• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

ALTEC 846B Trouble/week sound?

Theophile, I want to take a shot as to "why" the BMS driver and "why not" other drivers. I do not have them, nor have I heard them. I have no test equipment and am struggling to understand all of this new information. To me it appeers that the Altec drivers are lacking in HF extension above 10K. Some, in the past, have added a tweeter to help with this even though they may not have seen the need (according to a chart) or even been informed of the deficiency. Some have heard them and "subjectively" decided that something was missing or not to their liking. The choice of replacements/upgrades should be made the same way a manufacturer decides to use a specific driver, because it posseses certain attributes that are thought to be of value to the design and help attain the desired results. The BMS drivers offer measured performance specs and seem to be a good starting point for such an endeavor. The test data is just a reference from which to build from. :thmbsp: YMMV
 
macaltec said:
The choice of replacements/upgrades should be made the same way a manufacturer decides to use a specific driver, because it posseses certain attributes that are thought to be of value to the design and help attain the desired results. The BMS drivers offer measured performance specs and seem to be a good starting point for such an endeavor. The test data is just a reference from which to build from. :thmbsp: YMMV
Click to expand...
Yes, exactly.

I chose the driver based upon my best judgement of how the Valencia's issues might be resolved within the specific design constraints of the project, such as keeping the stock crossovers, no mods to the horn mounting, reasonable cost, sensitivity compatible with the stock woofers, the requisite compensation, etc., all from analyzing the specifications. I was also already familiar with the driver "family" and how it might sound, comparatively.

[Speakerlust would doubtless have used a different approach.... :p: ]
 
Hello Zilch

I have a question for you. Of all the combinations you have available which sounds best to you?? I am going to guess at least one is a set-up using the JBL 2435 Vertec driver using a Beryllium diaphragm. There's a reason for that and it's not flat response out to 20K. I can set-up a 2426 on a 2344 horn and a 2435 on the PTH1010 waveguide with almost identical frequency and polar response and yet the 2435 is the hands down winner. It's not always obvious from measurements which driver will sound better and even then we may not agree which one that is. That's just the way things are.

Rob:D
 
Robh3606 said:
I have a question for you. Of all the combinations you have available which sounds best to you?? I am going to guess at least one is a set-up using the JBL 2435 Vertec driver using a Beryllium diaphragm.
Click to expand...
Hi, Rob!

You'd guess correctly. The Be diaphragm purportedly operates pistonically out to 18 kHz in 2435HPL, and the difference is audible. "Pushing" an undamped titanium diaphragm as in 2426H, the VHF is generated from operating the diaphragm in breakup mode. At lower frequencies, we'd be calling it "distortion."

Aluminum diaphragms with their inherent damping handle it perceptibly better, and the Mylar ones, as well, though I have not seen laser interferomety of either of these in operation. Mylar's used in supertweeters operating at 30 kHz and beyond.

Of all of the drivers I've evaluated here, the large-format 4" aquaplas-damped titanium-diaphragm JBL 2452H-SL is the best sounding. I don't have any Be TADs to compare, but I'd bet they'd be close. I should have some aquaplas'd Be 2435s here to try soon.

There's no denying diaphragm material and driver design play a significant role in this, which cannot be ignored. Push any driver beyond it's limits trying to force out VHF out and it'll sound like crap. Yup, guaranteed. :yes:
 
There's no denying diaphragm material and driver design play a significant role in this, which cannot be ignored. Push any driver beyond it's limits trying to force out VHF out and it'll sound like crap. Yup, guaranteed.
Click to expand...

Hello Zilch

You can't push a driver beyond it's limits trying to force VHF out of it. That's not how compensation works. You attenuate the midband response. If the VHF is not there you can't put it there no matter what you do. If the VHF is rolled off too much no amount of attenuation will give you a usable response. The distortion levels are not going to change using attenuation to level out the response.

You can however significantly increase the distortion levels by trying to run these drivers too low or using a horn to its cutoff frequency. The lower you go the greater the diaphragm displacement and this increases distortion. If you use a horn to cutoff the diaphragm can unload and you will get more displacement.

"Pushing" an undamped titanium diaphragm as in 2426H, the VHF is generated from operating the diaphragm in breakup mode. At lower frequencies, we'd be calling it "distortion."
Click to expand...

