Jump to content

New, Hindsight-Design AR-5 Loudspeaker


tysontom

Recommended Posts

I have the parameters somewhere on the Advent woofer, but I can't put my hands on the information. A few years ago, I also measured the harmonic distortion of my pair of Advents compared with my AR-3as (using my HP-339A analyzer and a Earthworks M30 microphone and Earthworks microphone preamp), and the AR-3a was significantly better at anything approaching a high-output level, when each system was adjusted for identical SPL output. I also have a pair of Small Advents, which have the same low resonance, but that speaker doesn't seem to tolerate high power levels as readily as the Large Advent. I measured down to 20 Hz and up to 20-25 watts/rms equivalent input power, which might not seem like much power, but it is a significant power level at steady-state input loads at such low frequencies. Most testers over the years stuck with 1-watt input for this measurement. This definitely "separates the men from the boys," and the AR-3a woofer handled that power without distress.

—Tom Tyson

It is truly odd that distortion figures are usually measured at constant input level. Idea for measuring at similar SPL has surfaced time to time, but it has never been adopted too widely for tests. I scanned distortion figures from Nov 1976 issue of Swedish Hifi & Musik AR10pi/JBL L166 test. Both speakers are from same period, size and cost

Fig 2 there is JBL L166 anechoic respose and distortin figures at 2V drive.

Fig 3 there is AR10pi at similar drive. Distortion figures are quite similar in mid and hf but low frequency distortion is higher.

Fig 4 AR10pi is driven 6dB harder to increase SPL to meet L166 output. 2nd and 3rd harmonic distortion figures have been increased quite bit over L166 distortion figures especially in LF.

I do not know how typical this is, but it seems that distortion will rise quite sharply when SPL does increase even at moderate 2-4 V drive level.

I suppose that these low distortion figures do have something to do with this "Smoke on the water" effect. I do not know if you can assume that all high efficiency speakers do have good distortion characteristics... but at least JBL distortion figures seems be quite low whenever they have been measured.

Pete

I do have pair of Altec 604E´s in basement reserved for some project and reproducing Mastering Labs crossover has been on my mind... but no one has been able to solve if 0,15 mH tap of 1,5 mH inductor is facing tweeter negative or positive side. From pic it seems to facing tweeter positive, but can´s be sure.

Best Regards

Kimmo

PS in text it was mentioned that frequency response measurement was slightly compromised under 200 Hz due to the dimensions of anechoic room

post-126436-0-97484500-1437305072_thumb.

604xover_SP1.pdf

Link to comment
Share on other sites

  • Replies 66
  • Created
  • Last Reply

You mean to say that all those years your AR's were not making music? And now you'll tell us that

none of us can possible be hearing the music unless we get your current favorite and blessed speaker.

You are an internet BS artist, that is all.

You might consider taking a step back, Pete - you apparently have some long-term unresolved issues, and you seem sort of worked-up.

This isn't my field of training, but maybe just take a few deep breaths, or something.

And the name-calling? On a completely serious note, I'd appreciate it if you wouldn't do that.

In fact, since my actual name and real-life details are known by a few members of this forum, I'm advising you to stop doing it.

If you're not able to avoid internalizing something that I've written about a loudspeaker, I'm going to suggest that you simply refrain from reading anything that I write.

.

Link to comment
Share on other sites

Hi Kimmo, very interesting! I'm surprised that the JBLs do so well given that they usually do not

have a lot of Xmax but it is possible that at those low drive voltages they do not exceed Xmax.

Being high efficiency the magnet system is probably more deeply saturated which helps to keep

distortion low - this is probably what's going on here, as drive level is increased the woofers with

longer Xmax will have an advantage. I agree that distortion measurements should be done at a

reference SPL and best with the FR flattened.

Looking at the MAstering Labs crossover it seems that the 1.85 uF cap and pot are the very high

frequency path to the tweeter. The 2.3 uF path is second order with the 1.5 and .125 mH inductors

as the shunt inductance. The .26 mH causes this path to only boost the lower range of the

tweeter. I think that the tap must be as shown in the diagram with the .15 mH section at the top,

since otherwise the much larger inductance would not allow much signal to get to the tweeter.

