Some of this was discussed by others in this thread https://www.diyaudio.com/community/threads/18-sound-nd3st-compression-driver.300974/
but in its five pages, it meandered off. It was triggered by the Voice Coil review that showed extremely low distortion and good CSD of the ND3ST. Someone suggested that maybe it was measured at lower than stated 104 dB / 1 m, measured with the microphone 10 cm from the mouth of the horn. First of all, the fundamental level in the distortion plot is only roundabout 107 dB (as compared to ~126 dB for the NSD1095N measured a few years earlier under the same nominal conditions), and then the drive level to achieve 104 dB / 1 m was given as only 0.42 V for the ND3ST (and 2.2 V for the NSD1095N, both being 8 Ohm models). A 4015 series driver also measured in Voice Coil was similar to the 1095 in terms of sp in the distortion plot and drive voltage.
Ok, with that out of the way, I found 7 ND3 and 3 ND1460/80 drivers and compiled their key data and impedance and FR plots. I may have missed other 1.4 / 3 in drivers in their huge catalogue. Edit: I missed the ND1460/80A.
but in its five pages, it meandered off. It was triggered by the Voice Coil review that showed extremely low distortion and good CSD of the ND3ST. Someone suggested that maybe it was measured at lower than stated 104 dB / 1 m, measured with the microphone 10 cm from the mouth of the horn. First of all, the fundamental level in the distortion plot is only roundabout 107 dB (as compared to ~126 dB for the NSD1095N measured a few years earlier under the same nominal conditions), and then the drive level to achieve 104 dB / 1 m was given as only 0.42 V for the ND3ST (and 2.2 V for the NSD1095N, both being 8 Ohm models). A 4015 series driver also measured in Voice Coil was similar to the 1095 in terms of sp in the distortion plot and drive voltage.
Ok, with that out of the way, I found 7 ND3 and 3 ND1460/80 drivers and compiled their key data and impedance and FR plots. I may have missed other 1.4 / 3 in drivers in their huge catalogue. Edit: I missed the ND1460/80A.
Attachments
Last edited:
The ND1460 and 1480 share the same diapraghm, as evidenced by the spare part numbers. The 1480 just happens to have a stronger motor (2.2 instead of 1.9 T). Curiosly, this raises sensitivity only from 109 to 110 dB. The SPL plot in the 1460 data sheet is graphically mangled but the plots do look quite similar. The 1480 plot has a few more 1 dB wiggles at 2, 4 and 6 kHz than does the 1460. I am not even sure if the same smooting was used. German online magazine Hifi-Selbstbau measured the 1460 on a Limmer 033 horn. The report several 2-3 dB wiggles in the 2 - 8 kHz range. Not sure if this is due to less smoothing or the other horn model. The 10 - 20 kHz octave looks different, with the dip occuring at 11.5 rather than 14 kHz in the data sheet and the peak at 15 rather than 17 kHz, but the dip looks shallower.
The impedance curves of the 1480 in the data sheet has more pronounced fundamental and secondary maxima, as is to be expected with a stronger motor. So again, it is surprising that the motor added only 1 dB sensitivity.
The NSD1480N sports the TiN coating on the Ti dome, with the motor having the same 2.2 T field as in the ND1480. I would have thought the TiN coating would make the dome heavier but it actually increased 1 dB to 111 dB sensitivity. Maybe the used thinner titanium foil?
The frequency response up to 10 kHz looks a tad smoother than the 1480's, more like the 1460's, but with the 1460 and 1480 sharing the same part, this is probably smoothing or errant reflections in the lab. The TiN dome does have a shallower 14 kHZ dip and a tad more extension beyond 17 kHz going for it.
The impedance curves of the 1480 in the data sheet has more pronounced fundamental and secondary maxima, as is to be expected with a stronger motor. So again, it is surprising that the motor added only 1 dB sensitivity.
The NSD1480N sports the TiN coating on the Ti dome, with the motor having the same 2.2 T field as in the ND1480. I would have thought the TiN coating would make the dome heavier but it actually increased 1 dB to 111 dB sensitivity. Maybe the used thinner titanium foil?
The frequency response up to 10 kHz looks a tad smoother than the 1480's, more like the 1460's, but with the 1460 and 1480 sharing the same part, this is probably smoothing or errant reflections in the lab. The TiN dome does have a shallower 14 kHZ dip and a tad more extension beyond 17 kHz going for it.
