Saving a Speaker Design Disaster, Can You Help?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Good day,

I was in the process of designing a transmission line speaker for my self working around the demands of the spouse and the decor and have hit a stone wall. I personally do not believe that bass is an option for a full range speaker unlike many others in this hobby. Now before you jump to the conclusion that I am a ghetto buster, let me inform you that my music taste runs largely in the jazz to classic vein with some listening to an occassional pop or country artist when the mood strikes. I was a singer for many years and also play a number of musical instruments since I come from a very musically inclined family, and have first hand experience at what live music sounds like.

Here is the dilemna, I have to keep my wife happy by designing a pair of speakers the are decor friendly to her standards(I do really love her inspite of these demands) so, I proceeded to design and build along the lines of a TL. After building the enclosures I realized that I had made an error in my calculations and will only be able to tune the boxes to approx. 190Hz(yes you read right, big mistake!)or around 95Hz when stuffed since the line is only 36" or 900 millimeters long. The lines displace approx. 31 liters. The enclosures a octogan shaped with an inside dimension of 8" or 210 mil. I was going to use an 8" driver firing upward onto a dipersion sphere in which the tweeter would be mounted(see the Don Morrison speakers at http://www.donmorrisonaudio.com for the basis of this idea), the the TL would be downfiring so to speak through a vented base. So there is the beast that we are working with. After realizing that I could not salvage the situation as a TL, I checked on a the possibility of a passive radiator, no luck because of the downward position of the raditor. So I have now reluctantly decided try to make it ported, but I have no clue where to start.

I was planning to use the Peerless 8" 850136 woofer and would still prefer to do so. If not I will change to the Peerless 6 1/2" 850122. All the specs for both of these are available on Peerless website at http://www.peerless.dk

I would appreciate any suggestions or links that would further explain the thoery and procedures to build a ported speaker.

Thank you for time.
 
diyAudio Moderator Emeritus
Joined 2001
S3:

I went to the website trying to understand the internal construction of the speaker, but I must have missed something.

Anyway, you said the volume of the line was 31 liters. The speaker you chose, the Peerless 850136, tunes out quite nicely in a ported box of 31 liters. If you tune the box at 38 Hz., then you will be 3 dB down at 42 Hz.

42 Hz is lowest note of a bass guitar with normal tuning, and is not a bad cutoff point.

If you built the line straight, I would say that you would get a resonant "shot" at 400 Hz or so. Since it is folded, I am not sure that will happen. You probably will not need to stuff the line like a Transmission Line, but you should have some stuffing in there.

To tune the enclosure to 38 Hz, take a port with an internal diameter of 3 inches and make it 9.75" long.
 
What's the crossover frequency? What tweeter are you planning to use?

There's probably going to be a big problem getting any kind of coherent soundstage given the separation distance between the up firing woofer and the forward firing tweeter.

I give serious consideration to a using a coaxially mounted driver. Seas has some fairly nice ones.

BTW, tuning a ported box has nothing to do with the driver. It's totally a function of the size of the box and the desired tuning = port size. You can easily checkout any number of tunings quickly by using this online calculator
http://www.globalnode.com/users/stevenr/spkrs/
 
Thanks Kwizard,

Maybe I yelled wolf to quick. You are right, running the Peerless driver on Unibox 300 showed that it would tune quite nicely at 38Hz. I was thinking of using a 4" port 18 3/8" long which is according to the output of Unibox 300. I want to use as large diameter of port as possible to eliminate as much air turbulence and port noise as possible. What do you think of a 5" diameter 29 5/8" long?

A few clarifications, the line is straight 37" long and 32.55 liters empty. So figuring the drive and misc. volumes it should be around 30.25 liters with any stuffing or figuring out the wall volume of the port.

After runining this on Unibox 300 though I have found it is showing a massive impedance spike at around 70Hz to the tune of almost 55 Ohms!? and also another at around 15Hz at around 66 Ohms?! What do you think?

I have designed several seal boxes quite successfully I think(that is debatable), but this is my first venture into ported systems. Just for these reasons.

Thank you.
 
Thomas W,

I was thinking of crossing over at around 3000HZ to the Peerless 812978 Tweeter. It will also be up firing against a dispersion sphere. The frequency charts show the 8" is quite flat through 2000Hz and starts a rolloff on its own to 3000Hz. Crossover will be a combination 1st order Butterworth for the woofer and a 3rd order Butterworth for the tweeter. I got this idea from John Murphy in his article at http://www.trueaudio.com/st_mr2.htm - a walk on the wild side I know, but if I am not happy with the results I will use a 3rd order on both drivers, probably.

I know from reading your posts that you have a lot of experience in speaker design, so I would love to hear your opion on what I am about to do. I may be stubborn and still do the wrong way, but... well, at least I was warned!!!

As you can tell I am designing an omnidirectional speaker, at least I think that it what it would be classed as!!

Thank you
 
Hi

Well for any omni design I'd recommend a coax driver, just to keep the source time aligned.

IMO the correct way to design a XO is to measure the drivers in 'baffle' response, then decide on the XO points/slopes/types. Making an XO design based on someone's theory/ideas and or the mfgr's FR plots, just doesn't make sense.

just my $.02 :)
 
diyAudio Moderator Emeritus
Joined 2001
S3:

A ported box always has two impedance humps roughly equidistant from the tuned frequency. These are equidistant from 32 Hz-nothing to worry about. All it means is that at 70 Hz, your speaker will be equal sound output to the midpoint but will be drawing a lot less current. Hopefully, you won't be playing these much at 15 Hz, lol.

I take it you are going to have one of those pointed things pointed down at the woofer. Like Thomas W said, if you use a separate tweeter, even if you give it it's own pointed thing, it will be set off center from the woofer, which means that you have time alignment problems.

The only arrangement I can think of to keep things aligned is to have an arrangement where the woofer faces up, the tweeter is attached to a frame that allows it to fire down toward the center of the woofer, and a double sided pointed thing is between the two. Sort of a two shelf idea above the woofer cutout.

Otherwise, I guess you have to go coaxial.

KEF has coax units on it's UniQ system, but I don't know if the individual drivers are for sale.

I used a 3" port because it's diameter is larger than some of the ports I have seen Peerless recommend for it's 10" drivers. The larger the port, the less the port noise. So I figured it would be okay.

You must subtract the volume of the port from the volume of the box. If you use the 5" port, about a third of your box volume will be taken away by the port. If you use the 4" port, about 13 percent of the box volume will be taken by the port-which is not so bad. If you use the 3" port, it will be about 5 percent.

[Edited by kelticwizard on 10-26-2001 at 03:14 AM]
 
Kwizard,

I am confused(probably my ignorance) how if I center the tweeter over the woofer it will be off center? I am intrigued by your idea of facing the tweeter toward the woofer.

Let me try to get this picture clear to you, the woofer will have its own dispersion sphere centered over it and in the center of that the tweeter will be mounted. The centers of the drivers will be centered on axis. Yes, I will have to deal with baffle compensation and such either electronically or otherwise. Please if this will cause a time delay or be somehow off center, please explain.

I am open to learn anything I can. Yes I have considered using coaxials.

Thank you
 
diyAudio Moderator Emeritus
Joined 2001
S3:

Sorry, I did not understand what you were trying to do. If the centers of the speakers are lined up, then you should be aligned.

As for the crossover, I might point out that both these drivers are sold by Madisound, http://www.madisound.com. As of a few years ago, Madisound would do the crossovers if you bought the parts from them. You might want to check if they still do. I think they did it for free.

As I think the major dealer for Peerless drivers in the USA, it would not surprise me if Madisound has special software supplied by Peerless to cross over two of their drivers. You could ask.

Incidentally, have you considered the next step up in line in the Peerless 8 inchers-the 850490? It's the HDS line. It's about $20 more.

It has a longer excursion. The hump at 1500 Hz in the response, while slightly bigger, seems smoother and more easily taken out by an inductor. It has a better motor structure.

That aluminum shorting ring prevents the speaker from being "sucked in" to the back of it's travel as it nears the box tuning frequency, effectively clipping half the waveform! The progressive suspension of the 850136 helps too, but it is better to eliminate a problem at it's source. The bass response in your enclosure is virtually identical.

Just a thought.
 
Another issue is whether or not the suspension of the driver is designed for horizontal use, many aren't.

There's a calculator on the Adire site,http://www.adireaudio.com/Tech Papers/driver_orientation.htm This will give you a rough idea as to whether the driver will work, but it's best to contact the mfgr directly and get their advice. It would be a real drag to go to all this work only to have the suspension fail a few months after the speaker is completed.

Also the tweeter being centered over the top of the woofer won't do the job. Having the the 'acoustic centers' (sound originating elements) of the drivers aligned is what's important, and that's quite difficult to do without test equipment. Trying to align 2 separate drivers means a +/- tolerance of approx 0.5mm front to back. This is why coax drivers are a natural for this application.
 
diyAudio Moderator Emeritus
Joined 2001
S3:

If the top of your speaker is flat-I don't know what the omnidirectional guide looks like-then you might consider the following tweeter arrangements.

A) Affix one tweeter to the front of the speaker enclosure, (one of the eight sides), and one to the back. This will give dipole tweeter radiation, which should share some characteristics with the woofer's omnidirectional radiation. Also, if the woofer is centered, then the time should be aligned.

B) If the top of the speaker is flat, then make a small enclosure with two tweeters facing opposite sides, and put it atop the omnidirectional unit. This would have the advantage of being able to be moved back and forth so you can experiment with imaging, time alignment, etc., before you find the right positon to fasten it to the top. Since you seem to be into different shapes, (octagonal woofer enclosure), you should be able to come up with an interesting looking dipole tweeter housing.

Just spitballing a couple of ideas around that might be worth consideration. Let us know what you decide.

[Edited by kelticwizard on 10-29-2001 at 09:48 AM]
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.