Interesting, I've not looked at that XO in a long time. I'm fairly certain that there are people on the

net who've built these from scratch.

Link to comment
Share on other sites

Tom, here are pictures of a cone and VC from a masonite OLA woofer where you can

see the copper color and lack of vent holes. Also note that the distance from the top of

the windings to the cone is longer to make up for the frame depth difference between

the OLA and NLA woofers. This is the only Advent woofer that I've taken completely

apart at this point so there is no chance that I mixed it up.

Also, the wind height seems closer to 5/8" than 3/4" in this woofer even if the angle in

the picture makes it look like 3/4".

Edit: Just reread your post and the voice coil does not look right for an OLA - makes no

sense to me. I don't think even the NLA all metal woofers had an aluminum former like that.

Perhaps there were changes in later production, but that does not look at all like an OLA

voice coil to me. I've not yet seen a Jensen VC but that's what I'd expect it to look like.

----------------------------

Do you know if that woofer had a round magnet? Previously, I'd expect the magnet to

be square because the dust cap is fabric but I recently came across a round magnet

version (with a Jensen part number) and a fabric dust cap. They must have made the

changes incrementally, or as supplies changed.

post-101114-0-49976000-1437323266_thumb.

post-101114-0-17512800-1437323289_thumb.

Link to comment
Share on other sites

>First, I'll point out that I have an early Large Advent woofer cone and voice coil here in my hand, and it does not look anything like yours. Copper colored and without vent holes. I've refoamed a lot of Advent woofers and inspected them for bad glue joints but I don't cut the dust cap so I didn't see below the spider but I always remember a copper colored former. It is possible that there were production changes. I've posted a picture here in the past and will look for it but the search here does not work very well. [PeteB]<

Let's see an image of that speaker you were holding in your hand. This way we can tell if you are talking about the same thing. I think you might be a bit confused and are conflating different Advent models from different years. I'm referring to the first iteration of "The Advent Loudspeaker," designed by Henry Kloss in 1968. If your woofer doesn't look anything like the ones I described (mine are from 1971, and they are the original version of "The Advent Loudspeaker," exactly like the ones introduced in 1968), then you may be talking about something entirely different. I've attached some other pictures of my pair and some additional anechoic measurements that were made in 1976. Let's see some of your images and/or test results.

>I asked the question as a lead into what I think would have been needed to compete with Advent. Kloss states outright in the Advent literature that he built the OLA woofer on a bigger frame to allow for more "linear" throw. We had Large Advents when I was a young teen, and I was very impressed with the excursion capability, _and_ the fact that it was nearly impossible to bottom the woofer in the way that most others would make a loud crack sound. I was reasonably careful with them but did drive a single one with an amp that would do 100W into their 5 ish ohm load and by eye it seemed like the woofer would do far more that .25" one way, seemed to be closer to .5" one way without bottoming. [PeteB]<

Kloss built the Advent woofer on a 12-inch frame for one distinct marketing reason: he wanted to say his speaker had a "12-inch" woofer to compete with the bigger AR-3a 12-inch woofer. Period. It's a much cheaper, lighter-weight woofer than the AR-3a woofer. Calling the Advent frame a "12-inch frame" had nothing to do with allowing for a more "linear" throw. You don't need a bigger frame for a longer throw, and it's not like it's moving ±1-2 inches or more. Also, the cone diameter on the Advent woofer is quite a bit smaller than the AR-3a woofer (this falls in with Hofman's Iron Law) in the same-sized cabinet, thus allowing a slightly lighter, higher-fs than the equivalent, larger-cone AR woofer, and ending up with an equivalent fc. It's a very good design idea, but the point I was making was that the Advent woofer doesn't come close to comparing with an AR 12-inch woofer in terms of linearity and distortion.

