Hi,
after a rewarding build of the Woden Baby Labs for the CHN-50, especially for the price, I am evaluating my options for a box with more bottom output. I am not leaning against Frugal Horn Lite because of the footprint and more difficult room integration.
So my options are Woden Giraffe and and a mass loaded quarter wave design from the MA website. Both designs are quite similar and the Giraffe is targeted for Alpair5.2. Which box gives more bottom end? Are there other differences I can expect in for example max spl or bass authority?
My biggest full range at the moment is a wide CGR Planet 10 design for CHP-70. In comparison, how would a CHN-50 in Giraffe perform against this? I guess max spl is lower, but what about low end?
Maybe I have overseen options.
Greets
https://www.kjfaudio.com/wp-content/uploads/2020/09/CHN50-Compact-Mass-Loaded-Quarter-Wave.png
http://wodendesign.com/downloads/Giraffe-A52-MLTL-0v99-081015.pdf
after a rewarding build of the Woden Baby Labs for the CHN-50, especially for the price, I am evaluating my options for a box with more bottom output. I am not leaning against Frugal Horn Lite because of the footprint and more difficult room integration.
So my options are Woden Giraffe and and a mass loaded quarter wave design from the MA website. Both designs are quite similar and the Giraffe is targeted for Alpair5.2. Which box gives more bottom end? Are there other differences I can expect in for example max spl or bass authority?
My biggest full range at the moment is a wide CGR Planet 10 design for CHP-70. In comparison, how would a CHN-50 in Giraffe perform against this? I guess max spl is lower, but what about low end?
Maybe I have overseen options.
Greets
https://www.kjfaudio.com/wp-content/uploads/2020/09/CHN50-Compact-Mass-Loaded-Quarter-Wave.png
http://wodendesign.com/downloads/Giraffe-A52-MLTL-0v99-081015.pdf
Greets!
No clue without reverse engineering them, so Scott/others involved with their designs needs to answer, but for reference when folks want to eek more LF performance out of a driver it's good to know where the 'wall of science' snuffs out any further 'pipe dreams' in that the basic physics of the situation is the driver's measured T/S specs sets the lowest practical tuning (Fb = 0.42*Fs*Qes'^-0.96) and at max acoustic eff. requires a hyperbolic flare (long/huge) BLH, which means it will only produce a usable BW from its upper mass corner (Fhm = 2*Fs/Qes') to its (Fb) and because it's tuned so far below (Fs) will only have fractional peak power output, though its high acoustic efficiency ensures more peak SPL than most folks will want. 😉
(Qes'): (Qes) + any added series resistance (Rs)
No clue without reverse engineering them, so Scott/others involved with their designs needs to answer, but for reference when folks want to eek more LF performance out of a driver it's good to know where the 'wall of science' snuffs out any further 'pipe dreams' in that the basic physics of the situation is the driver's measured T/S specs sets the lowest practical tuning (Fb = 0.42*Fs*Qes'^-0.96) and at max acoustic eff. requires a hyperbolic flare (long/huge) BLH, which means it will only produce a usable BW from its upper mass corner (Fhm = 2*Fs/Qes') to its (Fb) and because it's tuned so far below (Fs) will only have fractional peak power output, though its high acoustic efficiency ensures more peak SPL than most folks will want. 😉
(Qes'): (Qes) + any added series resistance (Rs)
Giraffe works well.
The specific CHN-50 ML-TL will be very similar but a bit less deep and shorter (the base drawing Scott used would be the Girrafe). @Scottmoose
dave
The specific CHN-50 ML-TL will be very similar but a bit less deep and shorter (the base drawing Scott used would be the Girrafe). @Scottmoose
dave
Will have a gander in the morning -currently on the laptop & about to sleep the sleep of the unjust. 😉
Oh the specific ml-tl is also from @Scottmoose ? I havent know.
I would like to build a design of his.
Much of a muchness; Giraffe gets a touch lower, with a touch more gain, but we're talking here about when it's used with the driver it was designed for. The price is slightly lower power-handling, these things being relative for 3in nominal drivers anyway. IIRC I did Giraffe at a request for an individual set of conditions / requirements. As a rule, I'd recommend using cabinets that are designed for the intended driver though, so if it's CHN-50 & you want a compact floorstander, then go with one designed for that unit -in the case of the two mentioned, it should have a touch (a touch) higher power-handling.
Okay thanks for the detailed explanation. Its interesting that these Ml-tl´s are forgiving similar for similar drivers.
I guess room, room placement and damping to taste in this conditions would make differences neglectable especially what one can expect from small drivers.
I guess room, room placement and damping to taste in this conditions would make differences neglectable especially what one can expect from small drivers.
I said the opposite actually: when possible, use enclosures that are designed for a given driver rather than trying to put something else in them. Within limits there can be some flexibility, especially when Vas is swamped and there isn't too much difference between the driver's LF load requirements, but as a general guideline it holds.
I have recently completed the mass-loaded quarter wave horn you mentioned. It has a really nice sound balance and good bass performance for such a small driver. I am using my pair in my home office and it has plenty of dynamics for that space.
Thanks for your experience. Have you build the box like plans in 12mm? Did you brace the box or the driver?
I used 12mm birch plywood, and no additional bracing. I have them close to a slanted roof/ceiling on a bookcase, the sound balance is on the warm side like that. On my desk, they were "lighter." There is a lot of detail in the sound, voices are very clean. I don't think bracing would improve it a lot. I would have been happy with just "Bluetooth radio sound" but these are much more. They fail at reproducing a Mahler symphony at realistic levels, but hey. It's about as large as the screws that hold my 15" Fane fullrange drivers down. 😁
Last edited:
I have completed the ML-TL design for the CHN-50. They have played a couple of 10h and I think they are fully settled in.
In comparison to my Bloodhound (Baby Labs) bookshelf TL with CHN-50 they do play with more fundamental, they go a bit more deeper. But the Bloodhound with less extension has the better kick, especially when listening to electronic music. Its more prominent.
In comparison to my Bloodhound (Baby Labs) bookshelf TL with CHN-50 they do play with more fundamental, they go a bit more deeper. But the Bloodhound with less extension has the better kick, especially when listening to electronic music. Its more prominent.
Do tell, why would anyone bother with this driver?, it's NOT high fidelity, & frankly can't be massaged to vaguely approach high fidelity. Peruse the factory SPL to frequency graphs, at 140 hertz it is squeaking out only 87 decibels worth of output. As the frequency drops below that 140, it crashes to an inaudible 71 decibels worth at 60 hertz...kinda pointless really as it has no bass whatsoever, its response bobbing wildly from plus 10 Db to minus seven Db.
Further, as I had mentioned before, it is misrepresented as a "Three inch" driver...the effective cone diameter is more like 2.75 inch, the big flanges taking up more surface area than the cone itself.
Myself having a similar 88 Db worth five-inch DIY fullrange pair, the limited output power can be most problematic...limiting its application to a desktop home.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Rick...
Further, as I had mentioned before, it is misrepresented as a "Three inch" driver...the effective cone diameter is more like 2.75 inch, the big flanges taking up more surface area than the cone itself.
Myself having a similar 88 Db worth five-inch DIY fullrange pair, the limited output power can be most problematic...limiting its application to a desktop home.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Rick...
I'm sorry, am I missing some emojis? Or are you serious? Obviously, the response is measured under standardised conditions that do not include an enclosure to fill in bass frequencies, and does not include room influences. It doesn't make a factory frequency response plot useless, but it's not suitable for the kind of evaluation you are attempting here.
OK, I'll bite. I've not got much time at the moment, but FWIW...
All competent factory FR data is taken under anechoic conditions, and without any form of usable low-frequency loading. Measurements are generally taken of drive units mounted on either some kind of infinite baffle arrangement (such as poking through what is in effect a false wall), huge sealed test boxes, smaller sealed test boxes, or an IEC compliant test open-baffle. Each of these have different characteristics that affect the frequency response, particularly at the bottom end of the range, and sometimes up to a little over 1KHz, depending on size, driver and mic. positions etc.
Measurements for most of the midrange frequencies and > are usually taken at 1m or 0.5m, in the latter case with the levels normalised to the 1m value, with the manufacturer's choice of 1w or 2.83v as the nominal. Different companies then handle the low frequency range (and if relevant, the lowermost part of the midrange) differently depending on their preferred methodology. To give a few examples:
2.5in = 22cm^2 - 24cm^2
3in = 28cm^2 - 32cm^2
3.5in = 36cm^2 - 38cm^2
4in = 50cm^2 - 60cm^2
5in / 5.5in = 75cm^2 - 90cm^2
6in = 110cm^2 - 120cm^2
6.5in / 7in = 130cm^2 - 150cm^2
8in = 200cm^2 - 240cm^2
10in = 325cm^2 - 360cm^2
12in = 460cm^2 - 520cm^2
15in = 800cm^2 - 900cm^2
Perhaps they are in the market for a 3in size wideband drive unit for use in compact systems, desktops, bookshelves, or as a mid-tweeter in a 2-way? Those are the usual reasons anyway.
It looks like you don't understand how factory response plots are taken, and the various methods that are used in doing so. Here are some of the actual facts of what is involved.
All competent factory FR data is taken under anechoic conditions, and without any form of usable low-frequency loading. Measurements are generally taken of drive units mounted on either some kind of infinite baffle arrangement (such as poking through what is in effect a false wall), huge sealed test boxes, smaller sealed test boxes, or an IEC compliant test open-baffle. Each of these have different characteristics that affect the frequency response, particularly at the bottom end of the range, and sometimes up to a little over 1KHz, depending on size, driver and mic. positions etc.
Measurements for most of the midrange frequencies and > are usually taken at 1m or 0.5m, in the latter case with the levels normalised to the 1m value, with the manufacturer's choice of 1w or 2.83v as the nominal. Different companies then handle the low frequency range (and if relevant, the lowermost part of the midrange) differently depending on their preferred methodology. To give a few examples:
- Scan Speak use a very large, near-infinite baffle with the rear of the driver sealed off into a 320 litre volume, and measure at 1m. They have a chamber large enough to do this successfully, which is a luxury most do not have
- Seas's current approach is to use relatively small, narrow-baffle test boxes. They measure at 0.5m, normalise to 1m, and their measurements are in fact corrupted by both baffle-step, and edge diffraction, typically to just over 1KHz. To that FR graph, they then add the mathematical T/S derived low frequency curve so people can see what the nominal is.
- Dayton TTBOMK measure on large baffles with the rear of the driver sealed off. They then take a nearfield measurement with the mic. about 1/4in from the cone, level match it to the 1m value, and splice it in, typically at 450Hz.
- Markaudio use an IEC compliant open test baffle and measure at 1m (which is what Seas used to do), hence the consistent cancellation around 450Hz - 500Hz, and the characteristic dipole lift below that, until it finally rolls off. This is the method Seas used to use, although they also drew in the mathematical IB curve derived from the T/S parameters. I'm currently working with MA to either add that, or the level-matched nearfield measurement to future data-sheets, plus some other enhancements -extra data etc.
IIRC you've been told this before, but repeating for completeness, '3in' is a approximate size bracket description used for convenience across the majority of the industry. Typical Sd values for the more common size brackets run approximately like this:
2.5in = 22cm^2 - 24cm^2
3in = 28cm^2 - 32cm^2
3.5in = 36cm^2 - 38cm^2
4in = 50cm^2 - 60cm^2
5in / 5.5in = 75cm^2 - 90cm^2
6in = 110cm^2 - 120cm^2
6.5in / 7in = 130cm^2 - 150cm^2
8in = 200cm^2 - 240cm^2
10in = 325cm^2 - 360cm^2
12in = 460cm^2 - 520cm^2
15in = 800cm^2 - 900cm^2
Well, you've partially answered your own first question there 😉 but to that, you also need to add context, usage, available linear travel, effective output once losses are factored in etc.
Last edited:
Tell you what, take this driver, fabricate any enclosure you desire, apply any amplification & processing you want...we'll put the pair in one corner of a studio, and in the other corner we'll put a baby grand piano.
Take whatever recorded Piano music you wish...walk into that studio blindfolded...Do you really think one couldn't tell which was which?..."exactness in details" Merriam Webster.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------Rick...
Last edited:
Of course not, because nobody in their right mind would think that is what it's designed for. In the same way that no reasonable person would think a 5in direct radiator is designed for that. Or a 6.5in. Or any size of direct radiator with a reasonably wide BW. As you know, there is not one, of any available size, that is capable of accurately reproducing the response & dynamic range of a baby grand piano at 'live' SPLs & most typical listening distances, with a partial exception going only to Olson's LC-1A, and even that comes with a bunch of caveats as it's a hybrid triax. So your argument is a bit like criticising a baboon for not being an ocean-liner, since you seem to be applying criteria the majority of drive units were never meant to achieve. 'Exactness in details', including context, strikes again. 😉
Welcome to reality: 99.999% of 'hi-fi' systems are incapable of accurately reproducing the full frequency and dynamic range of a piano at live SPLs. Which isn't really a startling revelation, because most of their owners (certainly anybody who has even been near a piano) also know that. To do it in a reasonably sized space needs at least a couple of 15in woofers & a compression mid-tweet, and preferably more. The (latest) point you are carefully avoiding is that many don't care, because they don't want or need that abilty, and if they did, most couldn't afford what is necessary to achieve it either. I can't myself, and even if I could, the neighbours would be complaining 24/7.
Welcome to reality: 99.999% of 'hi-fi' systems are incapable of accurately reproducing the full frequency and dynamic range of a piano at live SPLs. Which isn't really a startling revelation, because most of their owners (certainly anybody who has even been near a piano) also know that. To do it in a reasonably sized space needs at least a couple of 15in woofers & a compression mid-tweet, and preferably more. The (latest) point you are carefully avoiding is that many don't care, because they don't want or need that abilty, and if they did, most couldn't afford what is necessary to achieve it either. I can't myself, and even if I could, the neighbours would be complaining 24/7.
Last edited: