deal
There is one other thing concerning the diaphragm resonance you might investigate.
Your previous panels used vertical diaphragm sectioning, so only a small portion of each section was in close proximity to the woofer. You current panels with their horizontal sectioning has the entire area of the lowest diaphragm section in close proximity to the woofer. You might try playing some bass heavy music and disconnect the ESL while continuing to play the woofer. Then listen with an ear near each diaphragm section and see if the lowest diaphragm section is being excited by the woofer to a larger degree than the other sections. If this is the case, you may consider adding some damping cloth to the rear stator of just this panel section to equalize its response with that of the other panel sections.
Last edited:
I just finished up installing the mode switches this evening so I'm now all done building my new segmented ESL's.
I have them playing in narrow dispersion mode at the moment and I've not yet decided which mode I prefer, as each has advantages.
The narrow mode gives the most amazing imaging and slam but can be a bit intense and piercing on loud passages.
Wide mode eliminates the "head in a vice" effect and is not as "in your face", yet still retains quite good imaging without giving up much slam (can play loud as hell, in fact).
Switching between wide/narrow dispersion modes takes about 90 seconds to crank volume to zero (too much voltage to risk switching under load), turn selector switches on each speaker, recall the saved EQ curve on the Behringer, and then crank the volume back up.
I'm officially calling this project a success...
Thanks guys for all the help and encouragement!
Panel, resistor filter network and mode switch box:
I have them playing in narrow dispersion mode at the moment and I've not yet decided which mode I prefer, as each has advantages.
The narrow mode gives the most amazing imaging and slam but can be a bit intense and piercing on loud passages.
Wide mode eliminates the "head in a vice" effect and is not as "in your face", yet still retains quite good imaging without giving up much slam (can play loud as hell, in fact).
Switching between wide/narrow dispersion modes takes about 90 seconds to crank volume to zero (too much voltage to risk switching under load), turn selector switches on each speaker, recall the saved EQ curve on the Behringer, and then crank the volume back up.
I'm officially calling this project a success...
Thanks guys for all the help and encouragement!
Panel, resistor filter network and mode switch box:
Last edited:
Very Very Cool Charlie!!
I will be headed your way very very soon in the near future.
Maybe I will have the time to stop by and give them a listen as well!!
jer
It would be great to finally meet you, Jer. I'm PM'ing my address.
Charlie
Very similar except the treble end doesn't need as much boost in the narrow mode (as it beams at you).
Both modes need some boost to offset the dipole roll-off-- down to where the woofers kick in.
Upper mids are very bright in both modes (probably characteristic of the transformers) and my ears are very sensitive to the 2.5kz band so I pull the mids down quite a bit.
BTW, the settings are to please my ears (nowhere near true flat).
Narrow mode EQ:
Wide mode EQ:
Last edited:
Hi !
I have been following this thread to learn about ESL to possibly build one. However I have a rather rudimentary question regarding wider dispersion achieved using Eggcrate Louvers. What is the qualitative difference between the flat perforated metal screen which yields narrow dispersion and the eggcrate panel ?
Jayant
I have been following this thread to learn about ESL to possibly build one. However I have a rather rudimentary question regarding wider dispersion achieved using Eggcrate Louvers. What is the qualitative difference between the flat perforated metal screen which yields narrow dispersion and the eggcrate panel ?
Jayant
Hi,
They almost always require a slight notching in the 1-4kHz range, otherwise high pitched Sopranos voices get a too shrill note.
Its only due to the vastly superior sonics of an ESL that it doesn't split Your ears and doesn't pull the inlets from Your teeth, as almost all dynamic drivers with such a response would.
Instead a very slight raise over that range increases 'directness' and 'speed' and influences on staging.
jauu
Calvin
This is the typical characteristic of tall thin panels.
They almost always require a slight notching in the 1-4kHz range, otherwise high pitched Sopranos voices get a too shrill note.
Its only due to the vastly superior sonics of an ESL that it doesn't split Your ears and doesn't pull the inlets from Your teeth, as almost all dynamic drivers with such a response would.
Instead a very slight raise over that range increases 'directness' and 'speed' and influences on staging.
jauu
Calvin
With the resistor delay network switched out (defeated) there are essentially no qualitative differences between my old perf-metal panels and the new egg crate/wire panels.
Both flat panels beam sound to a tightly focused sweet spot about a foot wide and their sound, as heard from the focus, seems to emanate more from the space between the speakers than from the speakers themselves. And oh yes, the constricted sweet spot indeed puts your head in a vise (not necessarily a bad thing but useless for more than one listener).
When the resistor network delay networks are switched in, the egg crate panels project a much wider sweet spot and their sound, as heard from the focus, seems to emanate a bit less from the space between the speakers and a bit more from the speakers themselves.
I attribute all audible differences between the egg crate and perf-panels solely to the external resistor networks and to their differing diaphragm tensions (resonances).
Last edited:
Thanks Calvin.
All of your posts are enlightening. I'm sure I'm not the only one here wishing you would compile your posts into a book and publish it.
Charlie
Last edited:
Charlie,
Thanks for your explanation. Then would it be possible to avoid egg crate Panel, substituting it with insulted perforated metal panel and use the rods and resistor network to create the similar wide dispersion ?
The reason I ask is that I am finding it difficult to get egg crate panel where I live. I can't carry it personally from the US as the size is beyond what airlines allow to be checked in without a substantial penalty.
Jayant
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.