W
W,
God Bless You, Ronald!
Was the 25 drivers measured in the completed arrays, or individually in your test box. Either way, that is a lot of drivers to mount and re-mount, wire and re-wire.
I would like to experiment with stuffing myself. I have been looking into ways to measure the impedance curve of my drivers. I could do it safe using the signal generator in True RTA, thru the power amp with a 200 -1000 ohm, 1/2 watt resister in series with the "+" side of the driver and my DMM. But looking at the wiggles on your graph, it would be too easy for me to miss something graphing this by hand. I know the same set-up can be patched back into the computer via a mic input, but I also know one could blow the mic input if one is not careful. So I am thinking if I set this up first with the DMM, I could then safely patch this into my computer. I am more tempted to use 1000 ohms with the computer thou. What does your set-up consist of? Also, what software are You using to plot the graph? I have read REW and Holm can be used for this substituting SPL for ohms on the graph.
Thanks in advance.
W,God Bless You, Ronald!
Was the 25 drivers measured in the completed arrays, or individually in your test box. Either way, that is a lot of drivers to mount and re-mount, wire and re-wire.
I would like to experiment with stuffing myself. I have been looking into ways to measure the impedance curve of my drivers. I could do it safe using the signal generator in True RTA, thru the power amp with a 200 -1000 ohm, 1/2 watt resister in series with the "+" side of the driver and my DMM. But looking at the wiggles on your graph, it would be too easy for me to miss something graphing this by hand. I know the same set-up can be patched back into the computer via a mic input, but I also know one could blow the mic input if one is not careful. So I am thinking if I set this up first with the DMM, I could then safely patch this into my computer. I am more tempted to use 1000 ohms with the computer thou. What does your set-up consist of? Also, what software are You using to plot the graph? I have read REW and Holm can be used for this substituting SPL for ohms on the graph.
Thanks in advance.
This is the page that you need to look at to use REW for impedance measurement.
https://www.roomeqwizard.com/help/help_en-GB/html/impedancemeasurement.html#top
A simple jig with a known resistance is all that is needed to use it with a soundcard. If you have a headphone amplifier output that is good to use as it can supply more current than a line out. Resistors in the range of 50 to 100 ohms seem to work best to keep the noise down from what I have read. Buy a 0.1% tolerance resistor if you can or measure with a quality DMM as the exact resistance value will determine the accuracy of the measurements. Also a line in would work better than a mic input.
I would not use a power amplifier as the power is not needed and if you short something by accident your soundcard will go up in smoke. If you really want to use a power amp then build an attenuator and put in some back to back to back zener diodes or diode strings to clamp the voltage to something that won't destroy your card. Basically the same method you would use to measure a power amplifier with the soundcard.
The link fluid has provided is exactly what I used. I tested each driver I used separately when I mounted them on my baffle. So those are all open baffle measurements.
I wanted to be sure I had no DOA drivers among them. Even if the baffle was unsupported it would show up as a wiggle in the impedance curve.
The test enclosure showed detailed impedance information. This stuff really works. You can see every ripple relate directly to an FR dip. Once you mount 25 drivers in an enclosure it's good to make sure you've got no gross errors. They would really stand out and be exaggerated in an array.
I measured using an old laptop to be save. I didn't want to blow my precious soundcard. I didn't care about exact numbers, just some piece of mind, averages being most important.
With the damping scheme, setup with the test box, the final impedance peak of the arrays landed on ~145 Hz. Lower than predicted with WinISD pro.
I wanted to be sure I had no DOA drivers among them. Even if the baffle was unsupported it would show up as a wiggle in the impedance curve.
The test enclosure showed detailed impedance information. This stuff really works. You can see every ripple relate directly to an FR dip. Once you mount 25 drivers in an enclosure it's good to make sure you've got no gross errors. They would really stand out and be exaggerated in an array.
I measured using an old laptop to be save. I didn't want to blow my precious soundcard. I didn't care about exact numbers, just some piece of mind, averages being most important.
With the damping scheme, setup with the test box, the final impedance peak of the arrays landed on ~145 Hz. Lower than predicted with WinISD pro.
One of the reasons I like to use Bassboxx pro for simulating sealed enclosures is it takes the damping into consideration and gives a better idea of the likely performance.
P.S I posted some information on my thread and posed a question to you if have the time to look
http://www.diyaudio.com/forums/full-range/303417-full-range-tc9-line-array-cnc-cabinet-3.html#post4984903
I'll have to get back to you. It's 5 A.M. over here, I just got back from a (rock) concert.
If I'm less coherent than usual it will be because of the amount of beer I consumed
I hope you enjoyed the concert and the beer, no rush whenever you can, PM sent to assist with the tracks.
I'll have to get back to you. It's 5 A.M. over here, I just got back from a (rock) concert.
If I'm less coherent than usual it will be because of the amount of beer I consumed
Nothing like the real thing!
We can just hope to get close to the real thing in our own living rooms!
Thanks, Fluid and Ronald. That is exactly what I needed to know...
It depends on which array build You are speaking. In the case of the current rebuild of the cabinets for my Cheap and Cheerful arrays, the columns will be fully assembled and I will measure and see if I can optimize the stuffing more. In the case of the Vifa build, a lot is going to depend on how the current build goes.
I will be posting photos on my thread soon, with the reasoning behind this little side adventure. Actually, another question for Wesayso: Which of the measurements on either of the two graphs You have in post #3260 shows an overstuffed chamber? No hurry, as it will be a couple of weeks yet before I can run tests. (It took me 8 hours this weekend to make two jigs for the baffle cut outs.
)
It's hard to see with to lines of near identical colour. What to look for? The impedance bump should go down and to the left. Overfill the chamber and it starts moving right again.
Look at the plot marked Dec. 02. My notes say: max stuffing, this was where the maximum damping was achieved. Any more and the peak would move to the right again.
You'll see that the angel hair/fiberglass combo measurement matches the impedance peak (in Hz) with a way less damping material in that one. It's an excellent damping material when combined with fiberglass fill for the bottom end.
Look at the plot marked Dec. 02. My notes say: max stuffing, this was where the maximum damping was achieved. Any more and the peak would move to the right again.
You'll see that the angel hair/fiberglass combo measurement matches the impedance peak (in Hz) with a way less damping material in that one. It's an excellent damping material when combined with fiberglass fill for the bottom end.
Last edited:
Thanks, Ronald
On that graph, it appears the angel hair/fiberglass performed better then the felt/fiberglass. But as You mention, the felt was not yet optimized. And if I recall correctly, You mentioned the felt performs better for mid/high frequencies and angel hair performs better for the low end. Did You ever try using both the felt and angel hair with fiberglass?
The angel hair is also only really effective from the mid upwards as only fiberglass performed best over the entire range(*).
The wool felt made it possible to get all wiggles out of the graphs.
* Yes I did try the combination of felt + angel hair + fiberglass, it made a small difference in height and placement of the impedance peak. I also tried angel hair by itself. It wasn't nearly as effective as the combination of the two. Fiberglass is still an underrated damping material, very effective. All of the other stuff I tried failed to do better, but combinations might work.
I've puzzled to keep angel hair in the mix, but after more tests the difference wasn't big enough for me.
It was the felt I used that was able to remove all remaining wiggles. I've done way more of these tests than shown here. That one was a partial result from the first day.
Here's another saved example of what I was looking for:
-Cyan is close to the final mix, you can see it being a lower overall impedance peak than the combo with the angel hair, but even more important: no more wiggles.
-Purple is open baffle driver only measurement
-Grey is box without any damping whatsoever.
I think I spend a straight 4 days total on it, just with the test box.
Even more time was spend testing the complete arrays.
The wool felt made it possible to get all wiggles out of the graphs.
* Yes I did try the combination of felt + angel hair + fiberglass, it made a small difference in height and placement of the impedance peak. I also tried angel hair by itself. It wasn't nearly as effective as the combination of the two. Fiberglass is still an underrated damping material, very effective. All of the other stuff I tried failed to do better, but combinations might work.
I've puzzled to keep angel hair in the mix, but after more tests the difference wasn't big enough for me.
It was the felt I used that was able to remove all remaining wiggles. I've done way more of these tests than shown here. That one was a partial result from the first day.
Here's another saved example of what I was looking for:
-Cyan is close to the final mix, you can see it being a lower overall impedance peak than the combo with the angel hair, but even more important: no more wiggles.
-Purple is open baffle driver only measurement
-Grey is box without any damping whatsoever.
I think I spend a straight 4 days total on it, just with the test box.
Even more time was spend testing the complete arrays.
I was almost ready for a new phase in this adventure. After thinking about what it should be, a new more powerful amplifier or adding subs I lean towards adding the subs. Mainly for added headroom.
However, I ran into a financial setback and need to be patient once more. I do have the permission, though I can't say my girl "gets" the why. For me it's curiosity driven. Seeing and learning what I can do. I'm ready for a new challenge. I've been listening non stop for months and have no valid complaints. So I might as well be patient as there's no real need for change.
However, I ran into a financial setback and need to be patient once more. I do have the permission, though I can't say my girl "gets" the why. For me it's curiosity driven. Seeing and learning what I can do. I'm ready for a new challenge. I've been listening non stop for months and have no valid complaints. So I might as well be patient as there's no real need for change.
One thing not being discussed here is the reason that many manufacturers no longer use fiberglass as a stuffing material. Fiberglass has a tendency to shed microscopic particles into the air and if you are not using a sealed enclosure those will get out through the port and into your breathing space. I think this was the driving factor in the use of the poly fill materials, that and the fact that just physically handling the glass is a real irritant to the skin.