I do agree with you that the Advent woofer made allowances for long, non-linear excursions, probably better than the AR-3a ferrite woofer. The Advent has a total linear excursion of around 0.375 inches whereas the AR-3a has a greater linear excursion of 0.5 to 0.625 inches. The AR woofer thus has greater volume/displacement due to the larger diaphragm and longer excursion. Total physical excursion of the AR-3a woofer is 1.1 inches (XMECH) before it strikes the back plate, whereas the Advent will continue to flap back and forth without striking the back plate, well out of the linear gap. The end result: the Advent woofer will enter higher and higher harmonic distortion, but short-term peak levels are tolerable and not terribly audible; the AR is much cleaner and less distorted, but it is prone to strike the back plate if over-driven, a design that was not changed until well into the 1990s. AR actually stiffened the skiver on later models (reduced compliance at extreme ends) to improve power-handling rather than raise the back plate, a mistake in my view. Roy Allison and Chuck McShane should have made a simple tooling change early on to prevent any trouble.

—Tom Tyson

post-100160-0-39563300-1437331661_thumb.

post-100160-0-73495300-1437331748_thumb.

post-100160-0-75257000-1437331761_thumb.

post-100160-0-59165600-1437331770_thumb.

Link to comment
Share on other sites

Tom, here are pictures of a cone and VC from a masonite OLA woofer where you can

see the copper color and lack of vent holes. Also note that the distance from the top of

the windings to the cone is longer to make up for the frame depth difference between

the OLA and NLA woofers. This is the only Advent woofer that I've taken completely

apart at this point so there is no chance that I mixed it up.

Also, the wind height seems closer to 5/8" than 3/4" in this woofer even if the angle in

the picture makes it look like 3/4".

Edit: Just reread your post and the voice coil does not look right for an OLA - makes no

sense to me. I don't think even the NLA all metal woofers had an aluminum former like that.

Perhaps there were changes in later production, but that does not look at all like an OLA

voice coil to me. I've not yet seen a Jensen VC but that's what I'd expect it to look like.

----------------------------

Do you know if that woofer had a round magnet? Previously, I'd expect the magnet to

be square because the dust cap is fabric but I recently came across a round magnet

version (with a Jensen part number) and a fabric dust cap. They must have made the

changes incrementally, or as supplies changed.

I think we are looking at variations of the same woofer. I do know that Kloss used a bronze former for some of his KLH woofers, and maybe he used copper for his very first versions, but I think the voice coil diameter and height look very similar to mine. The magnet was square on my version. The cone is identical, etc., so there must have been some variations in the various years of production. Bear in mind that Advent produced well over one-half million of the speakers, I believe. I also think the date stamp on my woofer was 1971, but I no longer have the magnet portion.

Link to comment
Share on other sites

I think we are looking at variations of the same woofer. I do know that Kloss used a bronze former for some of his KLH woofers, and maybe he used copper for his very first versions, but I think the voice coil diameter and height look very similar to mine. The magnet was square on my version. The cone is identical, etc., so there must have been some variations in the various years of production. Bear in mind that Advent produced well over one-half million of the speakers, I believe. I also think the date stamp on my woofer was 1971, but I no longer have the magnet portion.

Let me remind people that I do not like the sound of the stock Large Advents, however, with BSC or my crossover mod,

they are an outstanding value and with it I do like their sound as a bargain system.

I do not use LAs permanently installed in any of my systems. My parents use a pair connected to their HD TV with a

60W/ch amp - for those that were curious. They don't upgrade unless they really need to.

Let me also point out that my parents bought Large Advents in order to have a stereo system, it usually filled the house

with music while my mom cooked, yes she listened to them from another room, lol! They were played when we had

company, when they had parties, and my grandfather loved to hear Greek music and visited often. My brother and I had

our own systems and nobody bothered to sit down and listen to the Advent system. My dad was a very busy business man in those years. He was an electronics tech in the Navy and certainly built many things. He built an oscilloscope from scratch shortly after college, lol!

I only have one Advent woofer fully taken apart just to be clear that I am not mixing up parts and it is shown in my

hand in the attached picture, The date on the magnet is Nov 7, 1972, also shown in a picture.

Tom, I just noticed that there is an indicator in one of your pictures, note that the spider is attached at the point where the VC

connects to the cone. That is how a New Large Advent woofer is built. I know this, because the first time I refoamed

an NLA woofer I made note of it in the Advent section of this forum. If necessary I will find and take pictures of both types.

This is a list of the Large Advent Woofers that I own, mostly but not all refoamed:

8 Masonite OLA Woofers - Square Magnet

7 New Large Advent Woofers - Square Magnet

6 Jensen Large Advent woofers all with round magnets and Jensen part numbers, one with a fabric dust cap, most are paper.

Anyway, I agree that the diameter and the wind height are almost if not the same in our pictures, however,

I mentioned that the distance from the top of the windings to the cone are very different. Mine is about 1 inch,

and yours is less than half an inch perhaps .25". A picture is attached here with red lines showing this distance

that I'm talking about.

A second picture is attached showing where the spider is glued on an OLA woofer, I marked it with a black

sharpie pen so that I could reassemble that woofer, there is a red arrow pointing to the black mark. I can also

see some remains of the glue for that joint. There is clearly about .25" from the spider glue joint to the cone whereas your

picture has the spider right up against the cone. I'll take a picture of the spider glue joint on both types of woofers when

I have a chance.

post-101114-0-30190900-1437347647_thumb.

post-101114-0-17123300-1437347665_thumb.

post-101114-0-28707200-1437347700_thumb.

post-101114-0-70638700-1437347726_thumb.

Link to comment
Share on other sites

Let me also point out that my parents bought Large Advents in order to have a stereo system, it usually filled the house

with music while my mom cooked, yes she listened to them from another room, lol!

When I was in grammar school, my dad built a pair of speakers using parts he bought from one of the stores in NYC's Radio Row. Being in another room improved their sound immensely. :)

Link to comment
Share on other sites

Ha, good one!

I should probably add that most of what we study in engineering are linear systems with

some obvious non-linear devices such as rectifiers, clippers, and mixers thrown in for radio

circuits. But the detailed analysis of non-linear systems is very complex and not usually

covered in any sort of detail until graduate level courses. What I've observed is that by

adjusting the non-linear parameters in a synergistic way one can come up with much more

optimal systems than if these synergies are ignored. It is impressive that the LA woofer

with only 1/8" linear Xmax can easily do twice that without objectionable distortion and at

least 3 times that without total loss of control. I believe that this is key to their success. To

put it simply, you could beat the crap out of them and they kept on going _and_ they

didn't sound too bad. They certainly survived parties reasonably well.

The Dynaudio 21W-54 woofer was another one with outstanding non-linear performance.

Interesting thing about the 21W-54 is that it had 3mm linear Xmax (low) and a huge 14.5 mm Xmech,

most European woofers of that time had 10mm one way Xmech.

These drivers have headroom.

There is also a JBL that does very well, I'd have to look up the part number.

Most other woofers do not do very well when pushed beyond Xmax.

Edit: The Meniscus Eclipse line of woofers also had very impressive performance, they and Eminence

who fabricated them initially learned a few tricks to make them more linear without increasing cost.

These go back probably more than 20 years.

Link to comment
Share on other sites

JBL used to have a 10" 3-way model (JBL L80T) which was similar to AR-5. It has very low distortion, high efficiency and is extremely flat through the important midrange, according to the review in High Fidelity. http://www.audioheritage.org/vbulletin/showthread.php?34400-JBL-L80T-Reviewed-by-High-Fidelity-Magazine-June-1986

There are several proprietary technologies which helped the low distortion. Symmetrical Magnetic Gap(SMG) is one of them. I believe Scanspeak's SD(symetrical drive) motor design is a derivative of SMG technology.

Link to comment
Share on other sites

Attached is a picture of a Polk dB1040 auto sound woofer, look at that raised back plate.

