ALTEC 846B Trouble/week sound?

[bowtie427ss]

That's most impressive............... so much so that i'm going to scratch up the admission fee and see the show!

Best response graph i've seen until now is of the PSD2002 on the 811 w/Pi Speakers 1K6 XO (see attached), but the 1600hz bottom end limitation just won't do.

 

[Zilch]

That's most impressive............... so much so that i'm going to scratch up the admission fee and see the show!

COOL!

Both BMS and 811B are usable down to 800 Hz, but 1.2 kHz just sounds better. Use 511B to take it to 800 Hz authoritatively.

BMS is clearly the winner above 14 kHz, not that I can hear much of anything up there.

[Not with my ears, anyhow.... ]

Notes:

1) Those are metric threads to mount the driver. My hardware store had metric studs in various lengths, so that's what I used. I don't think the size is what's stated on the data sheet, tho.

2) Do NOT solder to the terminals; use the correct-size Fastons. They are not mechanically supported other than by the plastic back cap, and if you get the terminals very hot, they will pull out.

 

[bowtie427ss]

1) Those are metric bolts to mount the driver.

Just reading about that........ i'm a semi-retired machinist so i have metric hardware on hand.

and if you get the terminals very hot, they will pull out.

Voice of a bad experience?

Polyester diaphram eh?

After nearly 60 years of aluminum it stands to reason an improvement would be found.

Have you had a BMS 4550 apart?

My assumption is they're pretty familiar aside from the diaphram.

Can i make their diaphrams fit my altec drivers?

I was going to PM my questions so as not to go OT, but then decided others may share in the curiosity.

 

[Zilch]

I have had them apart, yes. Some details are here. The conical phase plug is part of the diaphragm assembly:

http://www.assistanceaudio.com/02_bms.html

If you go to the BMS website, there's a link to their patent showing more of the mechanical details:

http://www.bmspro.info/

You cannot retrofit these diaphragms into existing Altec drivers. Their different technology, "Re-entrant," is very old, actually. The diaphragm plays to the rear of the driver, where the sound is radially summed and fires back out the front from the center. As a machinist, you'll find the design of it intriguing, very likely, and clever. You can see the radial soundpaths through the transparent diaphragm in the pic there.

Yes, I've melted the terminal mounting on ONE BMS driver in the past.... :

The Mylar diaphragm is inherently damped, so the output is smoother than the familiar metalics. To me, they sound more natural. First time I played them flat, I recall thinking there was something "missing." After a bit of critical listening, I concluded that what was missing was the trash of parasitic resonances.

I cannot listen to undamped titanium anymore. Aluminum is still tolerable, but it's better damped, also. My very best-sounding drivers here are damped titanium and beryllium, but their cost is prohibitive for these sorts of retro-DIY pursuits. The BMS Mylars are right up near the top, tho....

 

[Steve O]

Well, for ~$250 the pair, you can be the first AKer to try this in their existing sytem, and then tell the forum at large whether the Zilchster's got his head up his ass or not.... :thmbsp:

[I don't sell none of this stuff, folks.... ]

Based on the plot legend, I assume that plot below was achieved using 811 horn/BMS4550 driver thru M19 xover (N1201A-8A) and that a parallel notch filter (0.75uF||0.8mH||6ohm) is in series with hi-pass output and HF driver. Did I get that right? By my quick calcs, the notch is resonant at ~7kHz. If so, what anomaly is being corected by this. In previous plots of 811/4550 response I didn't see anything around 7Khz that needed correction. Insights appreciated.

 

[bowtie427ss]

The diaphragm plays to the rear of the driver, where the sound is radially summed and fires back out the front from the center. You'll find the mechanics of it intriguing, very likely, and clever. You can see the radial soundpath through the transparent diaphragm in the pic there.

Lots of interesting reading there, particularly interesting are the large format coaxial drivers............ many thanks:thmbsp:

It'll be a week or so but i'll be ordering up a pair of 4550's, very anxious to hear them in a couple different applications

 

[Zilch]

Did I get that right? By my quick calcs, the notch is resonant at ~7kHz. If so, what anomaly is being corected by this.

It's a series notch filter across the driver to knock down the 6 dB, 19 kHz peak the BMS drivers seem to all have. I developed it empirically, and have yet to put it into Spice to confirm the values. Is it close to correct?

http://audioheritage.org/vbulletin/showthread.php?p=160865#post160865

 

[Zilch]

Lots of interesting reading there, particularly interesting are the large format coaxial drivers............ many thanks:thmbsp:

Well, part of the price is that you post your findings here.

