Beyond the Ariel

No matter how flat the frequency response is made, I hear a layer of metallic haze that interferes with the upper harmonics of orchestral strings and woodwinds..........but it should be no surprise that I can't live with a speaker system that steers me away from orchestral music.

+1

That said, having done some experimentation with both small and large format Altec aluminum diaphragms dating from the '50s thru the early '90s, the late, great small ultra-lights did best overall, but in each case the surround is where this is generated, i.e. it's just white noise for lack of knowing a more technically accurate description, which I damped by 'dusting' on a thin layer of Dammar with an airbrush, though some folks have found coating the whole diaphragm and accepting the loss of top end 'air' around the instruments.

GM
 
Great Plains Audio is the closest we can get to current-production Altec. The shop in Oklahoma is the former repair station for Altec, the tooling is the same, and the workers are the same. The Altec name was sold to the Taiwanese car-stereo manufacturer Sparkomatic many years ago, so GPA doesn't get to use it or the Altec conductor-in-tails logo.

Since Gary Dahl actually is a conductor, and the 416 and 515 Classic Series Alnico drivers are built by long-time Altec employees, that's close enough for me. Don't want to mess with eBay, Audiogon, or re-magnetizing vintage drivers.

Thanks for the tip about the ketone polymer Faital Pro 1" exit drivers. Three of them (HF104, HF106, HF108) look like interesting candidates for a supertweeter, since the factory curves imply response beyond 20 kHz.

Gary's technique of a VHF bypass around the autoformer (or L-Pad) is interesting, since beryllium will tolerate a HF boost better than most other materials. Applying a boost to a driver that's already in breakup seems wrong in principle, and beryllium neatly sidesteps that problem.
 
Last edited:
I thought he used the great plains copy of altec 416 for bass

I use the 416's in the upper enclosures (3 cf sealed) to cover the 70-700 range. Below that are the separately powered subwoofers. Each 5 cf enclosure contains one TD15-H woofer and two PR15 passive radiators (one on each side of the cabinet).

My original configuration used TD15-M's in the upper enclosures.

Gary Dahl
 
Two quick items: I've written a brief show report on the 2014 RMAF from the perspective triode-amp owners, and it will appear in Positive Feedback Online. I've written for David Robinson, the editor, since the early Nineties, and PF is the magazine where I published the original plans for the Ariel (before the Web existed).

I should also mention that I'm considering a version using Altec/GPA 515's and 416's in parallel. LEAP simulations (using Gary's T/S measurements) indicate that with a 3.3 cubic foot/50 liter closed box (for each driver), the combination sums to flat, with an F6 around 50 Hz, and efficiency in the passband around 100~102 dB.

The LEAP simulations show the two drivers with complementary responses: the 515 is tilted towards the midrange, while the 416 is tilted towards the bass. In effect, there's a fractional (about 2 dB/octave) acoustical crossover between the two.

Performance isn't as high as a straight (non-folded) bass horn, but it's not as physically large either. At first glance, the 515/416 direct-radiator combo seems midway in performance between a single 15" direct radiator and a bass horn.
 
Last edited:
Woof! Woof!

I had WinISD graphs of two Peavey 18" woofers which could be counter-force boxed for 102db/watt efficiency.

1) Ported in 15.6 cuft with F3 = 34Hz

2) Sealed in 9 cuft with 30Hz Linkwitz transform. A plot of amplifier boost power is included. At 30Hz the Linkwitz transform 6x amp power gives sealed Qtc and saves 6.6 cuft of box volume.
 

Attachments

  • Three Way  C-Force.jpg
    Three Way C-Force.jpg
    101.8 KB · Views: 642
  • CounterForce 18s SPL.jpg
    CounterForce 18s SPL.jpg
    105.5 KB · Views: 621
  • CounterForce 18s Linkwitz PowerBoost.jpg
    CounterForce 18s Linkwitz PowerBoost.jpg
    92.3 KB · Views: 623
Member
Joined 2012
Paid Member
Newbie Speaker Build: Part 1

oltos, Your questions are very diverse.

Q?: Do you want to invest significant money and time to "educate yourself" on crossover design, speaker measurement, and cabinet construction to create your own unique speaker? OR learn as you copy a proven design? OR copy the Heart of a proven design with cost reductions around the edges? Heart: GPA416-8 alnico in sealed box + LeCleach 350-425 horn + 1.4" modern compression driver.

Q?: 2 vs 3 vs 4 way??

If you carefully copy g3dahl's final 4-way construction and crossover posts you can be confident of great success without wasting money and time. Purchase more expensive speakers, but save on trial-and-error crossover components plus measurement tools.