It should be fairly obvious that the wool felt or any type of felt would be most effective at the higher frequencies, the density and thickness are the controlling factors there. So a combination such as we are talking about of felt and fiberglass would be a combination that would cover a much wider frequency band and more effective than just using fiberglass stuffing alone, it is the synergistic effect of both that is working here and it is a matter of having to do your due diligence to determine the correct amount for each application whether a line array or a typical bass reflex and tweeter design. There is no simple answer as to how much stuffing to use and the exact position to add it, it is to be determined it each design. Even if you made what you thought was a clone of Ron's array unless you used the same exact building techniques you are going to have to tweak the stuffing to achieve the same final result that Ron has been showing in his extensive testing and excellent build quality of his arrays.
It should be fairly obvious that the wool felt or any type of felt would be most effective at the higher frequencies, the density and thickness are the controlling factors there. So a combination such as we are talking about of felt and fiberglass would be a combination that would cover a much wider frequency band and more effective than just using fiberglass stuffing alone, it is the synergistic effect of both that is working here and it is a matter of having to do your due diligence to determine the correct amount for each application whether a line array or a typical bass reflex and tweeter design. There is no simple answer as to how much stuffing to use and the exact position to add it, it is to be determined it each design. Even if you made what you thought was a clone of Ron's array unless you used the same exact building techniques you are going to have to tweak the stuffing to achieve the same final result that Ron has been showing in his extensive testing and excellent build quality of his arrays.
Sorry to read about the setback. Truly hope it's temporary.
But, it's not like you have nothing to do!
I believe it is just as important to just sit back and enjoy the times we have, as it is to push the limits of what is possible.
You've been pushing for a while, so it's fair enough that you spend some time enjoying the present.
I'm sure the little bugging of trying something new will arise, but no rush....
Good post! I totally agree. If you're interested in running arrays sealed trough the entire spectrum, fiberglass insulation is a good base material to use. Combinations with wool felt can make it perform even better overall.
Don't assume, measure... it's worth the hassle.
I'm hoping it will be temporary. I'm enjoying the listening time (immensely) and still run little tests every now and then. Things like playing with my specific ambient combination is easy, just a few turns of digital knobs. Overall I'm both happy and impressed with what I hear and have. Being impressed over a long long time frame is a lot harder than to be blown away for only a short period by a single song thinking you've "cracked the code to audio nirvana".
I've learned to be patient over the years, no worries. No hurry as I'm not in a bad place. I try to participate in the journey of others to help out, which works out for me too as this gives me more data input to learn from and possibly new ideas.
Sometimes the hard part of that is to understand the goals of others, not trying to mix them up with my own weird ideas or ideals
.
The is what has been so great about this thread. You have shown us measuring is worth the hassle. But more importantly I have learned thru your example, what the measurements are actually showing us and how we can use them to optimize our systems.
Being impressed over a long long time frame is a lot harder than to be blown away for only a short period by a single song thinking you've "cracked the code to audio nirvana".
This is an accomplishment in itself. The two other systems I have built: Avebury and my four way open baffle, have their favorite recordings. (Although I am reworking the open baffle and have not tried DSP with them yet, so we well see if that takes them to a new level.) With the Cheap and Cheerful Arrays, I got closer to enjoying everything, although missing that "bit more" the other systems had on those "select" recordings. Hence, part of my reasoning for redoing the cabinets. The measurements showed Avebury had a cleaner Impulse response early on in the 0 - 3ms area, which I believe is due to the better cabinet, as it is placed in the same spot in the room as the arrays. It makes sense, as the cabinets of the arrays was an absolute hack build with lots of sharp corners. We will see how much better the new cabinets measure. We will also see how well the circular shape works too.
Either way, Know, Ronald, that You are a great inspiration on this forum.
Attachments
I'm hoping it will be temporary. I'm enjoying the listening time (immensely) and still run little tests every now and then. Things like playing with my specific ambient combination is easy, just a few turns of digital knobs. Overall I'm both happy and impressed with what I hear and have. Being impressed over a long long time frame is a lot harder than to be blown away for only a short period by a single song thinking you've "cracked the code to audio nirvana".
I've learned to be patient over the years, no worries. No hurry as I'm not in a bad place. I try to participate in the journey of others to help out, which works out for me too as this gives me more data input to learn from and possibly new ideas.
Sometimes the hard part of that is to understand the goals of others, not trying to mix them up with my own weird ideas or ideals
: ) new tendencies seen around and the high value of build information this thread have written a certain company if interested automatic deep product analysis and documentation down the road asking to fund 2x 30W/4558T00 shipped to Netherlands, maybe its not smart to ask for and even less smart write about it here but if against expectations happens then like a open honest style about stuff as that, and nothing happens by itself if never asked for.