TL alignement question

[IG81]

In a TL alignement with SO = SL, if the (constant) cross-sectionnal area is diminished, what would be the effect?

I would think it'd increase the acoustic impedance, further load the driver and add some damping to smooth the low frequency peaks, at the expense of LF output. Does that make sense or is it another story completely?

IG

 

[Scottmoose]

I'm not sure I entirely follow what you're describing (not at my most awake this evening) -are we talking a reverse-taper line here, i.e. one that narrows toward the terminus? I can't quite see how the area of a constant cross-section line can diminish otherwise; by defininition, it's constant, with a throat & terminus sharing the same CSA.

 

[IG81]

I'm not sure I entirely follow what you're describing (not at my most awake this evening) -are we talking a reverse-taper line here, i.e. one that narrows toward the terminus? I can't quite see how the area of a constant cross-section line can diminish otherwise; by defininition, it's constant, with a throat & terminus sharing the same CSA.

Assuming we are indeed talking about a reverse-taper line, then if you hold Vp constant, Fp is reduced for a given length (you could say they're inherently mass-loaded), gain also reduces, & you will typically see an additional reduction in the amplitude of harmonic resonances. Conversely, a positive-taper line (expanding toward the terminus) will exhibit a higher Fp & gain for a given length than a constant CSA or reverse-taper line, as it's essentially a conical horn & therefore it's resonant behaviour is that of, or approaches that of, a cone.

I mean a non-tapered line, thus SL = SO. By "constant", I meant it "constant throughout line lenght", sorry if it was confusing. What happens if the cross-section is diminished, so both SL and SO would be made smaller by the same ammount, leaving all else unchanged.

IG

 

[Scottmoose]

Ah, right.

Well, like anything else, a TL needs a minimum volume (Vp in this case rather than Vb). The pipe always resonates at a specific frequency determined by it's length; that's inherent. However, as you reduce the CSA LF output will become progressively 'strangled' by the lack of volume. See the attached -undamped of course, for clarity. Once you start damping it, things just get worse.

BTW, I junked the last bit in my first post as it was of no use.

 

[IG81]

Ah, right.

Well, like anything else, a TL needs a minimum volume (Vp in this case rather than Vb). The pipe always resonates at a specific frequency determined by it's length; that's inherent. However, as you reduce the CSA LF output will become progressively 'strangled' by the lack of volume. See the attached -undamped of course, for clarity. Once you start damping it, things just get worse.

BTW, I junked the last bit in my first post as it was of no use.

Thanks, I think I understand better.

Can you tell me what was the magnitude of the CSA change from top to bottom graph in your example?

I would have been looking to diminish CSA in my alignement to 25% of what the MJK alignement prescribes. Might be a tad much.

IG

 

[Scottmoose]

Those were very much generic, so don't expect the same to apply in all situations. For information though, the lower graph shows a pipe 1/3 the volume of the top. Mr Sod's Law decrees that demonstrations never turn out as well as you'd like, and in this particular instance, the length / CSA / driver I happened to use probably weren't the best examples. Remember too, that things rapidly get worse when damping is added to the smaller CSA pipe.

Less obviously, what the machine doesn't show in addition to the above is the severe subjective 'strangling' that the undersized pipe causes once damped sufficiently to smooth the response out to an acceptable level, particularly WRT dynamic BW & transient attacks.

25% is a pretty hefty reduction; I'd be very cautious there. As well-known an engineer as Joe d'Appolito fell foul of that one when he designed the Thor TL for Seas. Bad idea.

 

[IG81]

Those were very much generic, so don't expect the same to apply in all situations. For information though, the lower graph shows a pipe 1/3 the volume of the top. Mr Sod's Law decrees that demonstrations never turn out as well as you'd like, and in this particular instance, the length / CSA / driver I happened to use probably weren't the best examples. Remember too, that things rapidly get worse when damping is added to the smaller CSA pipe.

Less obviously, what the machine doesn't show in addition to the above is the severe subjective 'strangling' that the undersized pipe causes once damped sufficiently to smooth the response out to an acceptable level, particularly WRT dynamic BW & transient attacks.

25% is a pretty hefty reduction; I'd be very cautious there. As well-known an engineer as Joe d'Appolito fell foul of that one when he designed the Thor TL for Seas. Bad idea.

Yeah, I thought it was kinda steep. I would have liked the outlook to be a bit brighter, as material availability and cost would blow this project's low price point aim through the roof were I to stay with 100% prescribed CSA. I'll look into other solutions.

Thanks!

IG

 

[Scottmoose]

OK, what driver are we talking about here? System? Room? Material played? Goals? They all have a bearing, but we can probably help out.

 

[IG81]

Hehe, OK, if you must know...

It's mostly a case of feeling I need to do something with stuff that I have in a box. Room? Nothing planned, could even change building a few times! Music? Probably extremely varied, or whatever would end-up sounding good. Main goal, to build something at least half-decent (for which I need at least a good feeling of the design from the start which is not the case so far).

I posted in this forum looking for TL info as it seems to be a specialty of the Fullrange folks, but my project is not fullrange, but two-way. I'll still talk about it briefly, keeping to a minimum the 2-way stuff.

I have a nice pair of Vifa C13WG-19-08 that were grabbed for peanuts and some HT26 horn tweeters. I wanted to mount them into a TL, but since I don't have a garage anymore and don't want to work wood too extensively, I thought I'd mount them into ABS pipes, ¼ inch walls.

My alignement prescribed a CSA close enough to that of an 8" pipe. Over here, that size is not available in normal stores, can only be ordered with big plumbing supply places and the cost is rather prohibitive. Even 6" diameter is out of my budget. 4" is what I was looking at, being the standard, easily available and not to expensive.

I want to throw together something quick and cheap ($) that has at least an OK potential to sound decent, but what we're looking at with 4" pipe instead of 8" (25% of CSA) is not very enticing, IMHO.

So that's where I'm at now!

IG

 

[GM]

What about concrete column tube formers, AKA Sonotube? Really, if looks aren't a big deal, then making them out of heavy corrugated cardboard

Anyway, with a high Fs, low Vas and medium Qts, significant BSC plus designing based on a Q = 2 is required to get any LF out of it with a considerable loss of efficiency as the trade-off.

GM

 

[IG81]

What about concrete column tube formers, AKA Sonotube? Really, if looks aren't a big deal, then making them out of heavy corrugated cardboard

Anyway, with a high Fs, low Vas and medium Qts, significant BSC plus designing based on a Q = 2 is required to get any LF out of it with a considerable loss of efficiency as the trade-off.

GM

I was hoping to use ABS 90deg elbows and sanitary tee-joints for driver mounting and industrial look, which is not something that sonotubes would allow me, but yeah, I was starting to consider that as a plan B now.

What do you mean by "design based on a Q = 2" ? I have not seen (or don't remember) a reference to this in the literature I read.

If you saw the Fs = 55Hz, Qts = 0.39 specs from the Tymphany archives, my drivers measured differently. Fs is actually 80Hz (!) and Qts = 0.44. I used proven T/S measurement methods that I think were pointed out by you some months back in a former project of mine. I am a bit stumped by the huge Fs deviation.

IG

 

[IG81]

I just happened upon the following link that pretty much looks like what I had in mind.

http://www.geocities.com/qgap/Pipe_Dreams.html

I have nice little 4" JVC fullrangers I also wanted to plug into pipes, but I beleive their high Qts of 0.84 is what also caused my alignement to require a big CSA.

I'll keep working on that....

IG

 

[GM]

What do you mean by "design based on a Q = 2" ? I have not seen (or don't remember) a reference to this in the literature I read.

Fs is actually 80Hz (!) and Qts = 0.44.

You won't AFAIK, it's 'old school' before T/S, so wasn't readily available to the general public in any formal design routine, but I was tired and just 'thinking out loud' when I posted, so ignore it and use MJK's or Rick Shultz's design routines based on total system impedance (Re + Rs).

Well, me/Scott/whoever's responses can at best only be as 'accurate', pertinent as the thread's available info..........

GM