Seems the few on any speaker forum that care about distortion are in a minority.
Although pink noise is closer to the dynamics of "normal" music than sine wave testing, I'm not familiar of distortion testing using a signal other than sine waves.
I use Smaart, single sine wave tones, using the RTA to see the added harmonics. All my posted tests have been outdoors.
I would like to test at 10 meters, unfortunately 2 meters gives more accurate results due to distance to buildings.
Generally have used 5 Hz spaced tones near the Fc or Fb, then 10 Hz at higher frequencies.
You are correct, at frequencies below Fb, distortion in the ported or TH Lab 12 rises dramatically.
As far as 100% distortion at 400 watts, yes, if you push the Lab 12 hard well below Fb, it will distort.
It is common knowledge that ported cabinets must be high passed below Fb to avoid over excursion, I did not try to destroy my drivers using a sine wave signal (about 9 dB more average power than usual music signals) to achieve that goal.
If you use a high pass filter just below the Fb or Fc (as anyone using high power on a sub should) distortion on the Lab 12 will be low below Xmax.
Subtract 6dB for being outside:
A 35hz TB 6.5" design would be about 30% smaller. Here is a shoebox sized 40hz design (-5dB at 35hz), subtract 12dB for being outside.
A 35hz TB 6.5" design would be about 30% smaller. Here is a shoebox sized 40hz design (-5dB at 35hz), subtract 12dB for being outside.
Yes, but not 6-7% at 400 watts. It will be at that level down at moderate power levels, but at 400 watts, it is much higher than that all through the passband. Please check your results.
I like using stepped sines for checking distortion. It's easy to get a THD+N measurement, by simply bandstopping out the fundamental. Or you can use a passband filter at 2x, 3x, 4x, etc. to get all the individual harmonics and then add them together for THD.
At high power levels, the microphone can be overdriven, so you kind of need some distance. But then again, the SPL is high enough the noise floor goes down. You can always measure the noise floor by simply measuring a "sweep" where no signal is presented to the speaker. Knowing the noise floor level, you can know where measured values are valid. If the SPL is above the noise floor, it's valid; Once below the noise floor, it's not.
some interesting responses.
that ripole sub looks interesting, and the TH gets down low.
is there a basic rule of thumb technique for folding a horn once hornresp spits out the long straight horn image?
is the expansion rate the only critical detail, and is it (the expansion rate) the same on all THs?
looks to me like about 30:1 at a guess.....
that ripole sub looks interesting, and the TH gets down low.
is there a basic rule of thumb technique for folding a horn once hornresp spits out the long straight horn image?
is the expansion rate the only critical detail, and is it (the expansion rate) the same on all THs?
looks to me like about 30:1 at a guess.....
Last edited:
Tune'able TH?
call me crazy, but d'ya think it might be possible to build a tune'able TH?
im thinking the driver and first 7% of the horn could be built parallel not tapered, and have the ability to make the length longer or shorter by 10% each way??
that way the sub output/rear wave could be tuned/corrected if there was a flaw in the design or build, or just to taste (flat/weighted)
am i wrong?
call me crazy, but d'ya think it might be possible to build a tune'able TH?
im thinking the driver and first 7% of the horn could be built parallel not tapered, and have the ability to make the length longer or shorter by 10% each way??
that way the sub output/rear wave could be tuned/corrected if there was a flaw in the design or build, or just to taste (flat/weighted)
am i wrong?
Last edited:
Hi 60ndown,
In a tapped horn you have a lot of interactions between the various lengths and cross-sections. E.g.: When you move the driver position closer to S1 you decrease L12, and will probably also decrease L45, while at the same time increasing L23/L34. If you do that in a given cabinet you will also be decreasing S2 and S4 (this is for a 4-section TH). I'm not saying, that some tuning mechanism cannot be build, but it gets complicated in a hurry. A driver baffle that can be swapped around may be of interest, and in Art Welter's Keystone Sub you have a wide tuning range by changing the Keystone exit. By the way, I added a drawing for the Keystone Sub to Art's thread.