The ND3 series have a copper cap, 2 to 2.1 T field in gap and 112 dB sensitivity. Their min crossover is given as 1.2 kHz (sometimes with min 12 and sometimese with min 24 dB slope), as opposed to 0.8 kHz (12 dB) for all 1460/1480 drivers. If they achieve 2 dB more sensitivity with 0.2 T less field, it is likely that the winding length of the voice coil was shortened. This would also explain the increased minimum crossover.
Interestingly, the copper cap and the shorter coil should result in less inductive rise of the impedance curve. However, all ND3 drivers go from 6 R at 100 Hz to about 9.5 R at 20 kHz, whereas the 1460/1480 go from 6 to just 7 R. How to explain this??
The ND3T has an uncoated Ti dome, so it should be the direct equivalent to the 1460/1480. The FR looks a little busy in the 10 - 20 kHz octave but these are very narrow wiggles, at it holds up to 20 kHz (and even picks up beyond 15 kHz). The wiggles indicate break up modes, but they seem well controlled and result in more HF output, at least on axis. The downside is that that there is 2-3 dB ripple all the way from 1.5 kHz up, much busier than the 1460/1480. Maybe part of the sensitity increase is due to a thinner and lighter dome?
Interestingly, the copper cap and the shorter coil should result in less inductive rise of the impedance curve. However, all ND3 drivers go from 6 R at 100 Hz to about 9.5 R at 20 kHz, whereas the 1460/1480 go from 6 to just 7 R. How to explain this??
The ND3T has an uncoated Ti dome, so it should be the direct equivalent to the 1460/1480. The FR looks a little busy in the 10 - 20 kHz octave but these are very narrow wiggles, at it holds up to 20 kHz (and even picks up beyond 15 kHz). The wiggles indicate break up modes, but they seem well controlled and result in more HF output, at least on axis. The downside is that that there is 2-3 dB ripple all the way from 1.5 kHz up, much busier than the 1460/1480. Maybe part of the sensitity increase is due to a thinner and lighter dome?
Now let's look at the differences within the ND3 family.
The ND3ST has 2.1 T instead of 2.0 T for the ND3T but this does not result in higher sensitivity. The prose only mentions a "proprietary suspension geometry". The FR looks a little smoother below 10 kHz but has the dip at 14 kHz and the peak at 17 kHz we have come to know from the 1460/1480.
The ND3A looks a lot like the ND3T but smoother. There is no difference at all in the prose between the ND3A and ND3SA. There isn't any mention of the suspension nor an increase to 2.1 T either. The curves look extremely similar maybe with an extra microdip at 7 kHz and less of a microdip at 12 kHz for the ND3SA.
Then we have the ND3N which is the TiN coating on Ti. AS with the NSD1480N, it looks ultra-smooth below 10 kHz and really straightens out the top octave. Again, adding the TiN curiosly does not drop sensitivity.
The ND3SN adds 0.1 T field but makes no mention of an improved suspension or any other supension. The curves look very similar, with maybe an extra 1 dB added to the 2 kHz dip in the SN.
Lastly, there is the BE which is discontinued. With a spectrum that is probably the busiest of the lot aside from the simple T, I can understand why.
So, what does the mysterious S in the Ti, Alu and TiN models stand for? It supposedly introduces a better suspension, but only in the ST. It adds 0.1 T, but only in the ST and SN. This added field does nothing to increase the sensitivity. It changes the FR of the Ti dome a little but has miniscule effect on the curves of the SA and SN models.
The ND3ST has 2.1 T instead of 2.0 T for the ND3T but this does not result in higher sensitivity. The prose only mentions a "proprietary suspension geometry". The FR looks a little smoother below 10 kHz but has the dip at 14 kHz and the peak at 17 kHz we have come to know from the 1460/1480.
The ND3A looks a lot like the ND3T but smoother. There is no difference at all in the prose between the ND3A and ND3SA. There isn't any mention of the suspension nor an increase to 2.1 T either. The curves look extremely similar maybe with an extra microdip at 7 kHz and less of a microdip at 12 kHz for the ND3SA.
Then we have the ND3N which is the TiN coating on Ti. AS with the NSD1480N, it looks ultra-smooth below 10 kHz and really straightens out the top octave. Again, adding the TiN curiosly does not drop sensitivity.
