Some Thor models to mull over

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
diyAudio Moderator Emeritus
Joined 2001
I have a question about the ML TWQT worksheet.

Epsilon-driver position ratio. They don't have that on the Ported Box Worksheet or the Sections worksheet.

My understanding is that the driver position ratio is:

Epsilon (e) = length of driver center to top of box / total length of Line.

Is that the case?
 
A few suggestions for anyone starting to use Martin's seminal MathCad sheets for the first time. This is not a design proceedure I advocate per-se, but it's very useful for people unfamiliar with how these enclosures behave in learning how the different vaiables affect the response, and each other.

OK, remember, try if you can for a very slight downward slope below 100Hz. It's usually best to try for a flat response, and then modify the vent dimensions (width / length) to achieve this, if you haven't already hit upon it.

Try to avoid narrow vents if possible: there's greater risk of port-noise. A wider, longer vent is always my first choice, and 99% of the time, it's better to mount it a few inches up on the front panel (4"-5" is usually quite good). Don't be overly ambitious about how low to go: it's better to have a decent response than try to squeeze an extra 5Hz from a driver, and don't use too much stuffing. 0.5lbs ft^3 at the absolute maximum. A low Q driver (anything below around 0.3Qts) will probably need some series resistance to work properly.

Be systematic when you start out. Set an arbitarty, but sensible line-length. 40" is about right as a starting value for most mid-bass drivers.

Start with a small area for So and Sm (say, 2Sd), 0.25lbs ft^3 of stuffing, driver mounted at 0.25 the line length (so 30" from the base) a small port radius of 1.0" (so a 2.0" diameter) and short length (say 0.75" if you're using 0.75" build material, like most of us), then gradually increase the port-length by increments of, say 0.5" each time whilst holding all the other variables constant and watch what happens to the response. Make some written notes if you like (I do!) if you see something interesting, a pattern emerging etc. Don't try going for a port longer than 6" though. Ignore ripple over 100Hz for now: we can increase the level of stuffing somewhat to sort that out if required (it probably won't be though).

Done all that? Good. Now try increasing the vent radius to 1.5" (so 3" diameter), reduce the length to the shortest possible, and repeat the above proceedure. When you've done that, increase the radius to 2.0" (4.0" diameter), reduce the length to the shortest possible, and repeat once again. Don't go wider than that though: not worth it.

Spotted your favourite combination so far? Good. Note it down. Can your driver go lower? If so, increase the area of So and Sm to 3Sd with your favourite vent dimensions, and adjust the length and width to get the same response you favoured before (it should be a value near those you liked most for the smaller cabinet).

Still want more bass (providing the driver can go lower -check it's Fs to be sure -don't try going more than a few hz lower than that, if at all). Then increase So and Sm to 4Sd and again, adjust the vent dimensions to get your favoured response. As before, they should be fairly near those you found best before.

Right, once you've done all of the above, you can start refining. So & Sm=3Sd too small, but So&Sm=4Sd too much, for example?Go for a mid-value of, say, So&Sm=3.5Sd and again follow our established proceedure. Continue in that manner until you get the cut-off where you wish it to be.

Now, is there a bit of a problem at the 3rd harmonic? (generally in 90% of MLTLs, it'll be between 100Hz-200Hz, give or take) Try moving the driver up and down, whilst keeping the rest constant, which should help to supress it, if there is an issue there. if there isn't, all well and good. Rest of the response looking a bit ragged above 100Hz? Increase the level of stuffing to flatten it out, until you reach 0.5lbs. Then, if you feel like it, try tweaking the length a bit, up and down, and watch what happens to the response. Increments of 2" are usually good. You might need to adjust the port dimensions slightly, and also the driver position to compensate for any changes in the response, but by now, you should have a good feel for which way you need to go with these variables to get the response you desire.

This takes a long time, but with practice, you should be able to circumvent much of this in future as you'll know what you're doing then: as I said at the start of this post, this is just about observation and learning what the variables do and how they affect each other.

