The XSD Speaker XO thread is now up.
https://www.diyaudio.com/community/threads/xsd-speaker-crossover-and-waveguide-gb.386316/
Please discuss the GB there and keep this thread for technical stuff.
https://www.diyaudio.com/community/threads/xsd-speaker-crossover-and-waveguide-gb.386316/
Please discuss the GB there and keep this thread for technical stuff.
The Faital Pro 6FE100 is a lower cost pro audio driver with stamped steel basket, ferrite magnet, rubber surround and fiberglass cone. I like how smooth it sounds but there is higher distortion bump around 1000Hz (a very audible region). It probably does not have goodies like shorting rings etc. My favorite aspect about it is the fiberglass cone. I like the sound of fiberglass mids (ScanSpeak 10F/8424 for example). It is about 91 to 92dB sensitivit. This is a midbass and can play down to 70Hz probably in a vented box.
The PRV5MR450NDY is a higher quality pro audio driver with a cast basket, neodymium magnet motor, shorting rings (low distortion), paper cone with cloth surround, and very high sensitivity around 95dB. It is very low distortion and sounds very clear. It can be a bit rough above 7kHx but can serve almost as a fullrange driver with bandwidth well above 12kHz. The low distortion allows it to play loud without sounding strained. Plus it has a high Qm around 22 which gives it a great sense of dynamics. The PRV cannot play below 250Hz. It is a true midrange.
Both are great drivers but in the end, the PRV wins due to lower distortion.
The PRV5MR450NDY is a higher quality pro audio driver with a cast basket, neodymium magnet motor, shorting rings (low distortion), paper cone with cloth surround, and very high sensitivity around 95dB. It is very low distortion and sounds very clear. It can be a bit rough above 7kHx but can serve almost as a fullrange driver with bandwidth well above 12kHz. The low distortion allows it to play loud without sounding strained. Plus it has a high Qm around 22 which gives it a great sense of dynamics. The PRV cannot play below 250Hz. It is a true midrange.
Both are great drivers but in the end, the PRV wins due to lower distortion.
Hi Tomlang,
Thanks for your interest. It is a fairly straightforward speaker cabinet. Being an open baffle, there are only 3 sides and a base. The woofer array box is stacked MDF (or plywood) sheets and you are right - CNC is perfect for that. If you are interested in making one, consider signing up for the passive XO or CNC waveguide group buy.
Thanks for your interest. It is a fairly straightforward speaker cabinet. Being an open baffle, there are only 3 sides and a base. The woofer array box is stacked MDF (or plywood) sheets and you are right - CNC is perfect for that. If you are interested in making one, consider signing up for the passive XO or CNC waveguide group buy.
Hi XRK,
What is your process (if you are happy to share) for measuring the woofer output of the the XSD for the purpose of xover design? Do you measure at the slot (and presumably account for BS loss) or some other way? Does the open baffle nature affect how low freqs are best measured?
I've been playing with some DSP xovers recently, but am struggling to get a consistent low freq measurement that translates well in the room.
Thanks!
What is your process (if you are happy to share) for measuring the woofer output of the the XSD for the purpose of xover design? Do you measure at the slot (and presumably account for BS loss) or some other way? Does the open baffle nature affect how low freqs are best measured?
I've been playing with some DSP xovers recently, but am struggling to get a consistent low freq measurement that translates well in the room.
Thanks!
With a reflex speaker I measure the bass for the purposes of XO design at 0.5m on the axis of the driver of interest. Usually the tweeter in a conventional speaker and in axis with fullrange for a FAST/WAW. Below 100Hz, the bass is really dependent on room modes so I ignore that for development of the crossover. What I do use is the level near 100Hz as the nominal sensitivity, but this depends on placement (near wall or on stand, etc). It’s up to the designer to extract the bass below 100Hz to a level consistent with the sensitivity desired. If not, one has to use a shelving filter (baffle step compensation to attenuate the mids and highs to match the bass below 100Hz). To get the “anechoic response” of the bass without a proper anechoic chamber, I use Keele’s method of scaling the near field (at the port and the cone) vent and driver areas. A good write up on it by D’Apollito is here:
https://audioxpress.com/article/measuring-loudspeaker-low-frequency-response
Since the XSD is open baffle slot array, so it’s kind of a small vertical line array of bass, the room modes and floor bounce are not as significant. The bass is very consistent as you move about the room - I find this a major draw of the bass from the SLOB. The measurement of the bass output is then simply about the same as the near field (at 0.5m) that I use for the XO development. Ultimately, you have to tube the crossover (or voice it) to sound balanced. This is part is sort of the “black art” that is learned from experience of designing many speakers and crossovers.
One main thing to follow with designing speakers is to use the “Harman House Curve”. Let the response tilt down to the right. About -5dB from 50Hz or maybe 100Hz to 20kHz gives the best balance that is pleasing with all music types and is not fatiguing. Enhance the 100Hz to 500Hz range by 1.5 to 2dB for a richer more full sound on male vocals. Enhance 7k to 12kHz by a few dB for more sparkle. Enhance 1k to 3kHz by a few dB for more presence. Enhance above 15kHz for more “sparkle and air”. Some of the these enhancements will be natural to the driver choice. Hence the signature of the speaker. But keep it all within +/-2.5dB and tilted down with the Harman House Curve and you will have a great sounding speaker.
https://audioxpress.com/article/measuring-loudspeaker-low-frequency-response
Since the XSD is open baffle slot array, so it’s kind of a small vertical line array of bass, the room modes and floor bounce are not as significant. The bass is very consistent as you move about the room - I find this a major draw of the bass from the SLOB. The measurement of the bass output is then simply about the same as the near field (at 0.5m) that I use for the XO development. Ultimately, you have to tube the crossover (or voice it) to sound balanced. This is part is sort of the “black art” that is learned from experience of designing many speakers and crossovers.
One main thing to follow with designing speakers is to use the “Harman House Curve”. Let the response tilt down to the right. About -5dB from 50Hz or maybe 100Hz to 20kHz gives the best balance that is pleasing with all music types and is not fatiguing. Enhance the 100Hz to 500Hz range by 1.5 to 2dB for a richer more full sound on male vocals. Enhance 7k to 12kHz by a few dB for more sparkle. Enhance 1k to 3kHz by a few dB for more presence. Enhance above 15kHz for more “sparkle and air”. Some of the these enhancements will be natural to the driver choice. Hence the signature of the speaker. But keep it all within +/-2.5dB and tilted down with the Harman House Curve and you will have a great sounding speaker.
Thanks for the reply! That makes sense re bass. It certainly is a black art - I'm not sure the endless options afforded with DSP helps in this respect, although it is nice to be able to play around with different ideas. I'm crossing over woofers at 200hz currently which seems to work better than my passive iteration.
I think I've probably been working too hard to make things flat, rather than a more pleasing house curve as you suggest. It also hasn't helped that the rear cutout of my FR driver essentially creates a horn boosting around 1300hz (according to HR), which likely contributes to fatigue on certain tracks. Might need to put it back on the CNC at some point.
With your on-axis measurements of HF drivers, are you using windowed measurements? Or are you close enough to minimize room interference sufficiently?
I think I've probably been working too hard to make things flat, rather than a more pleasing house curve as you suggest. It also hasn't helped that the rear cutout of my FR driver essentially creates a horn boosting around 1300hz (according to HR), which likely contributes to fatigue on certain tracks. Might need to put it back on the CNC at some point.
With your on-axis measurements of HF drivers, are you using windowed measurements? Or are you close enough to minimize room interference sufficiently?
It will have more extended bass if near a corner but you diminish one of the main features of open baffle speakers: the sound radiates from the back as well and reflections from walls give a natural live performance soundstage presence. About 2-3 ft from wall is enough. I’m at about 2.5ft in my lab and it works great. The biggest change that I notice is that there is not a dependence on where you are in the room for bass modes. Normally stepping 3 ft to the side or back changes that with conventional monopole speakers. Now, it’s very uniform bass. Also, there seems to be no effect of floorbounce dip since the bass source is a taller line array of 4 slots.