The ND3SN adds 0.1 T field but makes no mention of an improved suspension or any other supension. The curves look very similar, with maybe an extra 1 dB added to the 2 kHz dip in the SN.
Lastly, there is the BE which is discontinued. With a spectrum that is probably the busiest of the lot aside from the simple T, I can understand why.
So, what does the mysterious S in the Ti, Alu and TiN models stand for? It supposedly introduces a better suspension, but only in the ST. It adds 0.1 T, but only in the ST and SN. This added field does nothing to increase the sensitivity. It changes the FR of the Ti dome a little but has miniscule effect on the curves of the SA and SN models.
Last edited:
In reality...not much at all. Slightly more powerful motor in the SN but it doesn't seem to make much of a difference.
The smaller ones are meant to be arrayed in line array waveguides.
The smaller ones are meant to be arrayed in line array waveguides.
Smaller ND3 or do you mean the ND2 and ND1 series? And if S is only about the motor, why does the ST's FR look better than the T's? The ST also is the only S model where they claim a proprietary suspension (without saying that the T does not have the same suspension).
Smaller ND3.
I wouldn't get too caught up with the marketing materials. They aren't always totally accurate and changes from batch to batch be made at anytime.
The best available material for a dome is Beryllium but it's still very expensive. TiN offers great benefits over most materials at a much more tolerable price.
What's your specific application?
I wouldn't get too caught up with the marketing materials. They aren't always totally accurate and changes from batch to batch be made at anytime.
The best available material for a dome is Beryllium but it's still very expensive. TiN offers great benefits over most materials at a much more tolerable price.
What's your specific application?
Thanks, though I am still trying to understand what the smaller ND3 are? Non-S? physically smaller motor?
From the published curves, Beryllium seems to be as poor at the ND3S, and it's been discontinued anyway.
Applicaton is home hifi with an XT1464 or Limmer 033 horn. I have ND1460s but was curious about improved membranes and motors.
From the published curves, Beryllium seems to be as poor at the ND3S, and it's been discontinued anyway.
Applicaton is home hifi with an XT1464 or Limmer 033 horn. I have ND1460s but was curious about improved membranes and motors.
The smaller ND3's simply have less magnet material so that they can be closely arrayed in large format prosound line arrays.
The nitride diaphragms are clearly superior to Radian aluminum alloy diaphragms in direct comparisons. There is nothing very obvious when comparing measurements but the nitride diaphragms sound more detailed. Less hash.
If sticking with 18 Sound drivers, go with a nitride coated dome. The ND3 series will easily go down to 800hz in a home setting.
Other interesting drivers are the HF1440, N314X, and Oberton composite diaphragms.
The nitride diaphragms are clearly superior to Radian aluminum alloy diaphragms in direct comparisons. There is nothing very obvious when comparing measurements but the nitride diaphragms sound more detailed. Less hash.
If sticking with 18 Sound drivers, go with a nitride coated dome. The ND3 series will easily go down to 800hz in a home setting.
Other interesting drivers are the HF1440, N314X, and Oberton composite diaphragms.
Last edited:
Now I get it, the drivers without the S have only 109 mm diameter as opposed to 120 mm in the S models. Difference in field strength seems minimal 2.0 vs. 2.1 T (with the exception of the ND3SA), no difference in sensitivity. I haven't been able to find replacement diaphraghms. Are they the same between S and non-S? If so, we can conclude the bigger motor makes essentially no difference.
Still amazing the ND3ST looks so much cleaner than the ND3T.
Still amazing the ND3ST looks so much cleaner than the ND3T.
After considering the ND3N or ND3SN for my next project I emailed 18Sound to determine if there was an audible difference.
ErnieM already shared the correct answer. Here is the response I received from18Sound Application engineer:
"the two drivers are basically the same, the ND3 is an even more compact version that we realized thinking about linearrays specially but you can consider them the very same component in two different formats.
There is no audible difference between them, only a slightly different power handling rate due to the dimension of the magnet, being a bit bigger in the "S" version it allows a better heat dissipation, that's all"
ErnieM already shared the correct answer. Here is the response I received from18Sound Application engineer:
"the two drivers are basically the same, the ND3 is an even more compact version that we realized thinking about linearrays specially but you can consider them the very same component in two different formats.
There is no audible difference between them, only a slightly different power handling rate due to the dimension of the magnet, being a bit bigger in the "S" version it allows a better heat dissipation, that's all"