It is well known that drivers are stressed very hard in auto sound and that is what it takes

to give them generous Xmech. It is actually a very well engineered woofer and the list

price is $99 but can be found for about $60. Nice job. These are also _not_ huge Xmax

subwoofers and I liked the T&S specs when I last checked them several years ago, but

I believe that they've changed them.

Looked up the parameters, the cone mass is more than double what would work well in

the Advent sized box - that is unfortunate. A custom woofer built on that platform would

work well. The motor is probably fine the cone is what needs to be replaced.

post-101114-0-81113200-1437356327_thumb.

Link to comment
Share on other sites

Pictures are attached of my restored Advents, left to right, OLA, NLA, 5002.

Second picture 5012 in oak veneer.

They are completely stock with original stuffing, no extra bracing, inductors tested

and new film caps.

I've collected quite a few more since these were taken.

http://baselaudiolab.com/LA_ADVENT1s.jpg

http://baselaudiolab.com/JLA_5012_2s.jpg

I'm rebuilding several more to give away for friends who are just getting started. They make

a lot of sense for young people who often need a more rugged system.

Link to comment
Share on other sites

It is truly odd that distortion figures are usually measured at constant input level. Idea for measuring at similar SPL has surfaced time to time, but it has never been adopted too widely for tests. I scanned distortion figures from Nov 1976 issue of Swedish Hifi & Musik AR10pi/JBL L166 test. Both speakers are from same period, size and cost

Fig 2 there is JBL L166 anechoic respose and distortin figures at 2V drive.

Fig 3 there is AR10pi at similar drive. Distortion figures are quite similar in mid and hf but low frequency distortion is higher.

Fig 4 AR10pi is driven 6dB harder to increase SPL to meet L166 output. 2nd and 3rd harmonic distortion figures have been increased quite bit over L166 distortion figures especially in LF.

I do not know how typical this is, but it seems that distortion will rise quite sharply when SPL does increase even at moderate 2-4 V drive level.

I suppose that these low distortion figures do have something to do with this "Smoke on the water" effect. I do not know if you can assume that all high efficiency speakers do have good distortion characteristics... but at least JBL distortion figures seems be quite low whenever they have been measured.

Pete

I do have pair of Altec 604E´s in basement reserved for some project and reproducing Mastering Labs crossover has been on my mind... but no one has been able to solve if 0,15 mH tap of 1,5 mH inductor is facing tweeter negative or positive side. From pic it seems to facing tweeter positive, but can´s be sure.

Best Regards

Kimmo

PS in text it was mentioned that frequency response measurement was slightly compromised under 200 Hz due to the dimensions of anechoic room

Iso,

I could barely interpret the figures in your attached picture, so I can't tell the measurement details. Could you kindly send a higher-resolution image of the test report? Most likely, the distortion in the bass was not taken into the lowest frequencies, and the figures were based on a fairly high SPL level in the mid-bass, which would favor the very efficient JBL L166. The distortion figures may also have been measured into a 360-degree solid angle, also favoring frequencies above the region of radiation loading.

In general terms and in recent years, there has been less emphasis on measuring bass distortion. Speakers today, in general, have improved in reproduction of bass frequencies, at least down into the mid-low frequencies. In the past, however, the technique for measuring distortion was more simplified and basic, and one could relate performance to distortion more easily. Instead of using the current dB-reference levels (which have to be interpreted in one's mind to percent distortion), the earlier methods were to measure a loudspeaker usually at a 1-meter distance, anechoically, with 1-watt or greater, constant input-power for devices with similar efficiency, sweeping the frequency down from 100 Hz or so to 20-30 Hz. This was a basic and straight-forward method of measurement used for decades prior to the introduction of internet-based, loudspeaker-testing software, where basically "anything" goes.

Earlier test methods were based primarily on specific "standards," such as RETMA-STANDARD-SE-103, ASA and 61 IRE 30.RP1. Insofar as most speakers today are designed as 2-Pi devicesi.e., designed to be placed against a wall or slightly out from a wallthen the measurement should be made with the device flush with a flat, 180-degree surface and measured into free space. Efficiency differences were compensated for by choosing a nominal output SPL level for all speakers under test and using that for the measurement level.