You may decide on the Be dome....

Thanks Line Source for critiquing my plan. I only wish that I had more time to be the DIYer that I used to be (I troubleshot LSI digital systems to the component level back in the day). And also because the rest of my sound system build has or will cost me $$$$$. So, I have to be content with a combination of all three approaches, though mostly paths 2 & 3. That is, going with a proven design, which should therefore all but guarantee at least 90% happiness in every way and also b.) where I’ll be learning (or re-learning) enough to build a 2nd and very different pair not too far down the road (i.e. LOBs with Lowther, Supravox or other wideband drivers).

As to the Heart of a proven design, the only thing at the moment now etched in stone is the choice of woofer: GPA’s 416-8 Alnico (AND that I’ll be sure to ask GPA to measure and print out all T/S parameters before they ship my pair, as some 416B fan on this thread laments that the numbers on some actual product deviate “all over the place” from the spec sheet downloads).

The other thing that I’m sold on is sealed cabinets for these woofers. I was all but elated when g3dahl pursued this and had great sounding success with 416-8s in 3 cu. ft. boxes { “g3dahl used LARGE MDF edge rounds” }. I had a nice chat with GPA support yesterday over this among other issues and they were interested enough themselves in the “416B in a sealed box” idea and may run the numbers and/or perform some tests in this regard some weeks from now.

As per your 2nd post, f3 would be a bit worse than 37Hz if 6 cu ft sealed rather than ported cabs were used, as GPA support also warned. However, when I asked about applying a bit of EQ, they all but said that the 416Bs were designed to be EQed-though only by about 6 to 8db maximum (or risk hearing damage). I now recall that Nelson Pass and Jon Ver Halen got the 416-8s in the Lowther LOBs to go down to 40Hz with a touch of EQ, and most likely without any audible distortion.
Thus, the benefits with the 416Bs from carefully metered levels of EQ may at least not be specific to enclosure type.

Furthermore, I may have some limitations and advantages that might better enable me to exploit EQ, in the event that I would need any. For one thing, my current room is only ~ 14 x 7 x 13. And even when the system’s moved to a slightly larger room, my hearing is likely too sensitive to tolerate even moderately high SPLs (and I take every precaution, minute to minute, to keep it that way). Also, as previously mentioned, there is always 3db of apparent loudness gained from the presence of the 2nd speaker. And with most stereo recordings bass information is typically mixed equally between the channels.

At the same time, referring to the curve on p. 2 of the 416B’s spec sheet, in a 6 cu. ft cabinet (albeit a ported one), f3 is 92db. Unless I’m wrong in one or more ways, not only does that seem to be a sizable SPL-even though the human ear is less sensitive at audio frequency extremes-but there’s a reasonably steady rise in the 416B’s output to 98db @ 60Hz or 90Hz, and which remains on to about 500Hz. And what would ears like mine in my small room really need 98db SPL for at any frequency?

Without it, an overall lack of dynamic range and/or other deficiencies?

OR, without it, no issues that would be very noticeable?

If no, then would something like “1/3 octave EQ” attenuation be the solution of choice-or some other manner of filtering that would thereby allow me to trade the 416B’s relatively high upper broadband output level for a flatter end below 60 Hz? OR would the overall sound suffer any objectionable distortion or coloration in the process?

If yes to EQ, then could this be done just as painlessly with software? For example, a plug-in or standard utility in a DAW platform like Pro Tools or Samplitude http://www.samplitude.com/en/ (my main source that feeds my DAC is a micro-ATX desktop pc with a hybrid switch mode power supply, similar to what Nelson Pass used with his First Watt B4 crossover).

I know that a room’s dimensions and its total acoustical environment weigh substantially for or against any speaker system. And I may not have the test gear, the complete know how or the money to apply all of the corrective schemes.

But with the 416Bs in 6 cu ft sealed cabinets, it appears that my goal seems to be to
get them to yield a reasonably flat SPL of between 83 and 86db from 30Hz through their midbass and lower midrange. And that this performance would be for my ears and in my 14 x 7 x 13 room (or one maybe ~ 3ft larger?). Are those numbers most probably a desired and achievable goal in my case?

And if 6 cu ft sealed boxes alone won’t do it, then is EQ a respectable way to make it happen?

If so, can you recommend any specific hardware or software EQ solution to do this?

