Xover design...

[aceinc]

Brian:

Do you do xover design for other people, or just for your speakers?

How do you go about xover design? Do you have a test chamber for measurments etc? What software tools do you use?

I am asking not because I want to do my own, but because I'm interested in having someone designing and perhaps building xovers for speakers I'm thinking about building.

I was looking at a multiple midbass with a single tweeter perhaps MMMMT. The drivers I am thinking of using would be PE 264-816 & 275-085. It looks like a crossover in 4khz range looks good on paper to tame the nasties above 5khz in the 264-816.

My choice of these drivers is simple, I have wanted build a speaker with a planar/ribbon tweeter, this one is fairly inexpensive and handles a fair amount of power. The reason for the midbass units, is I wanted something that would allow a reasonably high xover, was fairly small and had some bass. The reason for four of them is based on my understanding that 4 would increase the spl to match the tweeter and have the added benefits of more bass and in series/parallel would be 8 ohms, making the xover design simpler.

I worked with a fellow locally in South Florida that was talented, but we chose not to work together, he thought I was taking advantage of him, and I wanted prices to remain consistent between jobs.

At the moment I do not have the time to get good at xover design would be interested someone that would do a good job for a fair price.

Paul

[Brian Bunge]

Paul,

Sorry that I haven't responded sooner. I've been pretty busy at work and getting ready for my wedding next weekend. I think you'd be better served looking at some different drivers (especially the tweeter) but I don't have time to go into it right now. Give me until tomorrow to get all of my thoughts together and I'll give you a more in-depth response then.

[aceinc]

Congratulations on your impending nuptials...

I will wait patiently...

I was hoping someone might have had some experience with the inexpensive ribbon/planars... Perhaps they are inexpensive for a reason...

Paul

[Brian Bunge]

Paul,

I've seen the ribbons used in a few line arrays, and heard them in one line array, but that's the extent of my experience with them. From everything I've read the BG Neo 3 planar tweeter is a much better option. It costs a little more and is not as efficient but I think it's much better option. Also, I think you'd be better off going with something like the Dayton RS 5" drivers than the Tang Band 4". The Dayton is actually a lot closer to a 4" driver than a 5". Also, it's only a few dollars more than the TB but is a much nicer driver in my opinion.

I don't have any kind of chamber, as I do everything out of my home. I still have a lot of help with my crossover design but do have LspCAD and JustMLS and am learning to use both. As I learn to use JustMLS better I'll be taking measurements outside to eliminate the effects of the room and then import the data into LspCAD.

I wouldn't mind designing something along the lines of what you're talking about anyway so if you want to work together on something that would be great. But first I'd like to find out exactly what your needs and expectations are and we can go from there.

[aceinc]

Brian:

I was unable to find a Dayton 5" RS driver. What is the PE part #?

My design criteria is fairly fluid but basically I am interested in a thin speaker cabinet, perhaps floor standing, I want to play with a ribbon type driver, I would like something that would xover to a sub at 80hz. I think keeping the mid/tweeter xover above 2.5khz is a good idea.

I'm interested in drivers that don't need a lot of manipulation to be fairly flat. My concept for these drivers were TMMMM for the front & rear, and MMTMM for the center.

You mention that you heard a speaker with these tweeters in it, but didn't mention your impression... I assume it wasn't good, what was the problem with it?

Paul

[Brian Bunge]

Paul,

The driver is the Dayton RS125S-8 (PE part# 295-360). I think 4 of them would work pretty well with a 28L (1ft^3) enclosure. Assuming 3/4" material, I think a cabinet 38"H x 7"W x 12"D would work quite well. If you wanted to limit depth a bit the cabinets could always be made taller, but it really depends on what you want tweeter height to be.

Any special reason you prefer to keep the crossover above 2.5KHz? I know that the tweeter used might make that a necessity. My towers use dual RS 10's, dual RS 7's and the RS28a tweeter. My mid/tweet crossover is at 1.6KHz, but is extremely steep (around 8th order). For a center channel I'd recommend 2.5-way design so that you could eliminate comb filtering for anyone sitting off axis.

The line array I heard with the Dayton planar tweeter actually wasn't too bad. The main thing I didn't like about it was that the tweeter line wasn't long enough. It's been a good 3-4 years ago since I heard it though so it's hard to truly remember the sound other than them being very dynamic. I do know this tweeter was compared to the BG Neo 3, along with several higher dollar true ribbons. The BG was ruled the winner and the Dayton planar was considered last place.

[aceinc]

I like the Dayton RS 5", the specs look good, the size is only 5/16" larger, and the price is about $4 more per driver. That coupled with the low le, and the nasties don't crank in until about 12 khz it seems like it would work well.

Winisd shows a .64 cu ft, 17 litre box yeilds an f3 at 62 hz which should be acceptable with a sub.

What I don't like about the neo is the output is weaker, the power handling is 1/4, the impedance is 4 ohms, and when you add the mounting ring the cost is almost $20 more per driver. That coupled with my opinion that it is butt ugly makes it undesireable.

I suspect the neo would be good with two of the RS 5" drivers for a more refined system, but for home theater I want something I don't have to worry about the kids (My wife will say that I'm the biggest kid) cranking and frying...

I think a 12db crossover at 2.5 khz for this combination looks good on paper, might need something additional to attenuate the nasty hump at 12-15khz. Of course until the cabinets are built and the drivers are measured in the cabinets we really won't know.

I'm not familiar with a 2.5 way design, and only have a vague idea about the comb filter problem you are refering to. Could you point me to someplace that will educate me?

Paul

[Brian Bunge]

I use Unibox for my box designs most of the time and I find it much more accurate than WinISD. I think 17L is a bit on the small side but if you tune around 50-55Hz it doesn't look too bad. Oh, and what are you planning to use for a subwoofer?

A 2.5 way design is one where you have multiple woofers in a design. The upper woofer(s) cross over to the tweeter and play all the way down to the low frequency limits of the driver. Then your .5 woofer(s) will cross over at some frequency below the mid/tweet crossover and only reproduce the lower midbass/bass frequencies. Many times this is done to keep efficiency up by allowing the lower .5 woofers to handle the baffle step compensation.

I'll have to see if I can find something in particular about comb filtering. It's really an issue with short, wide center channels where the woofers flank the tweeter. The center to center spacing between the woofers ends up being so great that anyone that sits much off center of the tweeter will have a lot of trouble with vocal intelligibility. This is due to the listener being different distances from the two (or more) woofers and results in a very chesty sound, especially on male vocals. This is why I would suggest using a 2.5 way crossover so that only the two inner woofers are crossing over to the tweeter and reproducing the vocal range. Without doing this things would get very nasty for anyone sitting off center.

Here's an interesting center which is actually a 2.5.5-way using six 3" woofers.

http://www.partsexpress.com/projects...ure/index.html

Oh, and here's the article where the Dayton planar tweeter (same as the Silver Flute tweeter that Madisound sells) was compared to various other ribbon/planars. Read this before you decide you definitely want to proceed with this tweeter.

http://www.zaphaudio.com/nondomes/

[aceinc]

zaphaudio certainly has strong opinions about the silver flute planar tweeter. You'd think the designer made a pass at his wife & kicked his dog...

Paul

[Brian Bunge]

Yeah, he didn't care for that tweeter at all. Like I said, I only heard it once and that was in a line array several years ago. In that capacity it wasn't too bad. I'm just not sure how good it would sound with only a single unit running.

[aceinc]

Brian:

I finally built a center channel, totally different from what we had discussed. My original question in this thread is still important; Do you design, or help with the design of a xovers? Either for money, or out of the kindness of your heart.

I will be taking some measurements in a decidedly non anechoic room (I'll try to deaden things as much as possible with blankets and towels and the like) using REW for all the drivers individually in the cabinet. But I do not have anything that can give me impedance curves or anything like that. If you have any ideas on software or hardware that will help

When I get home I'll snap a few pictures so you can see what I'm dealing with.

The specs as I know them;

Cabinet is ~32" wide by 8-1/2" high.

There are 12 drivers in the cabinet, 4 - 5" 8ohm woofers, 4 - 3" 8ohm midranges and 4 - 5/8" 6ohm tweeters. All sections of drivers are connect in series/parallel, to hopefully yield their original impedance, as a section.

The 4 tweeters are stacked closely in the center, flanked closely by the mids stacked in pairs, and the woofers are outboard of that.

The mids are extended range, and should be capable of a fairly low xover point, (Hopefully this will limit comb filtering and mid range smear.)the tweeters need to be crossed over above 3,500hz.

The woofers are partitioned off from the mids/tweeters, and ported. According to winisd, I can expect < 50hz f3.

I am hoping for the improbable, a xover with a low parts count and good sound.

All input is welcome.

Paul

[aceinc]

I purchased a prebuilt xover for $5 on e-bay. It is a 3 way first order design with crossovers at 600 and 4500 hz. It was late last night after overcoming my last bit of silliness in cabinet construction, I had cut the holes too small for the mid's, the drivers would fit but when the wires were added, I found I had made the holes too small, I decided to see what the speaker sounded like. For $5 you don't get any documentation, so I hooked everything up, turned it on, and it sounded like I was listening to a speaker on a cheap modem.

I was shocked, so I pulled the woofers off the xover, and hooked them straigh to the input, and viola I had bass. So I played with various configurations, and didn't really get any thing that justified my effort. I finally hooked a cap to the tweets, and it sounded ok. I went to bed.

All day at work I was trying to find local folks with capacitor and inductors, with no luck. I posted the previous message and went home. When I got home I decided to look at the xover one more time. I am embarassed as the dickens to admit it, but I had hooked the input to the tweeter connections, and the tweets to the input leads.

I hooked the xover up correctly, and it sounded a lot better so I got the old Onkyo center, and did an AB compare. There was a lot more bass, but the mid was muddy, and too quiet. the Onkyo, while not great sounded better. After fiddling about some more, removing the crossover, and putting a cap back in, and it still was the same. On a hunch I got out the ohm meter and tested the three sets of leads, the woofers and tweets showed ~6ohms, the mids however showed 12ohms. For once I got a break, the first mid I pulled had a lead disconnected. I fixed that, and hooked everything back up, and it sounds great.

I decided to watch a movie and put off testing for tomorrow. As it is, it blends in well to the rest of the system. I will hopefully have some test sweeps over the weekend.

As I promised here is a picture.

[aceinc]

I am merging this thread with;

http://www.hometheatershack.com/foru...html#post35347

[Brian Bunge]

I still have help from someone else to do crossover design work for me. It's just not something I've had a chance to get into yet. It looks like you bought HiVi woofers and mids and maybe the Dayton neo tweeters. If so, you might be able to find someone on the PE forum who can help you out with XO design.

I'd be real interested to see how much combing you end up having with the driver arrays you have.

[aceinc]

Brian:

They in fact are HI-Vi & dayton. Is their an objective way to measure combing?

I don't believe that I have "head in a vise" syndrome, but it is to early to tell.

You can read more about the trials and travails of building this speaker in the other thread.

Paul