That said, it would be highly unlikely that an L166 (an otherwise fine speaker) is going to surpass an AR-10Pi in the low frequencies below 40 Hz, either in output or in distortion. In my personal experience, the L-166 isn't even close. The L-166 cabinet is a bass-reflex design, and i think it was tuned to around 50 Hz or so, meaning that roll-off below the port resonance is rapid. Therefore, when you test it for deep-bass distortion, it will have difficulty, whereas in frequencies above 40-50 Hz, it will look fantastic.

—Tom Tyson

Link to comment
Share on other sites

Tom

I have now no scanner available... so the following pics are made from pics taken with Fujifilm X10 camera and the original magazine pics were not razor sharp either. But these are best I can do now. I added original Hifi & Music review here... Turbon may be interested to see it.

It seems that in second AR10pi measurement level was matched at 2 kHz to JBL L166. As usual, measurement details were not filled in too well on measurement sheet. However it may be reasonable to expect that distortion curve has been elevated by 20 dB.

As it was mentioned that dimensions of anechoic room did have influence to response below 200 Hz, most likely they measured response at 1m in Sweden. These days most LF measurements made here in Finland were made at close field of woofer and this curve was joined to 1m anechoic curve to remove effect of anechoic room dimensions.

Best Regards

Kimmo

post-126436-0-56406600-1437419325_thumb.

post-126436-0-78554100-1437419345_thumb.

post-126436-0-86354100-1437419382_thumb.

post-126436-0-96439700-1437419404_thumb.

post-126436-0-66849200-1437419416_thumb.

Link to comment
Share on other sites

Kimmo,

Thanks for the higher-rez images. This is the first and only test I've ever seen that shows a JBL to equal—let alone surpassan AR woofer in deep-bass low distortion, and I certainly have no explanation for that test you supplied. During the years when these speakers (JBL L166 and AR-3a/AR10Pi, etc.) came along, I never read a single review that even hinted at the possibility that a JBL L-166, or similar JBL, with the smallish vented cabinet, could even approach the better acoustic-suspension woofers. The JBLs certainly had their strong points—and they were popular as rock-music speakersbut these speakers were never known for their deep-bass, low-harmonic distortion performance. There was never any report in the US of such performance, to my knowledge. It would have been widely reported, but I never saw a single reference at all to such performance, but perhaps others on this site can provide some earlier tests that validate it. The JBLs were well known for their bright, west-coast sound and very high sensitivity, and these speakers had a strong following.

I did go back to see if I could find any reviews on the L-166 done in the US that were tested in the same way as an equivalent AR speaker, and I did find a CBS Labs report (High Fidelity magazine) that tested an AR-3a in 1973 with their exhaustive harmonic-distortion testing (anechoic) and a JBL L-166 tested in 1976 the same way. As you probably know, the AR-3a and the AR-10Pi share exactly the same woofer and cabinet volume, etc. I have another report on the AR-LST that has even lower distortion, but I used the AR-3a as a reference in this case. CBS Labs tests were thorough and documented, and these tests were considered to be benchmarks at the time.

These tests show the JBL L-166 to have slightly less distortion (3rd harmonic) in the 300 Hz range, but over seven-times greater distortion than the AR-3a in the lower 80 Hz range at 95 dB/SPL, and it shows that the L-166 overloaded when trying to reach levels of 100 dB or more. I've attached some images of those tests.

—Tom Tyson

post-100160-0-22704800-1437429842_thumb.

post-100160-0-78762600-1437429865_thumb.

post-100160-0-75142800-1437430190_thumb.

post-100160-0-68862300-1437430206_thumb.

post-100160-0-68582200-1437430580_thumb.

Link to comment
Share on other sites

Tom

Thank you for your well thought response.

Measured 80 Hz and 300 Hz data at 90 dB from H&M and CBS labs are fairly similar for L166, even THD content should vary slightly in H&M test compared to CBS Labs. This is due to the fact that SPL of 90 dB is not constant over entire range and room dimensions do have some effect to fundamental to 2nd and 3rd overtone level.

It may be L166 vent that can not handle high SPL at low frequencies. Vent dia of 2,5" was quite typical these days for 12" woofer and this is not good enough for high SPL under 50 Hz.

However, H&M test at 90 dB´s should correlate very well to average or even slightly spirited listening conditions. Low THD and good response below 40 Hz are nice features, but their importance is something like absolute phase... not everyone can tell difference as very few instruments do reproduce so low notes.

10pi LF figures surely look odd. I suppose that A2 version of 20003-0 woofer was used when AR3a test was made and A4 was used for 10pi. It seems very unlikely that less compliant spider will increase distortion so much at 90dB. Maybe quality control has bee unable to detect defective woofer?

Any further ideas?

Best Regards

Kimmo

Link to comment
Share on other sites

Tom

Thank you for your well thought response.

Measured 80 Hz and 300 Hz data at 90 dB from H&M and CBS labs are fairly similar for L166, even THD content should vary slightly in H&M test compared to CBS Labs. This is due to the fact that SPL of 90 dB is not constant over entire range and room dimensions do have some effect to fundamental to 2nd and 3rd overtone level.

It may be L166 vent that can not handle high SPL at low frequencies. Vent dia of 2,5" was quite typical these days for 12" woofer and this is not good enough for high SPL under 50 Hz.

However, H&M test at 90 dB´s should correlate very well to average or even slightly spirited listening conditions. Low THD and good response below 40 Hz are nice features, but their importance is something like absolute phase... not everyone can tell difference as very few instruments do reproduce so low notes.

10pi LF figures surely look odd. I suppose that A2 version of 20003-0 woofer was used when AR3a test was made and A4 was used for 10pi. It seems very unlikely that less compliant spider will increase distortion so much at 90dB. Maybe quality control has bee unable to detect defective woofer?

Any further ideas?

Best Regards

Kimmo

Dear Kimmo,

Thanks for your response, too.

There is definitely something amiss in the HiFi & Musik test, in my view, but not knowing the test conditions, it's very hard to say what it is. It's also hard to reconcile the JBL L-166 going to 20 Hz with lower distortion than the AR-10Pi under any conceivable circumstances, and I think the CBS Labs test, which was done under anechoic conditions, not in a room, clearly show the weakness in a vented design as you go lower and lower in frequency below the tuned frequency. This results is the "unloading" of the driver, forcing it into excessive excursions, and this definitely did show up on the test with CBS Labs. And you also point out that the L-166 vent diameter was 2.5 inches, which would indicate an fc of 45-60 Hz or so. Therefore, I don't think that the L-166, though a fine design for the time, was a low-resonance vented design as are some of newer designs which do much better in deep-bass frequencies, with lower tuned-vent frequencies and greater driver XMAX). Electro-Voice did get into this low-resonance technology with some of their designs of the late 70s. But to my knowledge, at that time, JBL had not even tuned their 18-inch subwoofers for >20 Hz frequencies, so I'm pretty sure that the L-166 was tuned much higher in keeping with the efficiency and power-handling needs of the time.

As you pointed out, "Low THD and good response below 40 Hz are nice features, but their importance is something like absolute phase... not everyone can tell difference as very few instruments do reproduce so low notes." I think this was the rationale for the JBL designs for many years, and perhaps it's a moot point for music lovers who care for the majority of music.

I do think that at the extreme excursion limits, the version of the woofer used in the 10Pi might not perform quite as well as the earlier AR-3a version, but the differences in distortion are very small. Only under heavy input power would it be notices, I believe. However, some versions of the earliest AR-10 Pi used an almost identical woofer to most of the AR-3a versions, so it's hard to say. In any event, the distortion differences are very small, as the "stiffness" increases were set for the limits of the linear excursion, I believe.

I guess we'll just have to let it ride for now!

—Tom Tyson

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...