Steve,
The Metronomes are also quite interesting.... and not really a TQWT (in a strict sense). I can't remember whether we discussed it here or on the full-range forum. Because it tapers in both directions the taper is not linear, but quadratic. It does make for a very elegant looking speaker.... have you posted drawings anywhere?
dave
PS: always nice to see the "Made on a Mac"
The Metronomes are also quite interesting.... and not really a TQWT (in a strict sense). I can't remember whether we discussed it here or on the full-range forum. Because it tapers in both directions the taper is not linear, but quadratic. It does make for a very elegant looking speaker.... have you posted drawings anywhere?
dave
PS: always nice to see the "Made on a Mac"
Dave
Hmm
There's a slight curve on the taper
Does this mean it has an element of horn behaviour involved with it ?
Steve.
Hmm
There's a slight curve on the taper
Does this mean it has an element of horn behaviour involved with it ?
Steve.
Steve Cresswell said:
A Voigt can already be thot of as a form of conical horn. Here the taper is quadratic, as opposed to linear.
Normally a classic receipe Voigt has little chance of being close to optimal, but you have added a twist -- this has been discussed, but yours is the 1st i've seen actually built.
http://www.diyaudio.com/forums/showthread.php?postid=569253#post569253
I've sent the data off to Scottmoose to see if he would be so kind as to model it for us.
dave
This thread was split off from http://www.diyaudio.com/forums/showthread.php?postid=989597#post989597
It is about Steve's Metronome speaker
http://web.mac.com/scress1958/iWeb/Steve's Tube Trials/Metronome Speaker.html
(BTW Steve using spaces in URL name's is not recommended)
dave
It is about Steve's Metronome speaker
http://web.mac.com/scress1958/iWeb/Steve's Tube Trials/Metronome Speaker.html
(BTW Steve using spaces in URL name's is not recommended)
An externally hosted image should be here but it was not working when we last tested it.
dave
May i offer a few suggestions? I did a great deal of study on this action and even built some based on near the same principal several years ago.
ron
ron
I have some scrap wood around but no design skills. It there an easy way to change this design to suit 8 inch drivers?
This is a whole new avenue to explore (particularily if the performance differs drom a straight taper box.
I'm sure you could make one large enuff for an 8". Questions like is a 1/3 lenght an optimum offset? What happens if you mass load it? What do changes in taper rate get (or lose) you.
dave
I'm sure you could make one large enuff for an 8". Questions like is a 1/3 lenght an optimum offset? What happens if you mass load it? What do changes in taper rate get (or lose) you.
dave
Hi Dave -interesting design. Did you get my email?
Hugz -First port of call you need is Martin King's TL alignment tables. Necessary line length varies depending on taper.
Generic example lifted from Martin's tables: assume a 40Hz cut off. A straight line requires an 84in line. A TQWT with a 1:3 taper ratio requires 102.9in length. And a reverse tapered line with a 3:1 taper needs a 66.8in length. Providing you keep this basic principle in mind when you design, you're fine. If not, as Martin points out, you're in trouble. His tables are free to download, and provide ~optimum driver offsets for each rough geometry too. The quadratic flare will affect the spread of resonance, but it won't make much difference to the final cut-off IMO, though I speak under correction. Mass-loading should work in the same fashion as it does for a staight taper.
Hugz -First port of call you need is Martin King's TL alignment tables. Necessary line length varies depending on taper.
Generic example lifted from Martin's tables: assume a 40Hz cut off. A straight line requires an 84in line. A TQWT with a 1:3 taper ratio requires 102.9in length. And a reverse tapered line with a 3:1 taper needs a 66.8in length. Providing you keep this basic principle in mind when you design, you're fine. If not, as Martin points out, you're in trouble. His tables are free to download, and provide ~optimum driver offsets for each rough geometry too. The quadratic flare will affect the spread of resonance, but it won't make much difference to the final cut-off IMO, though I speak under correction. Mass-loading should work in the same fashion as it does for a staight taper.
Scottmoose said:
No, but i have been on the lookout since i sent you the metronome package.
dave
This is interesting! I will work up a few sims for what i did in the past and see if i can incorporate the principal into the H108.
ron
ron
planet10 said:
Blast! I sent it this morning, but it must've got itself lost in the system. I'll re-run the sims and get them over to you. One was for the design as it stands, the other a mass loaded version.
Hey Scott
It would be interesting to have a look at those sims to see if they do really cut off as sharp as I hope they do.
They seem to integrate well enough with my sub but sometimes I get a slight honk on records with a strong upper bass content such as some Motown and Atlantic/Stax stuff from the 60s.
It is not enough to be annoying and no speaker is perfect but I'd like to get some insight into a way of maybe getting rid of it.
Steve
It would be interesting to have a look at those sims to see if they do really cut off as sharp as I hope they do.
They seem to integrate well enough with my sub but sometimes I get a slight honk on records with a strong upper bass content such as some Motown and Atlantic/Stax stuff from the 60s.
It is not enough to be annoying and no speaker is perfect but I'd like to get some insight into a way of maybe getting rid of it.
Steve
PM dispatched. Yes, they do have the sharp cut-off according to MathCad. I think we can explain the honk too.
Best
Scott
Best
Scott
Steve,
I'm away from my home 'puter at the moment. I have a chart that shows the as built, modified to an ML-QQWT, and one lengthened to reflect the length needed to actually have the box resonate at a 1/4 wl of the Fs.
I've hesitated to post it so far because the straight taper equivalents are perfect duplicates, something i didn't expect, and therefore deserve a triple check.
The as built QQWT shows a roll-off at ~100 Hz with strong ripple -- a broad peak at the bottom, a deep null at about 400 Hz, and then another broad peak (plots only go up to 1k). The bottom peak could offset baffle step, the null is narrow enuff that it might be unobjectionable, but the upper peak indicates a lot of mid-range energy coming from the bottom which might colour the mids a bit -- this is into the region of floor bounce and other room effects so may well be swamped by other things.
The ML-QQWT exhibits much, much smoother response, with almost an extra octave on the bottom, the lengthened QQWT looks like the shorter one, but with everything shifted lower in frequency.
Scott also mucked about a bit with a bipole FE126 in an ML-QQWT and it looks quite good. Certainly an alternative to having to build a horn.
dave
I'm away from my home 'puter at the moment. I have a chart that shows the as built, modified to an ML-QQWT, and one lengthened to reflect the length needed to actually have the box resonate at a 1/4 wl of the Fs.
I've hesitated to post it so far because the straight taper equivalents are perfect duplicates, something i didn't expect, and therefore deserve a triple check.
The as built QQWT shows a roll-off at ~100 Hz with strong ripple -- a broad peak at the bottom, a deep null at about 400 Hz, and then another broad peak (plots only go up to 1k). The bottom peak could offset baffle step, the null is narrow enuff that it might be unobjectionable, but the upper peak indicates a lot of mid-range energy coming from the bottom which might colour the mids a bit -- this is into the region of floor bounce and other room effects so may well be swamped by other things.
The ML-QQWT exhibits much, much smoother response, with almost an extra octave on the bottom, the lengthened QQWT looks like the shorter one, but with everything shifted lower in frequency.
Scott also mucked about a bit with a bipole FE126 in an ML-QQWT and it looks quite good. Certainly an alternative to having to build a horn.
dave