Horn Speakers - Is it me or.......

[babadono]

I'm confused...of course you could put "pads" in front of an active crossover. I do it all the time, its called a mixing board. This will not help if the fear is a failing active crossover sending hot signal to a driver through its amp. Kind of like worrying about the accelerator on your car getting stuck..could it happen...sure.

 

[Flaesh]

the accelerator on your car getting stuck

Or an object might suddenly roll of nowhere under the brake pedal, preventing you from pressing it.
A friend's kitten crawled under the brake pedal when an oncoming car made a dangerous left turn. Miraculously, there was no collision with that car.
I think this is too much of a discussion of irrational fear. I just put the horn with 4550 to my ear, slightly worried that a full-power signal would suddenly be transmitted. But I heard nothing! Then I remembered the system was off.

 

[Newman]

A simple resistor divider between pre and power amps will do it.

@dped90 The reason I suggested putting it before the power amps was its simplicity.

Putting a divider after the power amps has advantages, such as it reduces noise (hiss, hum etc) if your power amps are noisy, and also reduces the loudness of any on/off thumps that originate from the power amps. But you have to be a bit more careful than with the simple divider before the power amp: careful to use resistors with sufficient power rating and to choose values that preserve the overall output impedance seen by the power amp to something it expects and that doesn't change the frequency response of the system itself.

cheers

 

[egellings]

@dped90 The reason I suggested putting it before the power amps was its simplicity.

Putting a divider after the power amps has advantages, such as it reduces noise (hiss, hum etc) if your power amps are noisy, and also reduces the loudness of any on/off thumps that originate from the power amps. But you have to be a bit more careful than with the simple divider before the power amp: careful to use resistors with sufficient power rating and to choose values that preserve the overall output impedance seen by the power amp to something it expects and that doesn't change the frequency response of the system itself.

cheers

Resistor on the power amp output clobbers your damping factor. Not a good place to put it.

 

[dped90]

I would put the L-pad between amplifier and speaker drivers, mostly because that's what I'm familiar with. If someone with experience using L-pads further up the signal chain tells you otherwise, follow their advice over mine.

If you decide to go with an L-pad between amp and speaker drivers, you could do this: Buy an inexpensive L-pad and use that to dial in the correct loudness for your horn, then disconnect the L-pad and measure the series resistance and the parallel resistance, and then you would know what values you want for the (presumably) fairly expensive resistors for your permanent, fixed-value L-pads.

A wise approach.

I like the way Peter Noerbaeck mounted the horn upside-down to slightly decrease the center-to-center spacing between horn and upper woofer.

Yes, though it looks like I won't get them quite that close together, especially if Troy recommends putting the M2 or XT1464 in some kind of a box.

Imo for good imaging across an area, not just right smack in the sweet spot, you want the off-axis curves to be essentially less-loud-replicas-of the on-axis curve, as this will give you the option of using time/intensity trading to get good imaging across the listening area.

By less-loud replicas of the on-axis curve, you mean x & y plots like this?
https://www.audiosciencereview.com/...jbl-m2-reference-master-monitor-review.26051/ But what is time-intensity trading and how does it maintain imaging integrity over the entire or most of the listening space (in the critical distance triangle) and how would the M2 and XT1464 compare in that regard?

Four feet is not a very wide listening area, so imo you don't need a really wide constant-directivity pattern. The off-axis curves of the 18Sound XT1464 track the on-axis curves exceptionally well across an arc 60 degrees wide from 1 kHz to about 11 kHz, and it's still pretty good south of 1 kHz and north of 11 kHz.

A 60 degree arc, but 120 degree coverage angle?

No, listening width will be between 9 and 11 ft, as probably will distance. What I meant was that I'd prefer a waveguide that could offer very good vertical coverage at both of those distances. How and why might I choose between those two waveguides?

And any reason to doubt the accuracy or veracity of these plots?

https://www.eighteensound.it/system/dragonfly/production/2021/01/16/22_27_09_849_XT1464_Horizz_polar.JPG

https://www.eighteensound.it/system/dragonfly/production/2021/01/16/22_27_09_819_XT1464_Vertical_polar.JPG

I've been using waveguides & waveguide-style horns that are constant directivity (or nearly so) for about twenty years and imo passive crossovers work just fine. But if you're going to be doing the crossover yourself, active crossovers are a lot more user-friendly.