How do you think they get the 2452 to have usable response out to 20K?? Damped or undamped it's the controlled resonances that are built into the diaphragms that get you the last octave. The aquaplas damps and lowers the spurious and unwanted modes. If anything the 4" diaphragm on the 2452 will have more issues with these modes than the smaller 1" drivers simply because the diaphragms used in the 1" drivers are much smaller in comparison. With the 2426 they are break-up free to almost 15K. The first mode in a 4" driver will be significantly lower and more into the audible band. If you want to remain pistonic stick to the beryllium drivers.

Rob:D
 
Hi, Rob.

I always get into trouble using "push" or "boost" in descriptions of how it's accomplished. There's no power added directly in the VHF, typically, rather the midband is attenuated, as you point out.

However, if the VHF is down 20 dB, and I attenuate the rest of the frequencies to balance with that, then 20 dB more power must be applied to restore the overall SPL up to the original level, with the VHF receiving the fullest measure of that increased power. Whatever the distortion component of the VHF out there is increases accordingly.

In 4430, JBL stopped at ~14 dB at 16 kHz. "Push" 2426H harder/higher than that, and it sounds nastier, is what I'm saying.

Altec Model 19, it's more like 9 - 10 dB, i.e., not as aggressive.

BMS 4550 requires just 7 dB, and less at VHF.

I don't know what's happening with 2452H-SL. There's no specs or technical information available, and the PWT FR response curves are "not for distribution." As you say, the VHF is coming from somewhere and the FR curve is certainly different. I'm told the phase plug is a revised design and a different material.... :dunno:
 
Hello Zilch

However, if the VHF is down 20 dB, and I attenuate the rest of the frequencies to balance with that, then 20 dB more power must be applied to restore the overall SPL up to the original level, with the VHF receiving the fullest measure of that increased power. Whatever the distortion component of the VHF out is increases accordingly.
Click to expand...

Actually it doesn't. You have to remember what your doing. There is no extra power applied when you are normalized for flat response. You don't change the sensitivity of the driver and horn, which in a 4430 as an example is 107db one watt. You attenuate the midband response. If you take a 2344 with a 2426 normalized for 93db for one watt, like in a 4430 to match the woofer, the actual power applied to the compression driver at say 2k for that 93db on axis is actually 13db down, to make it easy, or 50 milliwatts due to the attenuation from the network.

The power input at 16Khz is 1 watt for the 93db reference because there is no padding there. Look at the attached Voltage drive courtesy of Giskard for the 4430. You can see that the compression driver is attenuated at the crossover point. You have to subtract the attenuation from the power going in. That's what the compression driver really see's. You are not adding power to make up the original sensitivity. The distortion now is still that same as it was before you added the compensation circuit.

If you run 20 watts into a 4430 your going to get 106db @ 1 meter that's damn loud. I don't think many of us listen that loud. That gets you back into the area of the original sensitivity of the horn so at 106Db the driver see's 20 watts worst case only at the top of it's bandwidth well within it's power rating. How much music and actual power content is in the 10-20k bandwidth??

Rob:D
 

Attachments

  • 51-5334%20versus%2052-5130.gif
    51-5334%20versus%2052-5130.gif
    15.3 KB · Views: 13
Alas, sensitivity is not measured at 16 kHz.

2426H on 2344(A) is 107 dB (uncompensated, presumably,) representing the average between 1 kHz and 4 kHz:

http://www.jblpro.com/pages/pub/components/2426.pdf

To accomplish the requisite compensation for "flat" according to the posted network voltage drives, that gets attenuated 14 dB to 93 dB.

At 16 kHz, there is no compensation, so, in effect, the HF has been "boosted" 14 dB for any given output level, dB being relative as used here, not absolute.

Distortion is not specified for the driver alone, but it IS shown for that driver on the 2344A horn:

http://www.jblpro.com/pub/obsolete/2344A.pdf

It looks fine there, but both the distortion and the frequency response are shown only for the uncompensated condition. It's obvious from that curve why 16 kHz was chosen for the HF limit. It'd take an additional 14 dB to play it flat to 20 kHz.

