Altec 902-8B/511B Reunited!!!

[Guest " "]

Dean, Al, jc, Bob, whoever,

What do you think about this? I'm sure the values are way off, but it's the best I can do with those online calcs. I entered the K33 as 3.4 ohms (CT1526 specs) and 902 as 6.4 ohms (averaged between 6.1 and 6.7 ohms of both drivers). Also, this design would completely eliminate the autoformer and still attenuate the 902 the desired -9dB.

And yes, I know I changed my mind again from a 3rd order to now a 6th order. If someone could possibly work out the corrected values, that would be excellent! []

So I see you read the material.

I like this approach. put a diagram up with values, and ask for input pros and cons. I can try to find a 902 driver resonse curve to see if any zobel or notch filters are need. let me know.

What frequency is you diagram cut over at?

Did you pick up on the implied frquency distribution restriction of the 511 which was mentioned to begin at 7500hz.

Earlier you asked for a way to brighten up your projects high end. In light of the character of the 511 begining at 7500HZ, it may take going back to 3-way. Options at this performance level include jbl 2404, jbl 2405, t-350, st-350, and a few others.

BTW, there are some comments to consider 2nd and 3rd order xovers. Some reference to production use of this driver has even been cited. Keep in mind, most common production use was at 800/900hz, so 2nd and third order was a good fit. I'm thinking your limited as to how high you can go based on the upper end of the k-33. Anyone have an absolute value for this?

[chops]

So I see you read the material.

I like this approach. put a diagram up with values, and ask for input pros and cons. I can try to find a 902 driver resonse curve to see if any zobel or notch filters are need. let me know.

What frequency is you diagram cut over at?

Did you pick up on the implied frquency distribution restriction of the 511 which was mentioned to begin at 7500hz.

Earlier you asked for a way to brighten up your projects high end. In light of the character of the 511 begining at 7500HZ, it may take going back to 3-way. Options at this performance level include jbl 2404, jbl 2405, t-350, st-350, and a few others.

Of course I read it! [] BTW, the diagram is designed for a 600Hz HP/LP filter.

However, in a recent email from another member, I was reminded that to get a proper (closer) value on the woofer end, I should enter the K33 as a 6 ohm load, similar to what I think Al K does with his networks.

In fact, I remember Al saying something that even PWK got the woofer inductor wrong on most of the networks because he left out the 1mH inductance of the K33's voice coil. If this is true, and you set up one of those online calculators for a 1st order Butterworth filter, and enter 6 ohms for the woofer at 600Hz, the inductor value comes out to be 1.59mH. Add that to the K33's 1mH and you now basically have the 2.5mH inductor that comes on the Type B network.

So if this is the case, the actual crossover point in the Type B network for the K33 is more like 382Hz, again at 6 ohms, 1st order.

As for the 7500Hz rolloff of the horn, wasn't that referring to the 811B? I can't remember right now.

But anyway, the top end extension isn't all that bad now since I raised the crossover point back up to the stock 600Hz. Honestly, I can't believe how HUGE of an improvement it made on the top end.

If worse comes to worse, isn't there some kind of circuit that can be designed into the crossover to help with the top end? I think someone mentioned something about using a cap to bypass the autoformer?? [:^)]

[chops]

Also, here's a little snip from one of the PMs I got from Bob a while back concerning the K33 and/or Bob's own CT1526 drivers....

"

Charles,

Factory traces I have show the woofers dropping off very steeply

after about 1500 hz. That is a raw woofer with no inductor. With the

inductor, it should be done before 1 khz.

My CW1526 is essentially identical in performance to the Pre-1985 K-33E."

Now the only question that remains is how cleanly does the K33 sound up that high?

In Al K's notes, didn't he say there were rumors of PWK even running the K33 full out without an inductor at all? Not that I would do this because it would defeat the purpose of getting these 902 drivers for their midrange at that point.

[chops]

However, in a recent email from another member, I was reminded that to get a proper (closer) value on the woofer end, I should enter the K33 as a 6 ohm load, similar to what I think Al K does with his networks.

In fact, I remember Al saying something that even PWK got the woofer inductor wrong on most of the networks because he left out the 1mH inductance of the K33's voice coil. If this is true, and you set up one of those online calculators for a 1st order Butterworth filter, and enter 6 ohms for the woofer at 600Hz, the inductor value comes out to be 1.59mH. Add that to the K33's 1mH and you now basically have the 2.5mH inductor that comes on the Type B network.

So if this is the case, the actual crossover point in the Type B network for the K33 is more like 382Hz, again at 6 ohms, 1st order.

After thinking about this more and looking at some of my plots from a couple days ago, this is actually kind of confirmed!

With the 2.5mH inductor on the K33, that puts it somewhere near 382Hz. The 4uF cap on the 902 puts it up right around 600Hz. That would mean there would be a dip in the response between those two crossover points, right? (382Hz - 600Hz)

Now look at the plot I took the other day and you'll see a dip in the response between 200Hz and 600Hz! Am I getting close to something here or am I just going nuts?!

[Guest " "]

I was just curious what it was, if you have 600hz in mind, thats fine.

Yes, the inductance of the k-33 is a noted additional factor to consider.

Thanks for explaining the 300hz (382hz in this case) series numbers being used over the last few days.

The article was about the 511b, so I assumed the 7500hz issue was also about the 511b.

600hz vs lower xover point. Yes it will make a big difference.

Yes there are circuts to do tweaking. Zoble's, parallel notch filters, series notch filters, etc.

[Duke Spinner]

Duke,

Sorry, I had them on AuctionSniper... didn't know you wanted them.

Still bitter I missed bidding on them the first time he listed them... I forgot to. They closed without any bids at $249 with the 902s.

Shawn

I'm not mad, shawn .....

now i get to hear if thier a good short throw horn .......

...............for free ......[]

[sfogg]

Duke,

I kind of think the 594 will probably work out for short throw applications. There is a third one on ebay that I'm going to bid on since I buy in threes (L/C/Rs).

The little bit I played with the MR94s they sounded OK from pretty close up too. Still want to play around with them more. Big problem there is just how freaking huge they are. The load the 288 to way down deep though.

There are some interesting horns in the Mantaray series. For example I have some horns out of the Model 14 that have asymmetrical dispersion vertically. I think they are -10 degrees +30 degrees in an attempt to reduce floor bounce from the horn. There cutoff is around 1400hz though so they are more limited in what they could be used in. Would probably be a good horn for some wall mounted surrounds flipped upside down for better coverage to the listeners. They are pretty short too so they would fit in a cabinet easily. Someday I need to build something like that.

Did you see the Altec 31A on ebay? They could be fun to play with too. 300hz lower end and 120 degree dispersion horizontally. I'd love to hear 'em but I'm passing on them.

Too many horns, not enough money (or space).

Shawn

[chops]

Why are there 2 totally unrelated conversations going on in 1 thread?

[sfogg]

"isn't there some kind of circuit that can be designed into the crossover to help with the top end? I think someone mentioned something about using a cap to bypass the autoformer??"

Yes, a cap bypassing the autoformer *should* be able to boost the high end. But it would probably help to have Al simulate something like this first since I have never tried it.

You have what 9dB of attenuation through the autoformer? That basically means you have 9dB of possible boost available depending upon how you use it.

If you bypass the autoformer with a cap in effect you put a high pass filter bypassing the autoformer. As you go up in frequency more energy can bypass the autoformer which means you get less attenuation the higher up you go. You need to size the capacitor such that you get the amount of boost you want where you want it.

So for example... if you wanted your 3dB boost at 15kHz size the caps such that it was in effect a first order high pass at around 20,000Hz (-3dB point) bypassing the autoformer. At 20kHz the filter is down 3dB which means your end result would be to have 6dB of boost at 20,000 hz. (9-3dB) Half an octave away (15kHz) you would be down another 3dB through the filter... 9dB-6dB = 3dB through the cao. You would be letting 3dB of energy at 15kHz bypass the autoformer through the capacitor.

In theory anyway... the kicker may be the phase shift through the cap though....

[chops]

I understand what you're getting at, but I'm not sure on how to go about doing it. I'll have to bother Dean or Al for that trick. Thanks Shawn! []

Speaking of tricks, I just tried something to see how it works. At the moment, I have mixed feelings over the change. There seems to be a little more midrange, but whether that's a good thing or not is a different story. Also, due to the raise in midrange, the treble is now pushed back further again.

Oh, I guess I should let you guys know what I did! That would help! LOL []

Anyway, I shut down the system and did a quick rewire of the network. I moved the 902 over to the 3uF cap which I think should put it up around a 800Hz crossover point. Then, since I don't have any 1.3mH inductors laying around, I connected the K33s directly to the inputs so they get a full signal and roll off naturally.

I also took a few more readings and pics of the network mods....

And here's the plots.

1) This is before the change. 2.5mH to K33 / 4uF to 902 - Pink Noise - Sweet Spot - Stereo - +2dB boost at 10kHz

2) This is after the change. K33 run flat out / 3uF to 902 - Pink Noise - Sweet Spot - Stereo - +2dB boost at 10kHz

This is a high res. plot of the left channel @ 1m - Pink Noise - +2dB boost at 10kHz

[chops]

BTW, if you notice, there's a +2dB increase in output by letting the K33 run fullrange.

[Guest " "]

I was wondering what you were up to. I figured you were up to somenthing.....you dropped off the net for about 2 hrs.

[chops]

I was wondering what you were up to. I figured you were up to somenthing.....you dropped off the net for about 2 hrs.

LOL! That's bad if someone notices something like that! []

BTW, I think I've figured out what Shawn was talking about in regards to bypassing the autoformer with a cap.

If this online calculator is somewhat right, I would need a 0.12uF cap to get a 20kHz crossover point. I would then connect this from the main (+) input lead on the network to the (+) lead on the 902 along with the one coming from the autoformer. It's kinda like a reversed bandpass filter, allowing freqs to "seep" back into the 902, but this time from the extreme top end down at a -6dB slope which should bring up the level of the 902 on the top end.

Holy cow, did I confuse anyone there?! If you followed that, does that sound right to you?

[chops]

I was wondering what you were up to. I figured you were up to somenthing.....you dropped off the net for about 2 hrs.

LOL! That's bad if someone notices something like that! []

BTW, I think I've figured out what Shawn was talking about in regards to bypassing the autoformer with a cap.

If this online calculator is somewhat right, I would need a 0.12uF cap to get a 20kHz crossover point. I would then connect this from the main (+) input lead on the network to the (+) lead on the 902 along with the one coming from the autoformer. It's kinda like a reversed bandpass filter, allowing freqs to "seep" back into the 902, but this time from the extreme top end down at a -6dB slope which should bring up the level of the 902 on the top end.

Holy cow, did I confuse anyone there?! If you followed that, does that sound right to you?

Would these be what I'd be looking for if I were to do this?

Again, if that value of 0.12uF is correct, I would need 1) 0.10uF and 2) 0.01uF caps wired in parallel to get to the needed value, which would in turn get me to 20kHz?

http://www.partsexpress.com/webpage.cfm?&DID=7&WebPage_ID=70

[Duke Spinner]

Duke,

Big problem there is just how freaking huge they are.

.

Shawn

have you seen my 2360's ..???

[chops]

Duke,

Big problem there is just how freaking huge they are.

.

Shawn

have you seen my 2360's ..???

Can you PLEASE take your conversation elsewhere?!

[Guest " "]

I was wondering what you were up to. I figured you were up to somenthing.....you dropped off the net for about 2 hrs.

LOL! That's bad if someone notices something like that! []

BTW, I think I've figured out what Shawn was talking about in regards to bypassing the autoformer with a cap.

If this online calculator is somewhat right, I would need a 0.12uF cap to get a 20kHz crossover point. I would then connect this from the main (+) input lead on the network to the (+) lead on the 902 along with the one coming from the autoformer. It's kinda like a reversed bandpass filter, allowing freqs to "seep" back into the 902, but this time from the extreme top end down at a -6dB slope which should bring up the level of the 902 on the top end.

Holy cow, did I confuse anyone there?! If you followed that, does that sound right to you?

I follow it.

not having a drawing I have to ask, would the same effect occur if you use your values to bypass the autoformer at the autoformer?

there were some discussions earlier today with Al, and he introudced some really high ohms values that are involved in the autoformer in the way the autformer does its buisness.

your approach is intresting...if you wanted to widen the parametrics of your 6db slope, you could move the crossover point (larger value).

[chops]

I was wondering what you were up to. I figured you were up to somenthing.....you dropped off the net for about 2 hrs.

LOL! That's bad if someone notices something like that! []

BTW, I think I've figured out what Shawn was talking about in regards to bypassing the autoformer with a cap.

If this online calculator is somewhat right, I would need a 0.12uF cap to get a 20kHz crossover point. I would then connect this from the main (+) input lead on the network to the (+) lead on the 902 along with the one coming from the autoformer. It's kinda like a reversed bandpass filter, allowing freqs to "seep" back into the 902, but this time from the extreme top end down at a -6dB slope which should bring up the level of the 902 on the top end.

Holy cow, did I confuse anyone there?! If you followed that, does that sound right to you?

I follow it.

not having a drawing I have to ask, would the same effect occur if you use your values to bypass the autoformer at the autoformer?

there were some discussions earlier today with Al, and he introudced some really high ohms values that are involved in the autoformer in the way the autformer does its buisness.

your approach is intresting...if you wanted to widen the parametrics of your 6db slope, you could move the crossover point (larger value).

I'm not sure what you about bypassing the autoformer at the autoformer. [8-)]

But yeah, I'm really wishing Al or Dean would come on here and explain and/or walk me through this little trick. I'd really like to try it to see if it does what I want it to.

I would also like to know if I am really running the 902 at 800Hz right now or not also, being that I switched over to the 3uF cap.

Al, Dean... hint, hint... []

[Guest " "]

I was wondering what you were up to. I figured you were up to somenthing.....you dropped off the net for about 2 hrs.

LOL! That's bad if someone notices something like that! []

BTW, I think I've figured out what Shawn was talking about in regards to bypassing the autoformer with a cap.

If this online calculator is somewhat right, I would need a 0.12uF cap to get a 20kHz crossover point. I would then connect this from the main (+) input lead on the network to the (+) lead on the 902 along with the one coming from the autoformer. It's kinda like a reversed bandpass filter, allowing freqs to "seep" back into the 902, but this time from the extreme top end down at a -6dB slope which should bring up the level of the 902 on the top end.

Holy cow, did I confuse anyone there?! If you followed that, does that sound right to you?

I follow it.

not having a drawing I have to ask, would the same effect occur if you use your values to bypass the autoformer at the autoformer?

there were some discussions earlier today with Al, and he introudced some really high ohms values that are involved in the autoformer in the way the autformer does its buisness.

your approach is intresting...if you wanted to widen the parametrics of your 6db slope, you could move the crossover point (larger value).

I'm not sure what you about bypassing the autoformer at the autoformer. [8-)]

But yeah, I'm really wishing Al or Dean would come on here and explain and/or walk me through this little trick. I'd really like to try it to see if it does what I want it to.

I would also like to know if I am really running the 902 at 800Hz right now or not also, being that I switched over to the 3uF cap.

Al, Dean... hint, hint... []

I realize that you are not using a tweeter, and have put you horn on a different tap, but Just for food for thought. Assume for the purpose of explaining the bypass tweeking assume for a minute that you have a b-1 xover, look at the attached diagram.

In this example The diagram shows the tweeter connected at tap 4 and the squawker at tap 3. If this was your xover, you could put your .12uf cap across tap 5 and 4, you would also have to put a small inductor in line with the squawker. In you case, the in-line inductor would not be needed since you do not have a tweeter.

B-1.pdf

[Guest " "]

I was wondering what you were up to. I figured you were up to somenthing.....you dropped off the net for about 2 hrs.

LOL! That's bad if someone notices something like that! []

BTW, I think I've figured out what Shawn was talking about in regards to bypassing the autoformer with a cap.

If this online calculator is somewhat right, I would need a 0.12uF cap to get a 20kHz crossover point. I would then connect this from the main (+) input lead on the network to the (+) lead on the 902 along with the one coming from the autoformer. It's kinda like a reversed bandpass filter, allowing freqs to "seep" back into the 902, but this time from the extreme top end down at a -6dB slope which should bring up the level of the 902 on the top end.

Holy cow, did I confuse anyone there?! If you followed that, does that sound right to you?

I follow it.

not having a drawing I have to ask, would the same effect occur if you use your values to bypass the autoformer at the autoformer?

there were some discussions earlier today with Al, and he introudced some really high ohms values that are involved in the autoformer in the way the autformer does its buisness.

your approach is intresting...if you wanted to widen the parametrics of your 6db slope, you could move the crossover point (larger value).

I'm not sure what you about bypassing the autoformer at the autoformer. [8-)]

But yeah, I'm really wishing Al or Dean would come on here and explain and/or walk me through this little trick. I'd really like to try it to see if it does what I want it to.

I would also like to know if I am really running the 902 at 800Hz right now or not also, being that I switched over to the 3uF cap.

Al, Dean... hint, hint... []

I realize that you are not using a tweeter, and have put you horn on a different tap, but Just for food for thought. Assume for the purpose of explaining the bypass tweeking assume for a minute that you have a b-1 xover, look at the attached diagram.

In this example The diagram shows the tweeter connected at tap 4 and the squawker at tap 3. If this was your xover, you could put your .12uf cap across tap 5 and 4, you would also have to put a small inductor in line with the squawker. In you case, the in-line inductor would not be needed since you do not have a tweeter.

B-1.pdf