Regards,
In a tapped horn you have a lot of interactions between the various lengths and cross-sections. E.g.: When you move the driver position closer to S1 you decrease L12, and will probably also decrease L45, while at the same time increasing L23/L34. If you do that in a given cabinet you will also be decreasing S2 and S4 (this is for a 4-section TH). I'm not saying, that some tuning mechanism cannot be build, but it gets complicated in a hurry. A driver baffle that can be swapped around may be of interest, and in Art Welter's Keystone Sub you have a wide tuning range by changing the Keystone exit. By the way, I added a drawing for the Keystone Sub to Art's thread.
Regards,
It might be cool to see a large cabinet that was bristling with linear actuators and servos with sliding panels. The trouble I'd suspect would be keeping panels movable but at the same time reasonably air tight.
I like the idea because you could theoretically have different tunings for different sorts of playback material.
I've experimented with thread rods and cabinet tuning in the past, but that was mostly resonance tweaking.
I like the idea because you could theoretically have different tunings for different sorts of playback material.
I've experimented with thread rods and cabinet tuning in the past, but that was mostly resonance tweaking.
Wayne I don't doubt Art's measurements at all. Most of the competently designed modern drivers that I have played with will stay below 10% thd in a simple sealed enclosure even when driven very hard as long as the excursion is kept in check. Any resonant alignment driven below the cabinet loading will have rapidly increasing distortion. All should be operated with a hpf IMHO so anything below the cabinet loading is sort of for curiosity sake really. The only one that I have ever seen that did not exhibit this exponentially increased thd is your 12pi data and yes I have done measurements on subs with push pull arranged drivers. They exhibit the same rapid increase in distortion as any other. What your data shows is astounding to be honest.
I have a couple of questions...
You say a Lab12 at 400w will have much higher distortion but how are you judging 400w input? By voltage into a nominal impedance, measured impedance min, etc? What voltage are we talking? A lab12 will not handle a couple of 400w nominal sweeps? I find that hard to believe. As you said it seems that we are talking apples vs lawn chair here or something.
When a horn is operated below its cut off the fundamental tone is not amplified by the horn loading anymore but the harmonics will still be and may recieve a huge boost relative to the fundamental. Do you agree with this statement? This usually causes the thd to be even worse than a direct radiator in some areas below the cutoff from what I have seen.
Btw these are honest questions.
I have a couple of questions...
You say a Lab12 at 400w will have much higher distortion but how are you judging 400w input? By voltage into a nominal impedance, measured impedance min, etc? What voltage are we talking? A lab12 will not handle a couple of 400w nominal sweeps? I find that hard to believe. As you said it seems that we are talking apples vs lawn chair here or something.
When a horn is operated below its cut off the fundamental tone is not amplified by the horn loading anymore but the harmonics will still be and may recieve a huge boost relative to the fundamental. Do you agree with this statement? This usually causes the thd to be even worse than a direct radiator in some areas below the cutoff from what I have seen.
Btw these are honest questions.
The tests were done at 49 volts, nominally 400 watts per speaker.
I checked my test results before posting them.
The percent distortion only includes the largest harmonic, but addition of the lesser harmonic distortion would only amount to a fraction of a percentage point, so I did not go to the trouble of calculating it.
The method of calculation is time intensive enough as it is .
The Lab 12s did not exceed 10% distortion except when over Xmax.
Art Welter
Attachments
Of course distortion will rise dramatically when xmax is exceeded, quickly going above 10% even at low power levels and well over 100% at moderate to high power levels. But that's an extreme case. There are sources of nonlinearity within the normal operating conditions, like suspension nonlinearity and flux modulation. Some of the nonlinearities are symmetrical, others asymmetrical, so both even harmonics and odd harmonics are generated.
The LAB12 is a good woofer, but it has no shorting ring so flux modulation occurs. Some may take the position that distortion from subwoofers is less audible than other audio ranges, so increased distortion is not terribly important. That's probably true, and perhaps more relevant is the fact that shorting rings have proven to be not terribly effective below 100Hz or so. But the fact remains that flux modulation occurs, and it rises as power is increased. So does excursion, which increases distortion from suspension non-linearities.
Total harmonic distortion from a direct radiating LAB12 is above 6-7% in the 50Hz to 100Hz range before it even reaches 100 watts. At 400 watts, it is considerably higher. In fact, like I said earlier, a LAB12 at 400 watts is in its "red zone" and will come apart within an hour or two. At the higher end, around 100Hz, it will suffer from thermal stress, the voice coil adhesive will weaken and it will eventually fail. At the lower end, it gets closer to xmech, depending on the acoustic loading, of course.
What I think probably we're seeing in weltersys's data, is the RTA he's using isn't completely accurate. Or it could be that the power levels used weren't as high as stated. I'm not sure, but I do know that the THD of a direct radiating LAB12 is not even close to 6-7% at any frequency at 400 watts. It's a great driver when used at power levels lower than that, say in a vented or sealed cabinet for home hifi. It works great in hornsubs too, which limit excursion. But you do have to high-pass them (to limit excursion) and low-pass them too (to limit thermal stress), or ideally add cooling plugs to get the most from them.
The LAB12 is a good woofer, but it has no shorting ring so flux modulation occurs. Some may take the position that distortion from subwoofers is less audible than other audio ranges, so increased distortion is not terribly important. That's probably true, and perhaps more relevant is the fact that shorting rings have proven to be not terribly effective below 100Hz or so. But the fact remains that flux modulation occurs, and it rises as power is increased. So does excursion, which increases distortion from suspension non-linearities.
Total harmonic distortion from a direct radiating LAB12 is above 6-7% in the 50Hz to 100Hz range before it even reaches 100 watts. At 400 watts, it is considerably higher. In fact, like I said earlier, a LAB12 at 400 watts is in its "red zone" and will come apart within an hour or two. At the higher end, around 100Hz, it will suffer from thermal stress, the voice coil adhesive will weaken and it will eventually fail. At the lower end, it gets closer to xmech, depending on the acoustic loading, of course.
What I think probably we're seeing in weltersys's data, is the RTA he's using isn't completely accurate. Or it could be that the power levels used weren't as high as stated. I'm not sure, but I do know that the THD of a direct radiating LAB12 is not even close to 6-7% at any frequency at 400 watts. It's a great driver when used at power levels lower than that, say in a vented or sealed cabinet for home hifi. It works great in hornsubs too, which limit excursion. But you do have to high-pass them (to limit excursion) and low-pass them too (to limit thermal stress), or ideally add cooling plugs to get the most from them.
Wayne,
The voltage could be off by a volt or so. I have several VOMs, the variation between them is only a couple volts at 120, probably only a volt at 49 V. I set the voltage at 60 Hz, presumably a frequency the meter should be most accurate at.
The Smaart RTA seems quite accurate to me, why do you doubt it ?
A screen shot is attached below, can you point out some screen shot that looks "off" to you?
Where is your data supporting the statement that the "THD of a direct radiating LAB12 is not even close to 6-7% at any frequency at 400 watts." ?
Art
Attachments
I had considered putting T-nuts inside the Keystone front panel to attach a "step down" plate, but found that there was little music that the decrease in the low corner was worth the reduction in the rest of the range.
We're definitely not on the same page, because it's not a matter of a volt or two, or a decibel or two. The THD of a direct radiating LAB12 is ~25dB under the fundamental through the passband, sometimes higher, at 100 watts. That's the same numbers you're quoting for 400 watts, where distortion is much higher.
Now put that woofer in a horn, and distortion is reduced in the passband. Or use push-pull drive. Or both. But a single LAB12 run as a direct radiator doesn't get anywhere close to 6-7% THD at 400 watts. You can't even keep it in one piece at that level for very long.
Now put that woofer in a horn, and distortion is reduced in the passband. Or use push-pull drive. Or both. But a single LAB12 run as a direct radiator doesn't get anywhere close to 6-7% THD at 400 watts. You can't even keep it in one piece at that level for very long.
i cant, i dont use hornresp, and even if i did, i dont have the knowledge to get the calcs right (i dont even know what impulse behavior is?)
im not attached to the dayton in any way, i just want loud and low and outdoors
i have paypal for someone who is good with hornresp and calcs and knowledge,
and i am good with a tape measure and saw.
The drivers stayed in one piece, the test durations of the ported 2xLab 12 were long enough to also measure excursion. I tried to keep the tests as short as possible to avoid thermal compression, but the speakers were often hit for 15 seconds or so.
I have had no failures due to overheating, but the Lab 12 driver definitely warms up with 49 volts of sine wave going through it.
The Lab 12 is a good low distortion driver, it does not need horn loading to be low distortion.
Again,I'll ask if you have some data to back your distortion claims regarding Lab 12s in ported cabinets, post it, otherwise you are just blowing smoke.
Art
- Status
- Not open for further replies.