[We'll want plenty good PICS, of course, of what you do with them.... :huge: ]

 

[Steve O]

Close to correct is relative...

It's a series notch filter across the driver to knock down the 6 dB, 19 kHz peak the BMS drivers seem to all have. I developed it empirically, and have yet to put it into Spice to confirm the values. Is it close to correct?

http://audioheritage.org/vbulletin/showthread.php?p=160865#post160865

After checking out the link you provided I think I now understand what you're doing. You're using a series resonant filter parallelled across the BMS driver, not a parallel resonant filter in series with the driver. This is after passing thru the hipass portion of the N1201-A, right?

As far as being correct, the math would point to 1.581uF + 0.395mH + 8ohm as values that resonate ~20Khz. This is consistent with what you posted on the LH site a day or two ago:

"Xover3P says 1.581 uF, 0.0395 mH, 8 Ohms. Tried that several days ago with another crossover and it didn't work. That's when I checked with active EQ; cranking the 20 kHz band down 6 dB made it flat, so I know it's do-able with the right values".
The values you're using now work out to a notch centered on ~7KHz. I wouldn't have ended up with this particular set of values as a starting point myself but I sure can't argue with the results you got either. Bears further investigation.

 

[Zilch]

After checking out the link you provided I think I now understand what you're doing. You're using a series resonant filter parallelled across the BMS driver, not a parallel resonant filter in series with the driver. This is after passing thru the hipass portion of the N1201-A, right?

Yes, that's correct, but I believe both of us have slipped a decimal, 0.0395 mH, originally, no? And the final value was 0.08 mH, not 0.8; my legend is wrong on that value, sorry. Does that calc out any better?

In any case, I started with the calculated values, using 1.6 uF, 0.04 mH, and 8 Ohms, and tweaked from there. As you say, there's something going on I don't yet understand. I'll measure the driver parameters up there, and also see what Spice has to say. It'll be the weekend before I can get to that, tho. I ended up with about half the capacitance, twice the inductance, and 3/4 the resistance, but it's certainly working.

Edit: Forgot that I already had the circuits in Spice from originally doing the sims last October. Voltage drive for what I've got now appears below, and seems to be familiar and correct, or at least, as intended....

 

[bowtie427ss]

Well, part of the price is that you post your findings here.

I will happily oblige.................

It'll be a while before i build my N1201 clones, for the time being I'll be running active and i'd like to simulate the compensation circuit thru the use of PEQ in my digital crossover, can you tell me the parameters of the compensation circuit?

I understand i need to cut 6db at ~19khz, what kind of slope, or how wide or narrow should that filter be?

Sorry for the inane questions.................... best to get them outta the way now.

 

[dgwojo]

I understand i need to cut 6db at ~19khz, what kind of slope, or how wide or narrow should that filter be?

Question, should anyone be concerned about this peak at 19khz, I don't know about you but I can't hear crap at much above 15khz, I guess our pets will appreciate this?? :scratch2:

 

[Zilch]

I'll be running active and i'd like to simulate the compensation circuit thru the use of PEQ in my digital crossover, can you tell me the parameters of the compensation circuit?

The voltage drive curve is attached to post #70. That'll get you close. Ignore everything above 20 kHz; that's just me and Steve determining what's going on.

Post your parameters when you get something close, so we can all tweak with them. I haven't tried contouring with EQ on this one. Concensus seems to be it's best to do the compensation passively, even when actively biamped.

Two notch filters will render the driver listenable; it is well-behaved in this respect. AutoEQ can get you there in about five minutes flat....

Question, should anyone be concerned about this peak at 19khz, I don't know about you but I can't hear crap at much above 15khz, I guess our pets will appreciate this?? :scratch2:

There are theories as to how we perceive frequencies above what we can actually hear. Without even getting into that thesis, I made the decision to pursue it after looking at the distortion curves. The second harmonic tracks that peak, telling me that it's an artifact deserving some attention.

I can't hear whether it's there or not, but I'm an objectivist when it comes to the fundamentals; the curves don't lie.

Also, the fair sex hears higher than us knuckle-draggers. They count in this....

 

[dgwojo]

There are theories as to how we perceive frequencies above what we can actually hear. Without even getting into that thesis, I made the decision to pursue it after looking at the distortion curves. The second harmonic tracks that peak, telling me that it's an artifact deserving some attention.

