I understand that, a reference that you can trust. I have a Coustic RTA33, one of the first digital fft analyzer that hit some sort of comm. success. But it has a SPL function and a dedicated spl mic to 140 dB, which I have never used, but its calibrated with the rta mic, and then i could control with the Umic and REW.
But do you measure pink noise or a burst or a sinus ? do I use any weighting to compensate? what about duration? is it peak or a average reading we use?
I have read the findings of a guy named Floyd Toole, on how to reproduce sound and how to measure a room, very interesting, especially the double blind testing. And the loudspeakers cookbook, thats it
But now i have discovered this forum, so I am good.Hello all,
This is the first time I have asked a question on a forum like this, and I must say the response blew me away. I am truly impressed by how accurate some of you explain my problem and a solution , with so little info provided by me about what it is and how its put together. So a big big thanks.
This was not any kind of test or anything like that, I was just not able to figure how to best solve this one.. so a real case, ongoing now. Now I considered it almost solved and will wrap up and list some actions based upon advises received.
But I should have stated some more info from the beginning, and please understand, it didnt start out with 10 dacs and 25 channels of amp with only active crossovers, it grew that way over time. And the VU meters does not peak out on 2000W all day, I seldom use more than 400W all togeather to get normale listening level of around 100 db.
But I was tired of having big expensive bulky speakers taking up Livingroom space, so one design criteria for this project was to impact the rooms as little as possible, then I wanted it to play realistic and loud, like a good set of large high end speakers, placed in every room, and cover the most important frequencies 30-15000 Hz, target is still 130 dB SPL between 30-15000 hz +/.3dB, but will never reach it, due to off axis tweeter response and not in the bedrooms.
Another design goal was bluetooth so friends could play their spotify etc.. and this brings out some level issues. And it is fully automatic turning on/off the system itself , (but not yet functions but it will alter configuration feed 7.1 movie vs 2 ch stereo) , it is controlled by Google home, and it can be triggered by alarm in the house automation, playing annoying noises at 120 db while flashing all the 60 bulbs in the house.... this part is also not finished yet.
But I took the time to put some pictures together with some explanation on, just the speakers , since they have been discussed.
Thanks a lot again for all the help
This is the first time I have asked a question on a forum like this, and I must say the response blew me away. I am truly impressed by how accurate some of you explain my problem and a solution , with so little info provided by me about what it is and how its put together. So a big big thanks.
This was not any kind of test or anything like that, I was just not able to figure how to best solve this one.. so a real case, ongoing now. Now I considered it almost solved and will wrap up and list some actions based upon advises received.
But I should have stated some more info from the beginning, and please understand, it didnt start out with 10 dacs and 25 channels of amp with only active crossovers, it grew that way over time. And the VU meters does not peak out on 2000W all day, I seldom use more than 400W all togeather to get normale listening level of around 100 db.
But I was tired of having big expensive bulky speakers taking up Livingroom space, so one design criteria for this project was to impact the rooms as little as possible, then I wanted it to play realistic and loud, like a good set of large high end speakers, placed in every room, and cover the most important frequencies 30-15000 Hz, target is still 130 dB SPL between 30-15000 hz +/.3dB, but will never reach it, due to off axis tweeter response and not in the bedrooms.
Another design goal was bluetooth so friends could play their spotify etc.. and this brings out some level issues. And it is fully automatic turning on/off the system itself , (but not yet functions but it will alter configuration feed 7.1 movie vs 2 ch stereo) , it is controlled by Google home, and it can be triggered by alarm in the house automation, playing annoying noises at 120 db while flashing all the 60 bulbs in the house...
. this part is also not finished yet. But I took the time to put some pictures together with some explanation on, just the speakers , since they have been discussed.
Thanks a lot again for all the help
Attachments
To hit these kind of SPL you need PA speakers and loudspeaker management with limiting (as you will probobly turn it up even louder as it will sound cleaner at louder volumes). Your trying to build a multi room club system from Hi-Fi components where maximum efficiency and power handling are not the main design goals. Try drivers from B&C, 18 Sound, Faital Pro,RCF etc. You might want to consider AMT for the treble if you are trying to build slim speakers as horns have some depth. https://www.eighteensound.it/en/products/hf-driver/6-5/16/AMT Air Motion Transformer AMT200P
Didnt end up with too much visible wire, but you are right..there is appox 300-400 meter speakercableBloody hell....when I see all these wires everywhere and read bulbes are flashing...
Must admit some dark evenings, regretting the hole thing, just wanted to quit and buy a normal system in a store.I attached a pic, the balanced cables for high/mid, run through the house, from the crossovers to the amps. and this is just half ot the signals. And I always pulled to few. Now finally I have a spare signal/multipurpose cable with 32 single conductors , going from the living room and through many rooms before ending up in the amp closet.
Attachments
I have a 9 amp in four speaker out system here... It's a single 3x4 metre room, but there's easily several hundreds of metres of wire and cable here.
I've blown my tweeters twice -- I no longer power those speakers with a Yorkville AP1200. My tubes can't pass enough power to damage them... Just my ears. Luckily, I found replacements on Kijiji - twice!
I've blown my tweeters twice -- I no longer power those speakers with a Yorkville AP1200. My tubes can't pass enough power to damage them... Just my ears. Luckily, I found replacements on Kijiji - twice!
Hello, I did some quick measurements to verify yours (now ours) theory's, to my problems and after testing I think have made the following mistakes:
a) I have had too much faith in high-end/good quality hifi drivers ability to dispatch all that heat, created by excessive power.
b) I have not paid attention to excursion limits and sensitivity when choosing drivers vs SPl.
(This was neglected due to experience, "if you feed each driver with an undistorted/unclipped signal, containing only the frequency's it was designed to reproduce, you are good". I still think there is truth to it, but only up to a certain level.
This creates one big + one minor problem, a) I am going almost broke + very time-consuming to buy and replace drivers every month b) I am struggling to reach design criteria, which was a overall spl of 130 dB say +/-6 db. Only the first one needs real fixing. After measure my speakers individually and sub , I agree with what has been posted, 130 dB is not needed for me at home.
To verify that this is caused by heat, and not just thin voice coil driver wire with to low current capability, I did experiments, with the voicecoils pulled out in its full length, so I had a 8 and 11 meter thin voiccoil wire hanging in the air. And with the voicecoils still intact on the woofer cone/dome.
Then I passed 100W continuous (here it is even ok to use RMS) power , then 150 W, them 1000 W and finally until it melted, while I measured the wire temp /with a oven meter.
In short I was surprised by the amount of power I could pass through the wires: Both coils could pass 100W without any noticeable rise in temp. At 150W the temp went quickly up around 5-6 deg C, but stabilized with no problems. At 1000W the temp rised quickly to approx 130 and 150 C and higher , both were intact after 45 seconds, passing the current, but the ones in the voiccoil shape was starting to overheat, burning the Kapton and would not have lasted 5-10 min. In the last video I applied 1800 W to the 5" speaker voicecoil and it overheats and actually burns up the hole thing within 5-10 seconds.
Some short reflections - I tried, while applying 100W to increase to 1000 W for a second and back down, multiple imes, this seems to go fine, the wires does not have time to increase enough in heat. So its not the peaks that kills them, its too high continuous power.
The results from the B&W 5" woofer and B&W 1" dome was almost consistent, tweeeter giving up maybe 20-30% before woofer.
Pls note that this kind of testng have some limitations and this is not done anywhere near "lab conditions".
PS! My upstairs tvroom is a carpentershop now due to working and didn't plan to share this video, so excuse the bad video, no time for any editing either.
brg Simen
PS Needed to zip them, but checked for virus before sending
a) I have had too much faith in high-end/good quality hifi drivers ability to dispatch all that heat, created by excessive power.
b) I have not paid attention to excursion limits and sensitivity when choosing drivers vs SPl.
(This was neglected due to experience, "if you feed each driver with an undistorted/unclipped signal, containing only the frequency's it was designed to reproduce, you are good". I still think there is truth to it, but only up to a certain level.
This creates one big + one minor problem, a) I am going almost broke + very time-consuming to buy and replace drivers every month b) I am struggling to reach design criteria, which was a overall spl of 130 dB say +/-6 db. Only the first one needs real fixing. After measure my speakers individually and sub , I agree with what has been posted, 130 dB is not needed for me at home.
To verify that this is caused by heat, and not just thin voice coil driver wire with to low current capability, I did experiments, with the voicecoils pulled out in its full length, so I had a 8 and 11 meter thin voiccoil wire hanging in the air. And with the voicecoils still intact on the woofer cone/dome.
Then I passed 100W continuous (here it is even ok to use RMS
) power , then 150 W, them 1000 W and finally until it melted, while I measured the wire temp /with a oven meter. In short I was surprised by the amount of power I could pass through the wires: Both coils could pass 100W without any noticeable rise in temp. At 150W the temp went quickly up around 5-6 deg C, but stabilized with no problems. At 1000W the temp rised quickly to approx 130 and 150 C and higher , both were intact after 45 seconds, passing the current, but the ones in the voiccoil shape was starting to overheat, burning the Kapton and would not have lasted 5-10 min. In the last video I applied 1800 W to the 5" speaker voicecoil and it overheats and actually burns up the hole thing within 5-10 seconds.
Some short reflections - I tried, while applying 100W to increase to 1000 W for a second and back down, multiple imes, this seems to go fine, the wires does not have time to increase enough in heat. So its not the peaks that kills them, its too high continuous power.
The results from the B&W 5" woofer and B&W 1" dome was almost consistent, tweeeter giving up maybe 20-30% before woofer.
Pls note that this kind of testng have some limitations and this is not done anywhere near "lab conditions".
PS! My upstairs tvroom is a carpentershop now due to working and didn't plan to share this video, so excuse the bad video, no time for any editing either.
brg Simen
PS Needed to zip them, but checked for virus before sending
I believe you said that you used B&W 686 as donors? Those are fairly far from high end.
All speakers made by B&W are high end
Joke aside, I use 686 as donors because they give me most value. But the woofer you will find in designs between 1800 Euro (new 686) to 18000 Euro pr pair.. High end or not... these are very good drivers. But Seas, DynAudio, or even Klipch and Harmon Kardon to name some US brands have a similar speaker in their range. As long as your design criteria's are correct, the brand is of less importance. Here are the 2 drivers I have used the most:Attachments
Considering adding "light bulb limiters" to the tweeters. That's a common protection method in pro speakers. I can attest that they work; I've seen them shining through the speaker ports when a DJ pushed the amp into full clipping. The inline ("festoon"?) incandescent bulbs used for car dome lights may be the right type, but I've gotten away with using the 8W filament of an 1157 bulb to protect an Eminence compression driver. And, definitely, passive crossovers for the mid/tweeter crossover.
Thanks for your reply. I have already added fuses as a intermediate solution, (after I blew 2 more tweeters yesterday trying to measure max SPL). Fuses was also part of the reason for today's experiment, to find the proper size. Both fuses and bulbs are probably well functioning "budget" solutions that will work well in many systems. But not here, problem is not clipping or short time peaks the drivers can not handle.
My therory now, is that they fail most likely due to the fact that the voice coil reaches 100-150 C, and stays that way for long period of time, which eventually causes a premature failure on the clear coat/isolation on the wire in the voice coil and glue that holds the voice coil to the cone. So to translate back to the bulb, in my case the bulbs on mid/high will light up all the time, while the bulbs on woofer etc will then only occasionally blink if clipping occured. The power consumed by the bulbs will reduce the output of the speaker causing non linearity at high volumes, so its back to drawing board
H
And yes - sorry to break any illusions you might have about the subject, most speaker manufacturer does that. And its defendable because of several reasons, a)it's not the driver that separates (as a example) a B&W Matrix 802 and pair of 607 same maker, even if they share many components, its the hours in the crossover design and 70 additional kilos of MDF etc to perfect the 802. b) They introduce a exclusive driver , which they have poured a lot of development into, first on their top line, then later, when paid for , and volumes go up and cost down, the same driver is used in budget models.
And this element does not always come with a yellow cone, that's just because this version was fitted with kevlar cone, you get the same speaker with carbonfibre, polypropylene or even coated paper cone, and you get it with shielded magnets or 5 screws instead of 4, stuff like that. Still same speaker.