The plan is that whatever waveguide/driver combo I can decide on to best deliver clarity and imaging and sound stage realism while blending best with my midwoofers, Troy Crowe will build passive crossovers for.

Then for DACs for the main speaker amps and for my Rythmik subs, I first thought of this.
https://danvillesignal.com/dspnexus-dsp-audio-processor But Keith_W and even DSPNexus designer Al Clark warned me about Audio Weaver, which it runs on, being way too complex for DSP newbies. So, probably this. https://www.merging.com/products/interfaces/hapi and what software for newbies? Audiolense, which is "semiautomatic"?

I can see how far I can get on my own with room measuring for acoustic room correction and bass management. If I feel overwhelmed I will likely hire someone like https://accuratesound.ca/ and work with them to get it all done. At that point I might decide to biamp the system, provided that a good fail safe system to protect my hearing from high SPL (freak) accidents can be implmented.

 

[dped90]

Yeah, I agree! But 750 mW... Bob Daneliak's "Darling" amplifier (single-ended 1626 outputs) should do the trick!

see, e.g., https://evilscienceaudio.com/2013/11/03/single-ended-1626-darling-amplifier/

Wow, a tube amp proponent here at ASR? Blasphemy? Don't get me wrong, I think lots of tube amp offerings are wonderful, at least those which are about a clean as Pierre's VFET amp and/or the latest one he's built. https://www.diyaudio.com/community/members/pierrequiroule.9545/ And it's probably impossible finding any kind of solid state amp with wpc that low.

But aren't those tubes hard to source? Is it possible to get ~ 1 wpc from a single preamp tube, like the popular 6SN7? Or may just a ~ 3 wpc 2A3 amp? Look at me sounding like the tube expert that I hardly am.

 

[mhardy6647]

Easy to source, not (always) inexpensive. Chance favors the prepared mind.
The 2A3 has been in continuous production since the early 1930s.

My main amplifier is single-ended 2A3 and uses the 6SL7 (a twin triode not all that different to the aforementioned 6SN7, although the two triodes have higher amplification factor than do the 6SN7s) as driver. 3-ish watts per channel at reasonable (1%) levels of HD.
Yes, not mainstream ASR fodder.

 

[dped90]

I can't understand how limiting the signal at the DAC prevents hearing or driver damage compared to limiting at the driver. And the L-Pad at the driver reduces hiss, the filter capacitor reduces it even more, and also blocks DC in case you actually have a problem.

I completely agree.

I worry about damaging my hearing from listening to loud sounds for extended periods of time, for example public transportation which is surprisingly loud.

I always wear washable ear plugs whenever I am outdoors, even when in the car with windows up.

If you are worried about damaging your ears while fussing with your amps and drivers, wear hearing protection. Or turn the amp off while you are messing around. Otherwise the speakers will be fine as long as you don't abuse them.

No, what scares me is the chance, however tiny, that some kind of short may occur and with the drivers directly connected they would get the amplifiers' full power. But your solution put between the amplifier and drivers looks like a very good fail safe solution.

 

[dped90]

and also reduces the loudness of any on/off thumps that originate from the power amps.

Yes, a nice bonus. So, a circuit like MAB's looks like the right solution.

 

[dped90]

Resistor on the power amp output clobbers your damping factor. Not a good place to put it.

Sh*t! I knew there was likely a catch like this. But if the LPad network maintains the speaker impedance seen by amplifier then how could damping factor appreciably change?

Otherwise, what alternative?

 

[Newman]

Sh*t! I knew there was likely a catch like this. But if the LPad network maintains the speaker impedance seen by amplifier then how could damping factor appreciably change?

He's wrong. A simple dividing network could have that issue, but a properly calculated L pad network won't. That's what I meant by the relative simplicity of doing it between pre and power amps.

cheers

 

[MAB]

No problem in this application of an L-pad on a tweeter.

In the case of the D2, it also has a capacitor that rolls off the tweeter at ~800 Hz along with the L-pad. The passive network works with the DSP filter to produce the target crossover slope.

Here is the raw D2 response in red, the D2 with the above passive filter and L-pad in green, and the D2 with full DSP crossover and filters and the passive network in blue.

A full description including the JBL DSP is here. I used slightly different filters to get the blue response, but the approach is the same. I'm actually relying on the change in impedance for the net filter response, and in this case get a flatter impedance curve due to the L-pad. So no worries on the damping factor. And the passive network attenuates electronics noise dramatically. And the capacitor blocks any DC from the driver. If you didn't use the filter capacitor, you can use a larger value blocking capacitor to prevent DC and other nasty low frequency occurrences, just chose a value that puts the corner far below your crossover frequency.

Note: the above measurements were taken from the same relative position. They are normalized for SPL.

 

[egellings]

He's wrong. A simple dividing network could have that issue, but a properly calculated L pad network won't. That's what I meant by the relative simplicity of doing it between pre and power amps.

cheers

Between pre & power amps is the way to go, rather than at the speaker. L-pads do maintain a relatively constant load magnitude on the amplifier, but there is added resistance between the amp & speaker that lowers damping factor. The speaker won't see the tiny fraction of an ohm's worth of source impedance at the amplifier's output terminals through the attenuator (L-pad). I'm not wrong there. In many cases, though it may not matter, such as in a background (elevator music) setting.

 

[MAB]

People don't often talk about damping factor with regards to a midrange or tweeter, for good reason. It unfortunately it also commonly gets misused when describing the response of woofers. It's unfortunate naming: damping factor, since it evokes lots of incorrect analogies to what is actually happening electrically. We have a good thread on the basics. It's the ratio of the amp's output impedance to the impedance of the driver. And this is a discussion of a midrange/tweeter. I am not going to extend this to a woofer, that's a totally different post.

Let's look at what an L-pad does to a compression driver's impedance, and to the impedance of the resulting filter with the capacitor included. I measure the raw driver, the raw driver plus a filter capacitor, and repeat the measurements with and without the L-pad. Four measurements of impedance vs. frequency:

The L-pad negligibly raises the impedance below 400 Hz (red dotted vs. red solid traces) from 19 to 21 Ohms. This is nothing for a typical audio amplifier, which typically has 0.1 Ohm output impedance, even lower for some amps. Even if this was an 8 or 4 Ohm driver and was padded, still negligible for all but amps with absurdly high output impedance.

What is significant is the taming of the three resonant peaks in the impedance in the 700 Hz - 3 kHz region. This can be seen in the raw response (red lines) and the total response with the filter capacitor (green lines). These untamed peaks do lead to actual audible response variations that need to be dealt with additional filters in the passive or active crossover. It would be wrong to say those peaks are areas where the damping factor is really good, which is why DF is so misused here. It's likely some really high output impedance amps would struggle with the inductance of the unpadded driver, but amps with odd electrical characteristics is another topic.

Also, the L-pad significantly reduces hiss and noise. For the lowest noise amps (Benchmark, Hypex, PuriFi, to name a few), this isn't an issue. I compared some low-noise amps to some higher noise models, since noise is an audible, and passive components like L-pads and capacitors can make a fairly significant and audible difference in a quiet room, which might be why JBL uses them in the M2 and other active models.

None of this is elevator music only applications. The DF problem doesn't even closely apply to this application, there are other order of magnitude larger issues to deal with.

 

[Newman]

I am not going to extend this to a woofer, that's a totally different post.

But if the application is for a whole speaker not just one driver, you need to include the woofer.

 

[olegtern]

But if the application is for a whole speaker not just one driver, you need to include the woofer.

L-pad for a whole speaker?

 

[MAB]

But if the application is for a whole speaker not just one driver, you need to include the woofer.

The discussion was about use of an L-pad on a compression driver. Not on a woofer. Not on a whole speaker.

 

[Duke]

By less-loud replicas of the on-axis curve, you mean x & y plots like this?
https://www.audiosciencereview.com/...jbl-m2-reference-master-monitor-review.26051/

Yes.

A 60 degree arc, but 120 degree coverage angle?

No, the XT1464's total coverage angle is 60 degrees, so 30 degrees to either side of the centerline (horizontally).

What I meant was that I'd prefer a waveguide that could offer very good vertical coverage at both of those distances. How and why might I choose between those two waveguides?