I can't hear whether it's there or not, but I'm an objectivist when it comes to the fundamentals; the curves don't lie.

Also, the fair sex hears higher than us knuckle-draggers. They count in this....

I hear you, no pun intended, I've heard the symbiotik diaphragm 808's and although they claim to hit 16khz they really seem to lack the crispness of the aluminum phragm 802's that go a little higher. Yes, many woman have hearing like cats!!

 

[dgwojo]

Zilch, I wish I lived closer to you, it'd be really neat to pit the 802-8G alnico against the 902-8B ferrite magnet drivers, as much as I miss the shielding of the 802 the 902 still sounds best to my ears, maybe the 18,000 gauss of the 902 vs the 15,250 gauss of the 802 makes the difference??

 

[Steve O]

Now it makes sense...

Yes, that's correct, but I believe both of us have slipped a decimal, 0.0395 mH, originally, no? And the final value was 0.08 mH, not 0.8; my legend is wrong on that value, sorry. Does that calc out any better?
snip....

Yes, the 0.08mH value makes all the difference. Resonant freq is now calcs out to ~20KHz, just as your response plot and spice sim show. Your tweaked values produce a sharper "peak" than the values originally suggested by "Xover3P" and this makes sense too considering the relative sharpness of the 19KHz peak in the BMS4550 response.

FWIW, I used 0.0395mH in orig calc but can't type so that's where 0.395mH comes from. We need better man/machine interface.

 

[Zilch]

Your tweaked values produce a sharper "peak" than the values originally suggested by "Xover3P" and this makes sense too considering the relative sharpness of the 19KHz peak in the BMS4550 response.

Thanks, Steve.

Well, it's all good, then. :thmbsp:

Whenever I diddle in Spice, it gives me other ideas, but I'm done with this particular aspect for now. The results are good, so I'm not gonna mess with it. There's more benefit to be achieved "modernizing" the overall crossover design....

 

[Zilch]

Zilch, I wish I lived closer to you, it'd be really neat to pit the 802-8G alnico against the 902-8B ferrite magnet drivers, as much as I miss the shielding of the 802 the 902 still sounds best to my ears, maybe the 18,000 gauss of the 902 vs the 15,250 gauss of the 802 makes the difference??

If you want to send them out here, I'll be pleased to measure them for you.

I've tested 802, 804, and 806, the 802-8Gs being newly refurbished by GPA. You've seen the results. I can make any of them play full-range with compensation, but that's not an option for most users. There's also issues as to how hard one can "push" the VHF without the sonic quality deteriorating. It'd take a far more sophisticated crossover to have them approach what is easily achievable with modern drivers, and, in my view, much more effort than they warrant....

 

[Richard C.]

Which Altec diaphragms are being tested?

Zilch, Your testing has made me like my Altec 811's even more, now! A little damping and they won't ring.

Do you know which diaphragms were in the 802-8G that you tested? The original's are supposed to be #23744 which are light, fragile, and supposed to sound better. They are also very delicate. I have these in my Model 19 802-8G's and I had to repair the voice coil leads.

There is a normal one now (#34647), that is about the only one you can buy new from GPA. I also have a 902 light diaphragm #35480 which I use in my model 14's.

I don't use the black loading cap either, as I use a steep crossover slope (2 or 3-pole) and limit them to 1.2 kHz. I use felt in the aluminum cover.

The 902 drivers also come either, open thru the voice coil gap region to a small inner chamber, or with a blanking plate under the gap that shunts off this small chamber.

Lots of possibilities with Altec drivers.

Richard C.

 

[Zilch]

Zilch, Your testing has made me like my Altec 811's even more, now! A little damping and they won't ring.

I am certainly enjoying them, and the 511s, also. They are top tier. :thmbsp:

[And there's no shortage of horns for comparison around here....]

Do you know which diaphragms were in the 802-8G that you tested?

The diaphragms are new "standard" ones installed by GPA. I also have a vintage "light" one, which I'm going to try in an older (non-tangerine) 802 motor.

Problem is, the "light" is NLA, of course, even if it behaves perfectly.

With all of the variants, the Altec approach is thus far a bag of snakes. I sure wish someone would send me some Altec drivers that measured well so I'd have an answer that I can recommend....

Lots of possibilities with Altec drivers.

Right.