Hope it helps
Scott
 
kelticwizard said:
I have a question about the ML TWQT worksheet.

Epsilon-driver position ratio. They don't have that on the Ported Box Worksheet or the Sections worksheet.

My understanding is that the driver position ratio is:

Epsilon (e) = length of driver center to top of box / total length of Line.

Is that the case?


Quite correct. It's very useful as the position of the driver along the line dictates the degree to which the line-resonances are excited; vital in MLTLs and ML TQWTs. So, in, say, a 40" line, a driver at 0.25 line length will be 10" from the internal top of the line, or 30" from the internal bottom of the line.

Scott
 
Just a clarification to Scottmoose excellent proceedure write-up above.

The two graphs to watch are the Far Field Pressure Level response, and the Woofer and Terminus Level Response. This last one is the most useful in trying to see the interaction between port and driver.

Having set up one set of values to give you a set of graphs, look at the curves and decide which value to change next, eg port length.

Go to the start and enter the new length but do not press enter. Scroll down to the Woofer and Terminus curve. The curve will not have changed from your original entry. Watch the curve.

Now press enter and the curve will jump to the new port lenth value. This way you can more easily see the difference the new value makes.
 
Here's something new. Bored of rectilinear boxes? Yeah, me too.

How about a Thor ML TQWT? That's do-able. We could either do it the easy way, accept a very deep enclosure, and mount the drivers in the classic 'Voigt Pipe' layout on a 9" wide baffle, and accept a very deep enclosure.

However, we could also mount the drivers like those seen in Martin King's Project 2 ML TQWT, on the trapezoidal baffle, which would be my choice. Now, I'm conjecturing a little here, because I don't have an updated version of the ML TQWT sheet that can model the in-room baffle-step issues of a baffle of this shape, but my resoning at the moment is that if we work the dimensions so that the drivers 'see' a baffle approximately 9" wide, we should get away with it with any luck, and end up with a very handsome cabinet and novel design. Approximate response is shown below. I aimed for the slowly decreasing response seen to prevent the bottom end getting too heavy due to room gain.

It doesn't actually have to be much larger than the existing version for the FE164/7E either -couple of additional inches. All dimensions are internal.

60" tall
16.25" wide at the base
3.25" wide at the top
12" deep (so=1Sd, Sm=5Sd)
3"x3" port 3" up from Sm.
Tweeter centred at 0.5 line length.
0.4lbs ft^3 of stuffing.

Bit of fun anyway!

Best
Scott
 

Attachments

  • thor ml tqwt.gif
    thor ml tqwt.gif
    14.2 KB · Views: 1,404
Scott,

Some speculation on my part: (Mongo try to untangle words around concept) :scratch:

I think you are describing the baffle width around the drivers "averages" 9 inches but is wider below one mid-woof and narrower above the other mid-woof.

If this is the case, would the baffle step influence occur over a greater frenquency range?

Would the baffle step influence be less, if the above is true?

woof woof :D
 
Yep -the baffle above the top mid-bass will be narrower and the area below the lower unit will be wider. Tweeter centred in the cabinet with 4.5" on either side of its centre point (or thereabouts!)

As regards the baffle-step, I dunno. We'll have o wait for the new ML TQWT sheet to find out. But I think that any changes will be fairly mild, probably not such as to require re-jigging the BSC part of the crossover. I think you've hit upon it Ed -what changes there are will probably be spread over a slightly broader area. I doubt they'd be any more severe, and they might be a little less. I might see if I can refine this one: I like the tall ML TQWT enclosure.

Cheers
Scott
 
In a normal sized room orginal thor has great low frequency response alltough it plays to strong in the upper midrange that can make listeners believe it sounds weak in the lower bass ,
I tamed the thor using the orginal filter and then adding an resonance filter set to 3800 hz . adapting the db with an resistor of 12 ohms , The thors sounds like the perfect balanced speaker now , if any one wants to know the details feel free to add me to msn.
I believe the ears should be the judge for frequency response not any engineering frequency plot ,
 
How big is your room? Have you experienced voices and electric guitars to be to strong?
After my XO tweak the speaker vent from overanalytic edgy and painfull to organic easy flowing high end speaker a real bargain ,
Sounds a little like Audiovector M3 signature but with more low end and punch .
It can be good to use engineering equipment to know what end to start if the speaker obviusly sounds bad , but the final result is up to taste and ears .
The orginal Thor is gigantic! Making a fat thor is just not possible for most peeps .
 
Have been reading though the "Clarity" thread on the Thor, as well as this one, but I got lost in the "Short Thor" version......

1. Was there a simulation on the "Short" version....?

2. Is there some theory somewhere that can explain the "Short",
- or is so simple as the heightt of the cab' being the pipe length?
What about damping of this "line".. ?

3. Can one use the MJK sheets to determine the size of the loading port for a higher/slimmer version...??

The reason I'm asking, is that I was thinking of something like a MLTL version of a Stradivari cabinet......for the looks mainly, but also for the benefits of mid range dispersion....(a Strad "look-alike" would have very high WAF in our current living room - not something to be easily ignored..)
If this holds water at all, even a OB mid /tweeter combo should be possible ( if so wanted) ....
 
Yes to all. I designed Short Thor. There's nothing spectacular about it: it's simply a straight, mass-loaded quarter wave resonator. The straight bit simply means the line is not tapered, the mass-loading is the port. Sounds like a reflex design doesn't it? It's not. They work in a different way, and sound different too. The air-mass in the restricted area of the port provides an additional load on the quarter-wave resonances generated by the cabinet. This does two things. First; it reduces ripple over 100Hz. Secondly, it allows you to go lower for a given cabinet length. If I had, say, an open-ended pipe that resonated at 60Hz, by restricting the terminus area (mass-loading it in other words) I could reduce this resonance to circa 35Hz or so.

It's stuffed like other QWRs to damp out those higher harmonics that cannot be supressed by the mass-loading and correct positioning of the drivers (MTM makes it a pig in that respect: it's nonsense to say that the MTM configuration reduces ripple, it frequently makes it worse), but rather than a uniform stuffing, it's simply stuffed 0.5lbs ft^3 from the top of the enclosure to 15" above the internal base. That allows the stuffing to damp the unwanted higher harmonic modes whilst preserving the cabinet first mode, or Fundamental if you prefer, because the point of maximum velocity of this occurs at the terminus. Hense, we keep the stuffing away from the end.

Martin's sheets will allow you to calculate a taller, slimmer(? do you mean broader and shallower?) version, no problem. However, remember that when I designed Short Thor, I and Dave, who drew the plan of it, deliberately wanted to preserve the 9" width so no crossover modifications would be necessary to compensate for the changing point of baffle-step. I love the idea of a wide-baffle design though. I've experiemented with them in the past, and I thoroughly like the results. If you can re-jig the crossover it would be well-worth pursuing.

Linol -we clearly have different ideas about 'gigantic' -it's just a moderately large floor-stander by most standards. That said, as a bit of a single-driver fiend, I'm clearly more used to hefty enclosures than yourself. 6ft tall is about my limit, though I have built 8ft enclosures in the past to good effect. To answer your question, the room I heard Thors in was 25ft long by 16ft wide and 11ft tall. Wooden flooring, with a couple of (small) rugs, sealed brick walls, a few small paintings, two light-weight sofas and a chair. Ceiling was flat plaster. Couple of moderate book-cases, well away from corners. Generally quite a reverbarent room (I'm used to testing speakers). I didn't find a problem with the mid-bass, but low bass came there none. Effectively, there was zilch under 50Hz. You could detect it (just) down to 40Hz or so, but it wasn't usable. I checked drivers and crossover: no problems that I could find with either. Martin's simulation was spot on, allowing for a uniform 4db or so lift through room gain.

