Amen. That's the driving principle for me, behind increasing the number of multi-way sections.
Whatever a driver's frequency span is, excursion needs to increase 4X per octave decrease, to maintain level SPL.
I think that's one of the most overlooked speaker design issues.
For a compression driver (here in an extended waveguide), the excursion can look like this - for a constant voltage and for a constant SPL (dashed line):
- Some wouldn't use it below 1.5 kHz, I would go down to 600 Hz...
BTW, don't be afraid to measure the CDs without any protective filters, it's really not needed in most cases, it won't break them. The surround is so stiff that it won't let the diaphragm move a lot. It would need to be an insane level to damage the driver.
- Some wouldn't use it below 1.5 kHz, I would go down to 600 Hz...
BTW, don't be afraid to measure the CDs without any protective filters, it's really not needed in most cases, it won't break them. The surround is so stiff that it won't let the diaphragm move a lot. It would need to be an insane level to damage the driver.
Last edited:
Based on those excursion graphs, i'd be happy with 600 Hz too.
I end up just listening to what xover freq sounds best, because it's so easy to try any freq, with fully complementary linear xovers.
Another amen. Why folks think they need protection on CDs when measuring with low voltage sweeps like 2.83v (or for that matter, any HF/VHF driver) ,
has always made me go huh?
Heck, even if 10V kills your driver with no high-pass in place, junk the damn thing.
I'd say the crossover is a lot easier and also the reqirements for a woofer are less stringent regarding its quality as a midrange (which can be a pain to source, especially for sizes 8 - 10", if a physically smaller loudspeaker is a target). Simply the wider the bandwidth radiated by the waveguide, the better, it just relaxes many design aspects.
- This is the latest development, the smallish ATH-280EX (#13,808), STL is available via link on my website.
This one should be pretty universal. The gradually rising DI throught the 1k region makes it possible to be used with various woofer sizes, anywhere between 700 - 1500 Hz, depending on a CD used. Around 800 Hz it can be even matched to a "cardioid" midrage unit (DI ~ 6 dB).
- This is the latest development, the smallish ATH-280EX (#13,808), STL is available via link on my website.
This one should be pretty universal. The gradually rising DI throught the 1k region makes it possible to be used with various woofer sizes, anywhere between 700 - 1500 Hz, depending on a CD used. Around 800 Hz it can be even matched to a "cardioid" midrage unit (DI ~ 6 dB).
Last edited:
Surprisingly, if you ask what is the criteria of a woofer of high midrange quality, it turns out that no one can really tell. apart from break up resonance close to the upper passband limit.
The single most important criterion is a smooth frequency response, including the whole radiation pattern, through the intended crossover region and sufficiently above that. This alone can be pretty hard to find. Above 1 kHz virtually all of the drivers have some problems. If you can find one that doesn't, you're lucky. I've made enough loudspeakers of this kind that I know how tight it usually is, one virtually always wishes for a better woofer...
Or, you can set the crossover lower and forget about it. Almost anything will do up to 600 or 800 Hz. Then, we can look more closely on distortion, etc... There will be a lot more what to choose from.
Or, you can set the crossover lower and forget about it. Almost anything will do up to 600 or 800 Hz. Then, we can look more closely on distortion, etc... There will be a lot more what to choose from.
Last edited:
Two way monitor systems have been made for about 70 years. Usually with a 15" woofer and some horn/CD for the treble. Crossovers are often about 800Hz. A current example is the JBL M2 - driver spacing is about 17" or 1WL at 800Hz. I know the M2 is not exactly a 2-way. Perhaps a 15" woofer, with horn/CD spaced at 17", and 800Hz crossover, is close to ideal. Perhaps a 1" exit CD is on the small side for a 2-way with an 800Hz or lower crossover. A 15" woofer is on the large side for a crossover above 800Hz.
Good 15" woofers without large amounts of HF breakup are available. Different crossover frequencies and different spacing will change the pattern through the crossover region. The size, shape, and dimensions of the listening space will have an impact. Ideal must include some practical considerations.
Good 15" woofers without large amounts of HF breakup are available. Different crossover frequencies and different spacing will change the pattern through the crossover region. The size, shape, and dimensions of the listening space will have an impact. Ideal must include some practical considerations.
Last edited:
What if someone doesn't want to (or just can't) use as big woofer as a 15" and a correspondingly big waveguide? It gets complicated quickly, as it really doesn't scale very well - for some reason, the smaller woofers are typically not any better above ~1 kHz than an average 15".
(JBL M2 is exactly a 2-way, BTW)
(JBL M2 is exactly a 2-way, BTW)
Last edited:
I thought the current discussion is about using drivers and waveguides to accommodate a 600-800 Hz crossover between a woofer and horn/CD. I'll keep reading. I'm interested in your solutions.
I want a system that is infinitely small with an infinite bandwidth, with perfect frequency and phase response, and perfect dispersion for my room. I think I will Not get it.
The D2 CD tweeter in the JBL M2 is dual diaphragm, dual voice coil. Does each voice coil receive the same signal? Does each diaphragm have the same response?
I want a system that is infinitely small with an infinite bandwidth, with perfect frequency and phase response, and perfect dispersion for my room. I think I will Not get it.
The D2 CD tweeter in the JBL M2 is dual diaphragm, dual voice coil. Does each voice coil receive the same signal? Does each diaphragm have the same response?
Last edited:
Imagine yourself building a controlled directivity loudspeaker with a WG and a 10" woofer, say a pair of 10" in a tower, making a 2.5-way (which is already a "serious" design, just not that big.) Now, a 1" waveguide that allows to set the crossover around 800 Hz will be of great help in that case.
Yes, as far as I know.
Last edited:
I made a simple experiment, perhaps not as convincing as I hoped for , but neverthelsess -
This is 1" Selenium D2500Ti-Nd in the ST260 waveguide, connected a) directly and b) via a 70 mm conical duct starting at the driver's phase plug exit. All else equal, I kept the mic-driver distance constant during the experiment.
The direct ST260:
Via the extension:
Comparison of the on-axis responses:
My conclusions from this one:
, but neverthelsess -This is 1" Selenium D2500Ti-Nd in the ST260 waveguide, connected a) directly and b) via a 70 mm conical duct starting at the driver's phase plug exit. All else equal, I kept the mic-driver distance constant during the experiment.
The direct ST260:
Via the extension:
Comparison of the on-axis responses:
My conclusions from this one:
- The extension added 9 dB of SPL around 600 Hz.
- The extension adds FR disturbances caused by the throat reflection that need to be tackled in the filters (EQed)
- In this case, starting at the phase plug exit leads to a more coherent HF response - less beaming in the top octave.
BTW, this is a comparison of the D2500 in the above device (with the extension, top) and in the ATH-280EX, presented earlier (bottom):
(Absolute SPLs are not directly comparable. I'm only not sure this was the very same driver...probably not)
It seems it's always best to start at the phase plug exit. I'll also try to modify the ATH-280EX for this driver to be used this way, it should further clean it a bit.
(Absolute SPLs are not directly comparable. I'm only not sure this was the very same driver...probably not
)It seems it's always best to start at the phase plug exit. I'll also try to modify the ATH-280EX for this driver to be used this way, it should further clean it a bit.
Last edited:
Thanks for providing a hard criterion of yours.
Let's play with it based on a LR4, 24db/octave filter. LR4 is -6 dB at fc, so we can assume -40 dB is roughly reached at 1.5 octaves above fc (-6 dB -48 dB = -54 dB, +12 dB = -42 dB ~ -40 dB).
For a 800 Hz crossover any resonance / woofer breakup below/at 2400 Hz would be prohibitive.
I am using a crossover at 1250 Hz, which would require a response up to 3750 Hz free of resonances.
The woofer I am using has its breakup resonance at 2700 Hz:
https://phlaudio.com/fileadmin/user_upload/phl_audio/3411_SpecSheet.pdf
Okay, works within your premises of the 800 Hz crossover. How did you come up with -40 dB?
Experience is nice, but it must be formalized if it should help with modeling. This is why I was asking.