And now I am discussing the small handful of speaker manufacturer who makes their own drivers, most do not, they buy them from a manufacturer finished. a 3500 US dollar hand made design speaker from italy, some years back used a 12 cm midrange from Seas, which costed at the time 20 US dollar at the factory (2 km from here) and good Danish brand used the same element in a budget center speaker.. But if you see on the attached picture, naming and other small variations is present, so they can be sold in different markets, apart from that these 2 elements are same/same and interchangeable, despite the various cone material.
brg Simen
My bad, was by a B&W selling store here a couple of weeks ago, to check new price and spare part prices. And i took this picture (attat.) of the new small bookshelf speaker offered now, and it was 11000.- pr pair not piece (equals to 1250 US), they have sold their speakers pr unit in the past, so I assumed it was pr speaker, but that was wrong ( but in my book, 1250 US for a pair of 5.25" or 6" woofers and 2 tweeters and a 6db filter in each end is pretty steep).
And yes - sorry to break any illusions you might have about the subject, most speaker manufacturer does that. And its defendable because of several reasons, a)it's not the driver that separates (as a example) a B&W Matrix 802 and pair of 607 same maker, even if they share many components, its the hours in the crossover design and 70 additional kilos of MDF etc to perfect the 802. b) They introduce a exclusive driver , which they have poured a lot of development into, first on their top line, then later, when paid for , and volumes go up and cost down, the same driver is used in budget models.
And this element does not always come with a yellow cone, that's just because this version was fitted with kevlar cone, you get the same speaker with carbonfibre, polypropylene or even coated paper cone, and you get it with shielded magnets or 5 screws instead of 4, stuff like that. Still same speaker.
And now I am discussing the small handful of speaker manufacturer who makes their own drivers, most do not, they buy them from a manufacturer finished. a 3500 US dollar hand made design speaker from italy, some years back used a 12 cm midrange from Seas, which costed at the time 20 US dollar at the factory (2 km from here) and good Danish brand used the same element in a budget center speaker.. But if you see on the attached picture, naming and other small variations is present, so they can be sold in different markets, apart from that these 2 elements are same/same and interchangeable, despite the various cone material.
brg Simen
Attachments
It's not clipping that makes the bulbs light, it's just too much power. Clipping simply reduces the dynamic range of the music, it doesn't add dangerous amounts of high frequencies (as our primitive cave-dwelling ancestors believed).
If you want over 120 dB, it's time to visit the pro audio department. Monoprice has some powered speakers that look too cheap to be any good, but maybe they're OK; I'd have considerably more confidence in Behringer.
If you want over 120 dB, it's time to visit the pro audio department. Monoprice has some powered speakers that look too cheap to be any good, but maybe they're OK; I'd have considerably more confidence in Behringer.
Thanks Yes, that seems to be the case, after reading all advices and a small experiment, I summed up my 2 biggest mistakes with these two sentences:
a) I have had too much faith in high-end/good quality hifi drivers ability to dispatch all that heat, created by excessive power.
b) I have not paid attention to excursion limits and sensitivity when choosing drivers vs SPl.
Luckily only a) needs real fixing. Suspicious to horns, but probably just due to lack of experience with them.
In my head it has been only of interest for people with radio tubes amplifiers, since they only have 35 W and the 6-9 dB gain is probably wanted
The speakers cabinets inside the wall are 25 cm in depth give or take, , one is 80 cm deep. But would really like to try the AMT200P or very similar, but they at least used to be super expensive, like 500-1000 Euro for a pair. This is what used to be called ribbon tweeter, or similar? 103 dB given 1 watt is substantial higher than today's 92 dB from the Seas 27TFFNC.
But is there a reason for not posting off axis response? Never heard of 18 Sound, but not very updated on development in PA the last 10 years either. But they understand what they are doing and selling, (remarks like: 2 hour test made with continuous pink noise signal from recommended crossover frequency to 20 kHz. 3. Power on Continuous Program is defined as 3 dB greater than the Nominal rating. 4.12 dB/oct. or higher slope high-pass filter etc give them away), so why not add 30 and 60 degree off axis...
I agree with what has been posted, 130 dB is not needed for me at home. I measured today and the stations (satelites) was fluxing around 115, 116, 117, with 119 as highest, playing music at -3Db FS with low distortion. The sub of course way more, easily 126 dB, playing music, allowing for 1-2% amplifier distortion, but it needs to cover a larger area also. Turned off all weighting, mic on a tripod 1.5 meter away from speaker.. is that correct?
Attaching photo , look at what i had to do with the tweeters, ugly as... But reason for question above. Also added 2x10 sec video on how i measured, if you have the time
