Matt
If I get something that I don't think is correct, the first thing that I do is to measure the transfer function of the crossover, because most of the time that is the problem. Can you do that and send me the data. I can tell you right away of the crossover is working right. Make sure and load the crossover with a 4 ohm (or any value it doen't matter) resistor, not the driver. And its a good idea to disconnect the one not being measured.
The bump at 100 Hz certainly doesn't make sense.
I wouldn't unwind anything from an inductor until you are sure, because you can't rewind it back on if it gets worse.
If I get something that I don't think is correct, the first thing that I do is to measure the transfer function of the crossover, because most of the time that is the problem. Can you do that and send me the data. I can tell you right away of the crossover is working right. Make sure and load the crossover with a 4 ohm (or any value it doen't matter) resistor, not the driver. And its a good idea to disconnect the one not being measured.
The bump at 100 Hz certainly doesn't make sense.
I wouldn't unwind anything from an inductor until you are sure, because you can't rewind it back on if it gets worse.
Hi Dr. Geddes, I figured out all the problems, no worries. Thanks for the offer though. It turned out to be a mixture of problems with both the crossover, location, subwoofers, and pre-pro.
I unhooked everything, and took a better look at potential trouble area's in the crossover. The tweeter problem was a bad resistor and capacitor, as it turned out. I replaced those with new parts, and got that working correctly. Only odd dip is the 800hz one, which you said is normal.
The 100hz bump turned out to be a bass tone control I thought I had defeated, but actually had turned on full. The speaker was being processed with a 100hz highpass I forgot was on, the bass control on, and a messed up crossover, all causing extra problems. Once I traced all these problems down, I was able to get a better measurement.
Before you all think I'm a complete moron. When I moved everything I had to turn off and unplug all my components. The pre/pro uses a 1 farad cap to store the memory for settings, and everything was off long enough to cause it to reset to factory defaults. I knew this would happen, but when I took measurements, I forgot to check all of them, as I normally used the 100hz highpass, it wasn't part of my setup habit to turn that off.
I unhooked everything, and took a better look at potential trouble area's in the crossover. The tweeter problem was a bad resistor and capacitor, as it turned out. I replaced those with new parts, and got that working correctly. Only odd dip is the 800hz one, which you said is normal.
The 100hz bump turned out to be a bass tone control I thought I had defeated, but actually had turned on full. The speaker was being processed with a 100hz highpass I forgot was on, the bass control on, and a messed up crossover, all causing extra problems. Once I traced all these problems down, I was able to get a better measurement.
Before you all think I'm a complete moron. When I moved everything I had to turn off and unplug all my components. The pre/pro uses a 1 farad cap to store the memory for settings, and everything was off long enough to cause it to reset to factory defaults. I knew this would happen, but when I took measurements, I forgot to check all of them, as I normally used the 100hz highpass, it wasn't part of my setup habit to turn that off.
Nice paper from Troels 😉
Shows what happens with inductor values when cramped into small space
More than 2 coils easily leads to problems, if layout is wrong
http://www.troelsgravesen.dk/coils.htm
Shows what happens with inductor values when cramped into small space
More than 2 coils easily leads to problems, if layout is wrong
http://www.troelsgravesen.dk/coils.htm
Thanks Tinitus, but layout is fine. Non inductor is placed within 2-3 inches of another, and all are oriented in the correct recommended fashion.
Ok here is a picture of the Center's response taken as close to anechoic as I can get. I did bass measurement at 1" and spliced in around 375hz, as can be seen, using a special correction file designed for that. The response above that was taken mostly in my garage at 2 meters distance and 4 feet off the ground with my room acoustic treatments around key areas, such as the floor. I spliced in above 1.5khz measurements taken at 1.5 meaters in my makeshift anechoic chamber/closet I normally use for driver measurements. The chamber was designed per a recomendation from a B&K engineer for driver measurements, and is similar but larger to what is used for qc testing of driver production, so I was told. It's 10 feet wide, 6 feet deep, and 12 feet tall with an angled ceiling. The walls are lined with 8" of acoustic cotton batting like is sold for acoustic treatments, with 6" of the higher density rigid batting. The ceiling has the ability to hang some as well. I placed the speaker on it's back facing up, sitting around 12" or so off the ground, but only maybe 4 from the acoustic material. The mic is suspended above the speaker on axis using a boom mic stand. The door going into the closet isn't really acousticly treated, it's too cheap to treat and not wreck, so that may be a problem, but the measurement is pretty smooth overall.
The smoothing is 1/3, but the measurement was taken with no smoothing, it's the splicing three times that causes the smoothing to be so much. I think this shows a very smooth and accurate response.
Ok here is a picture of the Center's response taken as close to anechoic as I can get. I did bass measurement at 1" and spliced in around 375hz, as can be seen, using a special correction file designed for that. The response above that was taken mostly in my garage at 2 meters distance and 4 feet off the ground with my room acoustic treatments around key areas, such as the floor. I spliced in above 1.5khz measurements taken at 1.5 meaters in my makeshift anechoic chamber/closet I normally use for driver measurements. The chamber was designed per a recomendation from a B&K engineer for driver measurements, and is similar but larger to what is used for qc testing of driver production, so I was told. It's 10 feet wide, 6 feet deep, and 12 feet tall with an angled ceiling. The walls are lined with 8" of acoustic cotton batting like is sold for acoustic treatments, with 6" of the higher density rigid batting. The ceiling has the ability to hang some as well. I placed the speaker on it's back facing up, sitting around 12" or so off the ground, but only maybe 4 from the acoustic material. The mic is suspended above the speaker on axis using a boom mic stand. The door going into the closet isn't really acousticly treated, it's too cheap to treat and not wreck, so that may be a problem, but the measurement is pretty smooth overall.
The smoothing is 1/3, but the measurement was taken with no smoothing, it's the splicing three times that causes the smoothing to be so much. I think this shows a very smooth and accurate response.
For those of you looking for unsmoothed raw response data, here you go. I used an averaging method taken within a 1 cubic foot area or so to get this, with the speaker in it's normal location, taken at a distance of 9 feet. The decibels are corrected for distance to reflect 1 meter 1 watt type level. The speaker is basically sitting on the floor, so it looks like I was wrong about the cause of the 100hz bump, it's partly caused by the location. Since it doesn't show up in freefield measurements, I"m confident it's location related. In fact, looking at the associated acoustic phase data, it's clear there are room effects causing issues. At the exact location f that large dip at 133 hz is a complete phase reversal. Following the that is the rise below 133hz down to about 90hz, which is associated with a reversed phase from the rest of the response above that.
I can post phase data, but I find nobody ever reads acoustic phase data taken by the ATB correctly. It shows a good smooth change at the crossover point where it goes from roughly 0 to +45 and then declines to 0 again, smoothly, by around 8-10khz. After that phase gets really messy with up/down phase issues. In all likelihood this is caused by a mixture of the waveguide, compression driver, screen, waveguide, and/or foam. However, and Dr. Geddes can correct me if he feels I'm wrong, I don't see how anything that high will matter much, and I think correcting it would be very difficult.
Hey Dr. Geddes, as for future center channel modifications, what about an LCR with a very broad Q to bring down the 2-5khz range a few decibels, but then reducing the series resistor in the L-pad. Then reduce the bass inductor to something lower, say around 2-2.5mh to account for the boundary effect of having it mounted against the wall. These are guesstimate numbers based on the response I'm seeing, as a way to improve things. I haven't actually tried mounting it on a shelf directly against the wall yet, so it's possible I wont see as much of a response boost as I think, but given what I'm seeing with it 4" from the rear wall, with a sub cabinet and component cabinet sandwiching it, I think I will see further boost around that 150hz on down area.
Oh yeah, FYI, I turned all subwoofers off for these measurements. Going with an average spl of say 93db, the minus 10db point is 45hz, but with a room mode causing a boost at 21hz giving that a peak to 86db's. I'll get better measurements of the bass response of the room later with all subs on and all front speakers working full range. I expect the -3db point in my room to be below human audibility. I.e. well below 20hz. In fact, while I don't trust my mic's, I am measuring a lot og energy easily down into the 10hz range.
Informative. So now everything is checked and all the parts are right. Lets wait for the room response so to see if the 2-4k range integrates better. Because if it stays up, there will be an ''up front'' feeling.
Here is the shown from on axis to roughly 60 degrees off axis in around 10 degree steps. I didn't do this as carefully as I should have, I don't like lifting these speakers, so I didn't use my turntable, just moved them by hand using a triangle for angle measurements. It gives you a rough idea of the dispersion control. I see directivity control starting around 700hz or so, with a smooth response change until around 60 degrees when things start to change a bit. It's the best I've ever seen or measured.
Salas the last measurement I posted was an on-axis measurement in room taken at 9 feet. It wasn't quite the listening position, but it was in room on axis.
This is missing one of the LCR's, but that was a recommendation given by Dr. Geddes for better axial response. Maybe it deals with the peak there, I don't know. It's not huge, I think it's made to look worse because of the woofers response dip at 800hz, the 150hz bass rise from location, and the natural roll off of being in a well treated room.
pjpoes said:
Matt
That region is usually controlled with the tweeter sensitivity as either end is then controlled with other components. I'll look at that some when I do further measurements in the coming weeks.
Matt
Below 200 Hz those curves should lay right on top of one-another. And I would have expected a greater difference between the curves with angle. I suspect that there is still a lot of the room in this data.
Below 200 Hz those curves should lay right on top of one-another. And I would have expected a greater difference between the curves with angle. I suspect that there is still a lot of the room in this data.
THere is lots of room in the data, sorry. I should have better specified what I meant when I said, I didn't do this as carefully as I should. I should have done the measurements outside on the turntable some distance from the ground. Better than what I did would have been in my garage. What I did was in the middle of my living room, so there is Lots of stuff corrupting that data. I just throught it was better than nothing. However, comparing it with your data, I'm seeing a shift in the treble response to be 10 decibels down at almost the exact same place as you do. My last measurement was roughly 60 degrees, and you have one as 52.5 showing around the -10db point. I think it's more similar than you think. I didn't go to 90 degrees, so I'm sure you would see more and more decline.
I can't export that data you see to my other programs without saving each individual measurement, and importing them one by one. That would also be the only way to average all of those measurements.
As long as you guys accept the limitations of this measurement, I can show one that would roughly equal the averaged 60 degree response. Easily, the only way to do will be to move the speaker onto a pad on the floor, and manually turn it while the ATB software averages it. Let me see how that looks, I'm not sure if the pad rubbing will add a bass artifact.
I can't export that data you see to my other programs without saving each individual measurement, and importing them one by one. That would also be the only way to average all of those measurements.
As long as you guys accept the limitations of this measurement, I can show one that would roughly equal the averaged 60 degree response. Easily, the only way to do will be to move the speaker onto a pad on the floor, and manually turn it while the ATB software averages it. Let me see how that looks, I'm not sure if the pad rubbing will add a bass artifact.
This is the closest thing I could come up with to a response averaged across a window of roughly 110 degrees total. I.E. I went from 55 degrees off axis then rotated the speaker through to axial, then on through to 55 degrees off axis the other way, then back to axial, with the ATB software averaging the response the entire time. I cut it at 200hz, so that is 200hz to 22khz.
I did this in room on the floor, so again, lots of things corrupting that data. No idea if it shows anything useful, I can try better methods later on. These speakers are just too heavy for me to keep carrying around, so this is what you get for now.
The peak at 4-5 kHz is exactly what the LCR that you took out brings down. Adding it back will smooth this "average" curve, but will result in a slight dip directly on axis. The choice is yours.
Good to know my measurements are relatively consistent with yours.
This is the current setup. As I said, its temporary until I figure out what want to do long term. There is a small 3" bookshelf surround I built sitting next to it for size comparison.
This is the current setup. As I said, its temporary until I figure out what want to do long term. There is a small 3" bookshelf surround I built sitting next to it for size comparison.
OUCH! All that trouble getting rid of the diffraction from the speaker cabinets only to do it with the equipment and subs! I think one reason that I always get a slightly better rating on my systems here is that there is NOTHING up front near the speakers to diffract the sound. Its as bare as a babies bottom (or is that child porn and I should say a hook--s bottom?!).
yeah I know, I said it was temporary. There is a mixture of problems making it so I can't move those items. One of which is that the subwoofer is extremely heavy. Over 200 lbs with driver and amp. The other is that I simply don't have cables long enough to move the console, and the area where I am considering moving it doesn't have enough outlets.
I have to ask...
What is the little guy on top of the subwoofer? Is that a paperweight?
😀 just kidding with you.
Thanks for sharing the pic.
How does running in phantom center compare with actual center for you?
-Tony
What is the little guy on top of the subwoofer? Is that a paperweight?
😀 just kidding with you.
Thanks for sharing the pic.
How does running in phantom center compare with actual center for you?
-Tony
Don't know yet, I will let you know later.
I mean, comparing music played using DTS Music vs stereo, the soundstage is deeper and more consistent, but the center image is maybe a bit more vague. I tend to prefer music in plain stereo mode though, so others might prefer the other way. It sounds surprisingly good for such a bed center setup.
I mean, comparing music played using DTS Music vs stereo, the soundstage is deeper and more consistent, but the center image is maybe a bit more vague. I tend to prefer music in plain stereo mode though, so others might prefer the other way. It sounds surprisingly good for such a bed center setup.
markus76 said:
That might help the speaker - it would be outrageously expensive - but it won't help the rest of the room.