As you say “I agree with most folks here that a well stuffed sealed tuning sounds best.” But though I’ve been using sealed woofers since my Bozak days, IF you happen to think that the amount of EQ needed for this solution would likely be more sonically harmful than beneficial, I could opt to put the 416Bs in ported cabs (Btw, GPA Support mentioned that using two 2.5” ports rather than one 5” port might sound better). Still, since g3dahl seems quite pleased with his build, I’m confident that I could be happy with “uncorrected” sealed 6 cu ft cabs (maybe 5.5 cu ft if some issues arise with placement of the tweeter & midrange or wideband drivers). Thanks for your input and encouragement.

Richard’s gotten a whole lot more computer literate in the past year. Most of his clients call for specialty stuff related to home improvements. But everything he does for them and us is perfection. When he’s not freelancing, the last group he worked for probably has at least one CNC machine (Here’s a cool shot of one in action towards the end of this video http://www.msbtech.com/video/videoHome.php?video=analogDac ). And yes, the idea of turning out small hobby production lots has crossed my mind. If I can wow him with the sound of this build (fed from what I’d like to think will be a respectable front end)…..?? Those CNC machines can be rented of course, right? Who knows? Maybe by Christmas 2015.
 
I'm not sure applying EQ to the Altec/GPA 416 is a good idea. Unlike the vast majority of woofers made today, it has an underhung voice coil (a VC shorter than the magnetic gap). This type of construction is very linear provided the user does not demand long excursions which could drive the VC out of the gap. If long excursion is the goal, as required by deep LF, then a conventional overhung VC is a better choice.

In a direct radiator, excursion increases at a rate of 12 dB/octave as the frequency is lowered, so even moderate changes in the LF limit demand substantially more excursion. This is the underlying rationale for using the 416 in a moderate-volume closed box; it limits excursion, and VLF frequencies (below 50 Hz) are handled by a specialized, long-excursion subwoofer with its own dedicated subwoofer amplifier.

Applying low-frequency EQ to a bass driver does two things: increases excursion (possibly by several times), and also demands more amplifier power. This rules out low-power triode amplifiers in the 3 to 20-watt range, since a low bass note in the boosted EQ range would suck the amplifier dry. In other words, neither the bass driver nor the amplifier would perform satisfactorily in the EQ range.

Midwoofers (which is what the 416 is) and subwoofers are very different animals. The requirements of a smooth, peak-free lower-midrange and well-controlled breakup above 1.5 kHz are not the same as a very heavy, very stiff woofer cone, very long excursion, and the ability to absorb 500 watts of power. One is a midwoofer, and the other is a subwoofer. Although both are 15" drivers, almost everything else is different. (Analogy: one is a sports car, the other is a truck.)

It's true that Altec nearly always used the 416 in a large vented box, typically three to four times the volume of the closed boxes that Gary Dahl is using. In the Fifties, Altec, JBL, Bozak, and Electro-Voice didn't pay much attention to panel resonances and interior standing-wave modes, but it's a big deal today. It is not trivial to suppress resonances in a box this large, and the resulting box with all the interior crossbraces will be extremely heavy. Our moderator, Pano, has had experience designing and building very large enclosures, and it's neither simple nor light.

I'd rather band-split around 50~60 Hz than deal with boxes the size of a refrigerator. No thanks; this is thing is about as big as I want it to get.
 
Last edited:
Member
Joined 2012
Paid Member
Newbie Speaker Build: Part 1


Thanks a whole lot for the heads up on using EQ for bass extending the GPA 416. Again, I really only considered it because GPA support said that using it-though with great caution-is fine.

And also because Nelson Pass and Jon Ver Halen, who’ve had lots of experience with the 416 in LOBs, claim healthy bass improvement as well: “a small amount of bass boost was used,
so that no subwoofer was required. Bass was flat to approximately 40 hz.” Open Baffles Lowthers

In fact, the picture shows them being bi-amped with my First Watt J2 and F4 amps. Each amp is only 25 wpc into 8 ohms.

OR is this moderate bass boosting of the 416s only possible in LOBs and not in sealed boxes? If so, please explain why.

BIG question: How is the GPA 515C
http://www.greatplainsaudio.com/downloads/515_C.pdf
different from the 416
http://www.greatplainsaudio.com/downloads/416-8B Spec Sheet.pdf ?

The spec sheets show that the former’s Alnico magnet is twice the size and/or weight of the latter’s, and its LF response extends to 20Hz. So how would the 515C perform in a 3 to 5.5 cu ft sealed box?

And could it accept some degree of bass boost in those boxes to get down to 30 or 35Hz without penalty?

Also, the 416’s response goes to 1600Hz, while that of the 515C ends at 1KHz. Would it make sense to put a 16 ohm version of each in the same 5.5 cu ft sealed box, passively cross them and
drive a pair of them with my F4 amp FIRST WATT F4 ? How good would this work?


 
Last edited:
I agree with Lynn. The underhung voice coil makes the 416 a poor candidate for LF equalization. This applies to the 515 as well.

My 416's and sealed boxes combine to create a slightly underdamped alignment, providing a somewhat "rich" sound (before the addition of box stuffing). Increasing the box size a bit (or adding stuffing) yields critical alignment (Q=.707). Going past this point (larger box plus stuffing) pushes it into an overdamped alignment, resulting in a "lean" sound.

The GPA 515C is the basically the same set of soft parts (cone, voice coil, suspension) paired with a stronger motor. The bigger magnet lowers the Q (leaner bass), and also causes the driver to have an upward-tilted response...perfect for use in a suitable horn enclosure.

I believe that the results I am getting with the 416 are enhanced by the way I am using it. Because I am not asking it to play deep bass, the limited cone excursion minimizes IM distortion. Also, because I am not tuning it very low, I'm not getting that dreaded dip between the midrange and the bass; the whole range sounds as one. This makes it easy to integrate the subwoofer.

If I were to use my system without the subwoofers, I would probably increase the enclosure size a little bit, maybe to 4 cf. Reducing the Q just a little would be beneficial. But if you wanted to go further, a vented box would be most appropriate.

Gary Dahl
 
Member
Joined 2012
Paid Member
Newbie Speaker Build: Part 1

Thanks for setting me straight about the 515C.
But at the risk of being somewhat redundant, what about the 515-LF http://www.greatplainsaudio.com/downloads/515_LF.pdf ?

The spec sheet says “the Model 515-LF’s rear vented magnetic assembly enables it to operate extremely well in a variety of different enclosures, including direct radiator cabinets, bass horns of all types, as well as sealed enclosures.”

Of course, I wouldn’t know if it’s just as much the use of a ferrite-rather than the alnico magnet-as it is the “rear vented magnetic assembly” that allows the 515-LF to “operate extremely well” in sealed enclosures. But assuming that is the case, do you believe that the tonality below the midbass would not sound as good or the same as the 416, due to the 515-LF’s Ferrite magnet?

If the answer is no, and since I would prefer to avoid subwoofers, would it make sense to put a 16 ohm version of each in a ?? cu ft sealed box and passively cross them, where the 416 would handle the midbass and the lower midrange?

Might this set up achieve what I too agree are the most important goals: Limited cone excursion to minimize IM distortion; avoiding that dreaded dip between the midrange and the bass, so the whole range sounds as one?

BUT IF the 515-LF’s low bass will sound just as the 416’s-only at least a bit lower-then why not go for it? But will it?
 
Member
Joined 2012
Paid Member
Newbie Speaker Build: Part 1


Ooops! I just realized that a 416B and a 515-LF would hardly fit in a 4 cu ft box!
And-presumably-like the 416, the 515-LF’s low bass would also sound “thin” if put into a larger sealed box.

Therefore, what about:

The 515-LF in a 4 cu sealed box and the 416 (for midbass & lower mids) in its own sealed box sitting on top?

If yes, what’s the smallest size box that I could use for the 416 and still have it give perfect midbass?

OR the 515-LF and the smaller 414B
http://www.greatplainsaudio.com/downloads/414-8B%20Spec%20Sheet.pdf (for midbass & lower mids) in one sealed box?

If yes, what’s the right size box that for the 515-LF’s low bass and the 414’s midbass?

OR is the 416 alone in a 4 cu sealed box, crossed with a midrange driver still the best way to go? If yes, please explain.

 
Agreed and have proposed this combo on occasion, but to date no one has tried it AFAIK, so hope you go through with it.

GM

I've used twin 416a in a big Onken and a sealed 515B with TAD 4002. In my room it was OK. The system was triamped. I think the 515 is better on a front horn but not the best driver for that either. Maybe the 515 is best used in the old Altec cabinets with better bracing
 
oltos,

Since you are considering the Altec 515 / 416 ,I would look into some used jbl 2441 ( ebay ? )compression drivers and a pair of tractrix horns from DIY Sound Group and be done with your search. There are other good compression driver choices depending on your budget. Mount the horns on top of your Altec box and when you can, grab a pair of Fostex compression tweeters ( 900 series ) add a 6db crossover and you are off to a great start. Then you can kick back and listen to your tunes on a high efficiency, high quality speaker system for a reasonable amount of money. Just my opinion.
 
It would be cool if a YouTube video of this system could be put up playing some tunes and maybe a walk through of the design as it is so far in Gary's system :cool:



That's a pretty neat idea, but we'll need to find someone that's better with a camera than I am!

A small contingent of the "Sound DIY" Club made a trip to Gary's house yesterday to hear not only his speakers , but also other system changes.
I won't attempt to speak for the others, although there seemed to be general agreement.

First, I'll just make a general statement: "Gary's speakers sounded much better than they did last time."

The upper midrange/ treble was much improved by removing the ribbons. Part of that improvement was due to Gary's reconfiguration of the crossover to allow the horn to play the treble. To me it sounded very smooth (natural) with much better integration of the upper registers without, what previously had seemed to me to be subliminal sense of straining and even perhaps some conjestion while attaining those higher notes. This is just a very vague sort of opinion and if others object or disagree, they certainly won't get much of an argument from me.

I am content to say they do sound quite a bit better to me!

The other thing that Gary had to show us was his new Music Server that he recently built. "The Silent Server" is a design that "Sound DIY" Club member Renan Jeffereis designed for the DIY Club members. It's a fairly inexpensive, up to date design that Dr. Jeffereis came up with that uses a lot of the knowledge that he acquired while working at Microsoft.
Gary's build sounded quite nice indeed and he certainly seems very happy with it!

We played a fair amount of music while we were there. After departing, we had dinner at a Thai Restrant the Gary recommended and then drove to the Concert Hall to listen to the Bremerton Symphony Orchestra's wonderful program of Russian Composers (and Yes, Gary was right there on stage!).
It was a very enjoyable day for me and I assume that the others enjoyed it as well.
:D

Best Regards,
TerryO
 
Last edited:
Agreed and have proposed this combo on occasion, but to date no one has tried it AFAIK, so hope you go through with it.

GM

I will. Currently shopping for local area Northern Colorado wood shops with NC tooling and the ability to work with bentwood processes. The first bid came in close to $4000, so I'm still looking.

On a related note, I've written a brief show report on the 2014 Rocky Mountain Audio Festival, with a focus on commercially available high-efficiency speakers. Since the Ariel construction articles originally appeared in Positive Feedback magazine some 21 years ago, it seemed appropriate to send the show report to David Robinson, who's still editing the magazine.

High-efficiency speakers aimed at triode-amplifier owners are still kind of thin on the ground, long after the big splash made by Sound Practices magazine in the Nineties. When you remove speakers with DSP-processed prosound drivers, and the many whizzer-cone fullrangers with odd-looking backhorns, there's not much left in the high-end audio commercial sector.

More than half the rooms at the RMAF had compact stand-mount speakers with the inevitable 5~7" midbass driver paired with a Chinese-made ribbon, powered by very expensive transistor amps, and the much-despised digital music server as a source. Not that all the servers sounded bad (although some did), but you only get to hear the same tedious audiophile favorites over and over again, through dinky little speakers with big price tags.
 
Last edited:
On a related note, I've written a brief show report on the 2014 Rocky Mountain Audio Festival, with a focus on commercially available high-efficiency speakers. Since the Ariel construction articles originally appeared in Positive Feedback magazine some 21 years ago, it seemed appropriate to send the show report to David Robinson, who's still editing the magazine.
In that report, Lynn, the description of the sound in the exaSound room is where the action's at. Interesting that it used two elements of systems that were capable of producing convincing sound, that I heard at my last show, Bryston amplifier and Magnepan speakers. The sound there was not "lucky", it was just correct - and it seems staggeringly difficult for most people to get their systems to that point ...
 
The source material at exaSound was stupendously good; it's very rare at hifi shows to hear classical, choral and organ music at concert levels with no obvious flaws in the sound. The exhibitors usually select less demanding material for the obvious reason: it sounds better on speakers with rough spots in the response, and doesn't drive the little stand-mounters to their limits.

Parenthetically, the exaSound room was really small, about the size of a small guest bedroom. (The room photo was taken with a fisheye lens; I was sitting so close a normal lens would have shown only one speaker.) If I had the same setup in my living room, at 2.5X the listening distance, the Bryston amps and Magnepans might well have been driven to their limits. It all depends.

The gross nonflatness of many of the speakers at the show was appalling, and I regret to say the high-efficiency (whizzer-cone) speakers at the show did not portray the HE world in a favorable light. I was hoping for more HE exhibitors with good-sounding product, but there weren't many. It's probably still a small market, despite the oodles of tube-amp owners all over the world.
 
Last edited: