The light bulb trick works pretty good actually. I would do that in car audio installs some times. I would use those festoon type light bulbs that look like fuses because they where easy to wire and worked pretty good. There also is another option for that. I forget what they are called but they are part of a circuit that under a certain temp they let all the current flow through unhindered but when the temp hits a set point it shunts the current through a resister which allows the music to keep playing just at a lower level until the temp comes down.
Real good idea for everybody (and for all sorts of mishaps tweeters are prone to), esp with needlessly power amps driving tweeters. Likely much damage to tweeters from absence of that kind of protection.
But trick is to find membrane caps with adequate capacitance and avoid being tempted by electrolytic ones.
B.
There are multiple things going on here that need addressing.
As a person who likes to listen loud most of the time (mainly to minimally compressed music while having the entire dynamic range at my disposal), I consider 110 - 115 dB C weighted music peaks to be VERY loud. This SPL range can be considered average live concert volume levels with rather heavily compressed rock/pop music. The issue is how much compression there is and where the main concentration of energy is relative to the entire frequency range. Rock, pop, hip-hop, edm, metal etc all have different concentrations of energy in various frequency ranges, which can weigh the power distribution differently to the individual drivers as dictated by the crossover.
Based on this, you can have a typical 10 - 15W rms tweeter live happily playing along in a 2 way system with a 6-8" woofer, able to play most types of music for several hours at around 100 - 103 dB levels without harm (excluding your hearing). A tweeter crossed at 2500 Hz or so with a 2nd order HP will only see about 15 % of the total music power bandwidth when being fed most styles of rock, pop, jazz and classical music. This whole situation changes with heavily compressed bass heavy edm and guitar driven, treble laden metal/hardcore/punk. The tweeter is usually the first to suffer a thermal VC meltdown with this type of music and its heavy concentration of treble energy (sometimes 2x or more the percentage power distribution).
The other issue is overloading a driver when its being driven hard around its resonant frequency. The diaphragm on a dome or compressiom driver can go into an unstable rocking mode allowing the VC to rub on the upper pole plate. Most cone drivers can suffer a similar fait when driven hard around its resonance and/or if the air load placed on the cone is uneven ie. by unsymmetrical port or LF driver placement. This can make the VC rub and possibly fail as well. At a minimum it can cause excessive audible distortion.
Power ratings are heavily exaggerated by some manufacturers and can be based either on a driver's actual failure point rather than surviving for several hours of repetitive abuse (more realistic and practical). I generally dont consider manufacturers published power handling specs unless its a high SPL system overbuilt to survive constant use and not fail, even with a bit of extra headroom factored in for reliability sake.
Always consider VC diameter and former materials when judging power handling. VCs with aluminum formers handle more power than kapton/polyimide/nomex thanks to its better thermal conductivity. Fiberglass formers are also slightly better then kapton but not as good as aluminum. Copper VC windings are more durable than aluminum or copper clad and less prone to power compression as well as thermal fatigue. IMO, realistic continuous thermally limited power levels for most LF/MF drivers with varying diameter VCs are:
25mm/1" VC - 35 to 75w
38mm/1.5" VC - 75 to 125w
50mm/2.0" VC - 100 to 175w
65mm/2.5" VC - 150 to 250w+
75mm/3.0" VC - 250 to 350w+
100mm/4.0" VC - 300 to 500w+
Continuous power ratings for HF and compression drivers are about 30-40 percent less.
As a person who likes to listen loud most of the time (mainly to minimally compressed music while having the entire dynamic range at my disposal), I consider 110 - 115 dB C weighted music peaks to be VERY loud. This SPL range can be considered average live concert volume levels with rather heavily compressed rock/pop music. The issue is how much compression there is and where the main concentration of energy is relative to the entire frequency range. Rock, pop, hip-hop, edm, metal etc all have different concentrations of energy in various frequency ranges, which can weigh the power distribution differently to the individual drivers as dictated by the crossover.
Based on this, you can have a typical 10 - 15W rms tweeter live happily playing along in a 2 way system with a 6-8" woofer, able to play most types of music for several hours at around 100 - 103 dB levels without harm (excluding your hearing). A tweeter crossed at 2500 Hz or so with a 2nd order HP will only see about 15 % of the total music power bandwidth when being fed most styles of rock, pop, jazz and classical music. This whole situation changes with heavily compressed bass heavy edm and guitar driven, treble laden metal/hardcore/punk. The tweeter is usually the first to suffer a thermal VC meltdown with this type of music and its heavy concentration of treble energy (sometimes 2x or more the percentage power distribution).