Regards
Scott
 
In my room there was a strong peak around 4khz making voices sound thin and painfull (also electric guitars) .
I had the Canton RCL previously they are the low end masters and for small rooms like mine this low end is only destroing the soundstage , My thors goes almost as deep as the Cantons but with lower db , but its all there bottom to top. Cant wait until i pass the 100 hour limit.
 
"the room I heard Thors in was 25ft long by 16ft wide and 11ft tall...
I didn't find a problem with the mid-bass, but low bass came there none. Effectively, there was zilch under 50Hz. You could detect it (just) down to 40Hz or so, but it wasn't usable"

Interesting - Scott, were the Thors a modified version, or one of the decendants??? I'm one of the 'thousand' DIY's that have followed the Clarity thread intently. I'll build when I believe the 'best of' version has been posted.

Best regards - Rob
 
when I first heard the Thors in the audio shop there was superb image really deep image an holographic soundstage overall but none bass below 50 hz or so maybe the room was to big or the hundreds of other speakers in the same room swallowed the low end by destroing the phase
 
They were the originals, no changes from standard specification, and believe me, I checked given that I've given many, many hours over many, many weeks to examining and re-engineering / designing Thor.

I've no arguments about their imaging, which was very good, or the depth of the soundstage, all of which was fine. I've no real comments to add about their behaviour above 1KHz either (bit hot, but I've known much worse): I'm interested in the enclosure, which is operative up to circa that point. Above that, its only function is to provide as solid a mounting point as possible for the drivers.

A large room should nominally improve low bass, not decrease it. OK, I'll qualify that: it should make the speaker's life easier. The smaller the room you have, the more power the speaker needs to put out at low frequencies when you reach, and try to go below, the room's natural resonant frequencies. A room about 15ft long, for example, will have a mode of 36Hz, give or take, if memory serves. OK, so that's the simple version, because in practice, the nature of the furnishings, materials etc will all contribute a significant part which is why I stressed that it was quite a 'live' (i.e. reverberent) room that I heard them in. Phase problems won't case the bass to slump to any significant extent -its excessive damping that'll do that. Although while we're on the subject, I think the drivers in Thor are a bit too far apart to be ideal.

'Best' version thus far -well, I reckon the Small Thor MLTL is my personal favourite. It doesn't reach quite as low as Short Thor, but the enclosure size is much more managable -basically similar to the standard enclosure in external dimensions. And it maintains full output down to the mid 30Hz regions, unlike the standard enclosure. The ML Fat Thor is the one to go for if you want to terrorise the local subwoofers, but it's not as smooth. Short Thor MLTL is the best compromise for the low-bass fiends. 24Hz easy. But, like Fat Thor, a huge enclosure. As they're only MDF, it shouldn't be too difficult, or expensive to try a couple to see what you think.

Regards
Scott
 
in large rooms low frequencys will be more audable since the wave length is longer and cant be audable in small rooms but for example place the thors in a big arena and there will not be so much low end any way .
Bigger speakers for bigger rooms thats a fact anyway.
The output in small rooms will be greater for any speaker alltough some other problems will occour in an small room like colorations etc...different rooms have different resonant frequencys ,
Its not easy to get perfect sound, as my old teacher said you cant just buy your way in to great sound it needs great knowledge, testing and patience .
 
I have read through pretty much all of the second half of the Clarity thread, and through all of this one. My interest lies in building a Thor-style (one of the newer ones, of course) enclosure to house a pair of Dayton RS180 woofers and an RS28A tweeter - the Modula design (Natalie P version) pioneered by Jon Marsh and the htguide.com bunch.

Anyways, of the Short Thor, Fat Thor and Small Thor, I only have specific cabinet dimensions and drawings for the Fat Thor, out of the Clarity thread. Can anybody tell me where I could find drawings for the other two?

Also, anybody care to elaborate on whether or not I'm heading in the right direction with this particular driver/enclosure combination?

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