Finally, in the 4430 specs, we see the distortion in the compensated operation of that driver and horn, clearly illustrating the consequence of the compensation in Fig. 15:

http://www.jblpro.com/pub/obsolete/443035.pdf

4430 sounds fine, of course. My thesis is merely that pushing the response flat out to 20 kHz would be problematic. The usable limit is defined by the horn/driver response curve....
 
Alas, sensitivity is not measured at 16 kHz.
Click to expand...

Yes it is, its 93db at 1 watt :banana:

At 16 kHz, there is no compensation, so, in effect, the HF has been "boosted" 14 dB for any given output level, dB being relative as used here, not absolute.
Click to expand...

What's important is the absolute. The absolute tells you how to accomplish what you need to do.

Finally, in the 4430 specs, we see the distortion in the compensated operation of that driver and horn, clearly illustrating the consequence of the compensation in Fig. 15:
Click to expand...

Yes we do and it's relevant to any compression driver using a CD horn.

4430 sounds fine, of course. My thesis is merely that pushing the response flat out to 20 kHz would be problematic.
Click to expand...

If it sounds fine why would you want too?? If it's not broke don't fix it.

Rob:thmbsp:
 
Here's 802-8G on Clio..................... Your mileage may vary

Not my analysis, nor my drivers, you can currently find them on that auction site.

802-8Gonclio.jpg
 
bowtie427ss said:
Here's 802-8G on Clio..................... Your mileage may vary

Not my analysis, nor my drivers, you can currently find them on that auction site.

802-8Gonclio.jpg
Click to expand...

That's actually not half bad looking.
 
COOL, Bowtie. Thank you!

Looks kinda familiar, no?

From Italy, he says that's the response on a 1 kHz Tractrix horn.

And here's his impedance plots, with a clear resonance peak at 500 Hz:
 

Attachments

  • 96fc_12.JPG
    96fc_12.JPG
    30.3 KB · Views: 8
  • 9831_12.JPG
    9831_12.JPG
    31.6 KB · Views: 7
macaltec said:
That's actually not half bad looking.
Click to expand...
Compare to Gary's GPA-refurbished pair, which measure a bit better than those, actually. The VHF response is also better on 811B horns than those particular Tractrix:

http://audioheritage.org/vbulletin/showthread.php?p=160605#post160605

Italian seller's citing the measurements as evidence they're "matched."

Not all that well matched, actually.

[I'm certainly pleased you guys are getting this.... :thmbsp: ]
 

Attachments

  • 811B 802-8G3.jpg
    811B 802-8G3.jpg
    52.6 KB · Views: 658
Now what would happen with 2 average cone speakers against each other? Would they also show the differences of 3-5dB like these 802s in the plot above?

I really don't know but that would explain a few things to me in my own little world. Like why one of my drivers just sound louder than the other, same with speakers. Maybe my hearing ear to ear isn't quite as bad as I thought.
 
Cones do better:

http://audioheritage.org/vbulletin/showthread.php?p=120126#post120126

Compression drivers require tolerances to be tightly controlled; the smallest deviations show up in the response:

http://audioheritage.org/vbulletin/showthread.php?t=6050

http://audioheritage.org/vbulletin/showthread.php?t=6368

There are exceptions. Some vintage designs are more forgiving. JBL LE85 comes to mind:

http://audioheritage.org/vbulletin/showthread.php?p=159517#post159517

http://audioheritage.org/vbulletin/showthread.php?p=135642#post135642

Contemporary manufacturing methodology does a better job building quality compression drivers:

http://audioheritage.org/vbulletin/showthread.php?p=147389#post147389

Scroll on down to see a dozen more.

BMS (a newer design) can get them so close you have to blow up the plots to differentiate them:

http://audioheritage.org/vbulletin/showthread.php?p=159367#post159367

Compare to the vintage Altec 806A pair on that same plot (red, orange).

:scratch2:​
 
Anonymous Testing

jackgiff said:
Those 806A plots look awfully familiar. Are you exposing my 806A's, Zilch???
Click to expand...
I've got a pile of them here, Jack, and as a matter of discretion, I don't name names.

You'll have to cross-reference the numbers, 'cause I forget which are whose, now.... :p:
 
You must log in or register to reply here.
Back
Top Bottom