The other issue is overloading a driver when its being driven hard around its resonant frequency. The diaphragm on a dome or compressiom driver can go into an unstable rocking mode allowing the VC to rub on the upper pole plate. Most cone drivers can suffer a similar fait when driven hard around its resonance and/or if the air load placed on the cone is uneven ie. by unsymmetrical port or LF driver placement. This can make the VC rub and possibly fail as well. At a minimum it can cause excessive audible distortion.
Power ratings are heavily exaggerated by some manufacturers and can be based either on a driver's actual failure point rather than surviving for several hours of repetitive abuse (more realistic and practical). I generally dont consider manufacturers published power handling specs unless its a high SPL system overbuilt to survive constant use and not fail, even with a bit of extra headroom factored in for reliability sake.
Always consider VC diameter and former materials when judging power handling. VCs with aluminum formers handle more power than kapton/polyimide/nomex thanks to its better thermal conductivity. Fiberglass formers are also slightly better then kapton but not as good as aluminum. Copper VC windings are more durable than aluminum or copper clad and less prone to power compression as well as thermal fatigue. IMO, realistic continuous thermally limited power levels for most LF/MF drivers with varying diameter VCs are:
25mm/1" VC - 35 to 75w
38mm/1.5" VC - 75 to 125w
50mm/2.0" VC - 100 to 175w
65mm/2.5" VC - 150 to 250w+
75mm/3.0" VC - 250 to 350w+
100mm/4.0" VC - 300 to 500w+
Continuous power ratings for HF and compression drivers are about 30-40 percent less.
The OP has not posted again, which is happening sometimes.
I would advise a hearing test, and go from there.
Here we have a saying, that (foolish person), wants to buy a buffalo to eat his grass.
My original suggestion was to put smaller amps, but he wanted to keep the ones he got a good deal on...that was a minimal expense and effort solution.
He says got a good deal on two cinema amps, then mentions three.
The amps actually in use, and the connections are not mentioned.
What happens when you connect ten resistors in parallel?
Ohm's Law, figure out what happens when you put ten speakers in parallel. Without crossovers or any other device except the speaker coils.
He is either trolling us, being super intelligent, or a total deluded person of low intelligence.
And the more experienced people here, what kind of sub is needed to handle 2 kW? How big, and how many?
I would advise a hearing test, and go from there.
Here we have a saying, that (foolish person), wants to buy a buffalo to eat his grass.
My original suggestion was to put smaller amps, but he wanted to keep the ones he got a good deal on...that was a minimal expense and effort solution.
He says got a good deal on two cinema amps, then mentions three.
The amps actually in use, and the connections are not mentioned.
What happens when you connect ten resistors in parallel?
Ohm's Law, figure out what happens when you put ten speakers in parallel. Without crossovers or any other device except the speaker coils.
He is either trolling us, being super intelligent, or a total deluded person of low intelligence.
And the more experienced people here, what kind of sub is needed to handle 2 kW? How big, and how many?
Last edited:
Coupling caps ( as I mentioned in post 27) for use with active crossovers are usually large enough to be so far from the crossover as not to be in the equation. If course, nothing wrong with hybrid, first order to the drivers and the rest by DSP.
21 inches or so, one is enough.
They have not specified max. dB sound level though.
And I still wonder how big is the dwelling, and bad is his hearing loss.
I run my bedroom/ computer room at about 4 watts into two speakers, more than that is rare, though I am about 30 feet from the next neighbor.
The amp is good to 20 w/ch/8R RIAA, it is a Philips with Matsushita chips. But at high levels, it gets tiring.
They have not specified max. dB sound level though.
And I still wonder how big is the dwelling, and bad is his hearing loss.
I run my bedroom/ computer room at about 4 watts into two speakers, more than that is rare, though I am about 30 feet from the next neighbor.
The amp is good to 20 w/ch/8R RIAA, it is a Philips with Matsushita chips. But at high levels, it gets tiring.
Max SPL depends on several things.
The 21” driver above is rated 2000 watt. This is +33dB theoretically. It’s sensitivity at 97dB then tells us it could potentially reach 130dB at certain frequencies assuming xmax and resonances is avoided mechanically. Box design, subsonic filter and the music program are even more crucial in realistic levels. Distortion in the amp too… Not to mention room modes and placement. The power rating and sensitivity only gives you a hint really.
Thanks a lot digitalthor, I think this
. Will be implemented asap as a cost reducing action. However they are somewhat unpredictable, (fuses that is, so will continue to investigate, and read advices until solved. Ref your other rational suggestions for problem areas: a) yes you are correct, to set correct gains takes one full day:-/ and is a pain in the b..t to be honest, first outputlevel on the 10 channel interface (dacs), then output on 14 channels of preamplifier (line drivers) and finally input sensitivity on 21 channels of amplification (some is bridged). b) But when done...matching crossover hz/slopes and levels to driver sensitivity/location is quick and easy. c) Speaker wire "std copper electrician stuff" 1.5 sqmm for mid/tweeter on short runs, 2.5 sqmm for long runs and all midbass, 4 sqmm for subbass, all double and tripple checked, and found good. d) no clipping anywhere in the chain, below -6dB FS. But clip will show up on VU meters) brg Simen
Agree, this looks like a very capable tweeter
Damn... that's a persistent way of killing drivers
A typical dome tweeter can't handle this amount of abuse, you need a horn with a compression driver and a pro midrange - maybe also in a horn. It's not the average power but more likely the peaks that kill them - and because you play seriously LOUD for longer periods of time.
Build a horn loaded speaker, with a bigger midrange, compression tweeter and good directivity with pro drivers that can take this type of violent beating - a copy of some cinema pro speaker like this:
If this Celestion 10" only plays from 100hz and up, and you add your subs - then it should be fine. Or else build two more speakers or chose a driver that can handle even more - others have to chime in here... because I would not at the moment know of a proper upgrade. Possibly even the compression driver could be different, and take even more power.
Pro drivers can simply take tons more abuse before they burn, since they are mostly designed to play all night at high levels - which most hifi drivers are not - especially not your levels
Hello iamjacklope, thanks a lot, been building home and car audio systems for nearly three decades, and still not clever enough to remember to fuse the speaker inputs
. Will be implemented asap as a cost reducing action. However they are somewhat unpredictable, (fuses that is, so will continue to investigate, and read advices until solved. Ref your other rational suggestions for problem areas: a) yes you are correct, to set correct gains takes one full day:-/ and is a pain in the b..t to be honest, first outputlevel on the 10 channel interface (dacs), then output on 14 channels of preamplifier (line drivers) and finally input sensitivity on 21 channels of amplification (some is bridged). b) But when done...matching crossover hz/slopes and levels to driver sensitivity/location is quick and easy. c) Speaker wire "std copper electrician stuff" 1.5 sqmm for mid/tweeter on short runs, 2.5 sqmm for long runs and all midbass, 4 sqmm for subbass, all double and tripple checked, and found good. d) no clipping anywhere in the chain, below -6dB FS. But clip will show up on VU meters) brg Simenyes good call
.... first version of my wall speaker Mk1 had dual 6.5" midrange drivers, because i was worried for low output in midbass, it gave me 3dB more soundpress. from the added speaker, and 3dB from the double amount of power (2x300W vs 1x300w). Reason for abandoning the design was that having double set of speaker creates often a new set of problems, (phase changes and double amount of reflections), which happened and another reason was that even feed 90 Volt in 5 ohm, they could not produce the needed level in the 80-250hz range, so it made sense to reduce the speaker count with one midrange and just give the other some more power.. and add 2x8" or 2x6,5 and also 1x10" taking care of 80-250 Hz.Hello thanks, yes persistent. expensive and time consuming. I am buying 2-3 pairs at the time of the Bowers & Wilkins 868 speaker, used on Finn.no (like craigslist? in Norway) I have stopped ordering the tweeter and mid as replacements from B&WDamn... that's a persistent way of killing drivers
A typical dome tweeter can't handle this amount of abuse, you need a horn with a compression driver and a pro midrange - maybe also in a horn. It's not the average power but more likely the peaks that kill them - and because you play seriously LOUD for longer periods of time.
Build a horn loaded speaker, with a bigger midrange, compression tweeter and good directivity with pro drivers that can take this type of violent beating - a copy of some cinema pro speaker like this:
If this Celestion 10" only plays from 100hz and up, and you add your subs - then it should be fine. Or else build two more speakers or chose a driver that can handle even more - others have to chime in here... because I would not at the moment know of a proper upgrade. Possibly even the compression driver could be different, and take even more power.
Pro drivers can simply take tons more abuse before they burn, since they are mostly designed to play all night at high levels - which most hifi drivers are not - especially not your levels
at 120 Euro a piece. But what you write sounds like the best solution. Find a tweeter and midrange that can "handle this amount of abuse" since the rest works great. The main problem is the mids approx. 250-3000hz. that keep burning up... so yes most likely a then a 6-8" PA driver, never touched one in my life, but a large alu voice coil + pole venting is more important than linearity in frequency and off axis, those i can fix, but not the burning
. If i need a horn tweeter with a driver, I will build it, but there is some recommendations for a morel tweeter with some very good specs for my purpose, just some post up, so maybe I do not have to do..
Introducing a non linear element in line with a driver is not the best for sound quality, but it may save some drivers. BTW my personal experience is that in all the years I have driven the high end from overpowered amps in active systems, never had a blown one. The only tweeters I have blown were the concentric ones in the small but amazing Altecs, but that is almost half a century ago. Passive of course.
Hello
The kind of level matching you are talking about above is seldom demanding, (unless you have passive crossovers and uses resistors to adjust levels). You choose a 50 W amp for tweeters, a 200W for mid and 500W for sub, (example values) so only minor adjustment is needed. 5 min listening & adjust, entering recommended crossover settings as a start, and find where the system distorts or clips first, thats is your max, reduce the rest to achieve linearity.
Then you average a series with sweep measurements captured from most used listening positions, and calculate the final level/eq settings/bandwiths fine tune crossover slopes) in REW or similar to reach the +/-3db or whatever you design goal is..
The problem is that the recorded level on your music/input source varies, sometimes with 20 dB.. Which results in Dr.Dre often being played back at high volume even when set to low, and you can not get enough volume to hear the quiet passage of a Beethoven symphony. This I suggested to fix by increasing the gain to have some xtra volume to boost low levels in the material, and "exercise some control on the volume nob" to avoid blowing speakers playing -0dBFS brg Simon
Hello again, maybe, or we are just talking past each other. With gain setting I was referring to the "one time" matching of output level (voltage) of the preamp or other music source, to the input circuit of the amplifier, this is usually done at installation. On car audio and PA amps done by turning a pot meter which regulates the amount of gain the amplifier need to add or decrease to the signal to ensure it will work with the huge range of signals from various brands, from -10 dBV but in some extreme and good examples also +20 dBV, (that is a voltage difference maybe from 0.4 Volt to 8 Volt ).. If you have your gain set too low, or your input signal just is to low , the amplifier will not be able to reach full power and set to high, it will clip without using the full scale, add noise and more..
The kind of level matching you are talking about above is seldom demanding, (unless you have passive crossovers and uses resistors to adjust levels). You choose a 50 W amp for tweeters, a 200W for mid and 500W for sub, (example values) so only minor adjustment is needed. 5 min listening & adjust, entering recommended crossover settings as a start, and find where the system distorts or clips first, thats is your max, reduce the rest to achieve linearity.
Then you average a series with sweep measurements captured from most used listening positions, and calculate the final level/eq settings/bandwiths fine tune crossover slopes) in REW or similar to reach the +/-3db or whatever you design goal is..
The problem is that the recorded level on your music/input source varies, sometimes with 20 dB.. Which results in Dr.Dre often being played back at high volume even when set to low, and you can not get enough volume to hear the quiet passage of a Beethoven symphony. This I suggested to fix by increasing the gain to have some xtra volume to boost low levels in the material, and "exercise some control on the volume nob" to avoid blowing speakers playing -0dBFS brg Simon
I can't offer any useful advice on your current sound system but I can guarantee that this will be the sound system of your near future:
They come with a convenient charging case and the fidelity isn't too bad...
not like healthy working ears but yours seem destined for a pair of these. I've had them for about a year and they're the best I have tried so far. Based on what you must have spent on your current system I think the price will be well within your comfort zone!
They come with a convenient charging case and the fidelity isn't too bad...
not like healthy working ears but yours seem destined for a pair of these. I've had them for about a year and they're the best I have tried so far. Based on what you must have spent on your current system I think the price will be well within your comfort zone!
I have the good fortune of having a wonderful niece who is an audiologist, if you don't I think they are in the $4K-$5K neighborhood.