The M2's waveguide has a much wider coverage pattern. It's specifications say about 120 degrees horizontal and 100 degrees vertical. Based on eyeballing the waveguide geometry, I suspect that 120 degrees horizontal pattern width ONLY applies within a fairly narrow horizontal band, and likewise I suspect that 100 degrees vertical pattern width ONLY applies within fairly narrow vertical band. I can explain why if you'd like. My point is, I think that if you are significantly off-axis BOTH horizontally and vertically, like if you stand up and walk to one side of the speaker's axis, I think the direct sound's frequency response will no longer be what the measurements predict. That being said, unless you plan to listen a lot from various standing locations in the room, I don't think it matters. In the horizontal plane that your ears will be in while you are sitting down, I think they will perform as the measurements predict.

The XT1464's coverage pattern is 60 degrees horizontal by 50 degrees vertical, so much smaller than the JBL M2's waveguide. Of course there is output beyond these angles, but the high frequencies start falling off more rapidly than the rest of the spectrum outside these angles.

Assuming the XT1464's horizontal coverage is wide enough for the listening locations you have in mind, such that it is at least theoretically a viable choice, imo your choice is between more room interaction (M2's horn) vs less room interaction (XT1464). And obviously if the XT1464's pattern coverage isn't wide enough, the JBL makes more sense.

And any reason to doubt the accuracy or veracity of these plots?

https://www.eighteensound.it/system/dragonfly/production/2021/01/16/22_27_09_849_XT1464_Horizz_polar.JPG

https://www.eighteensound.it/system/dragonfly/production/2021/01/16/22_27_09_819_XT1464_Vertical_polar.JPG

Those are consistent with my measurements.

But what is time-intensity trading and how does it maintain imaging integrity over the entire or most of the listening space (in the critical distance triangle) and how would the M2 and XT1464 compare in that regard?

Here is a post where I describe time/intensity trading:

Speakers for imperfect listening positions

I'm wondering what I should look for in speakers for a room where I really won't ever be sitting in the 'sweet spot'. Sometimes I'll be sitting to one side or the other (pretty much directly in front of a speaker, but 8' or so back). Sometimes I'll be on the floor below the speakers and up...

www.audiosciencereview.com

Here is a link to an online paper on the topic:

https://libinst.com/PublicArticles/Setup%20of%20WG%20Speakers.pdf

And here is a post where I talk about what I believe to be a depth-of-soundstage advantage of time/intensity trading as compared with using a dedicated center channel speaker:

Is there an ongoing discussion of phantom centers?

So, over the last month or so I have made many many changes to my home theater based on all that I am learning, and happily I'm learning more every day. One of my latest points of interest came from a new purchase. I recently upgraded my speakers from a Klipsch Reference Premiere bundle to a...

www.audiosciencereview.com

Either the M2's waveguide or the XT1464 will work in a time/intensity trading configuration, BUT the M2's waveguide will do so across a wider listening area.

 

[dped90]

Given their differing front form factors, would you expect the XT1464 to integrate better with my midwoofers than the M2, and thereby yeild more coherent sound?

Altec 416-8B in 100L Sealed

In this blog post I test the Altec 416-8B in a 100L sealed enclosure. Testing was done outdoors for the frequency response and then brought indoors for the distortion testing. The woofers are new from Great Planes Audio. A customer has sent me the finished cabinets with drivers to integrate...

josephcrowe.com

It's specifications say about 120 degrees horizontal and 100 degrees vertical. Based on eyeballing the waveguide geometry, I suspect that 120 degrees horizontal pattern width ONLY applies within a fairly narrow horizontal band, and likewise I suspect that 100 degrees vertical pattern width ONLY applies within fairly narrow vertical band. I can explain why if you'd like. My point is, I think that if you are significantly off-axis BOTH horizontally and vertically, like if you stand up and walk to one side of the speaker's axis, I think the direct sound's frequency response will no longer be what the measurements predict. That being said, unless you plan to listen a lot from various standing locations in the room, I don't think it matters. In the horizontal plane that your ears will be in while you are sitting down, I think they will perform as the measurements predict.

Standing up from my seated position, 9 to 11 ft from between the speakers while listening, I don't think that I would be moving beyond a 5 ft area. That said, here’s Erin’s review of the JBL M2. https://www.erinsaudiocorner.com/loudspeakers/jbl_m2/
and some discussion. https://www.diyaudio.com/community/threads/geddes-on-waveguides.103872/post-6793279

 

[Duke]

Given their differing front form factors, would you expect the XT1464 to integrate better with my midwoofers than the M2, and thereby yeild more coherent sound?

I think the M2's waveguide would integrate better with your Altec woofers.