I'm working on a 4-way active system, where the tweeter is a Beyma TPL-150H crossed around 2kHz to a midrange crossed around 350Hz to double 10" driven by Hypex amp, crossed at 60Hz to a pair of subs.
My design goal has been using a stereo SET for the tweeters and a stereo SET for the mids. But, alas, I only have a push-pull tube amp and I haven't settled on the midrange driver.
So I'm thinking of building a 2-way unit with the TPL and the 10NDA610 driven by the tube amp. The system would then have 3-way active channels where the top channel would in turn have a passive crossover within. Active crossover generated on Acourate, applying room correction, time alignment, etc.
What kind of 2-way passive xo?
I'm starting with a 1st-order in an attempt to keep it simple and keep it time-aligned by sloping the front baffle.
From my reading, the general guideline seems to be the drivers should work well 2 octaves beyond the crossover point (2kHz).
Do you think this pair would be well suited for 1st order 2-way passive xo?
My design goal has been using a stereo SET for the tweeters and a stereo SET for the mids. But, alas, I only have a push-pull tube amp and I haven't settled on the midrange driver.
So I'm thinking of building a 2-way unit with the TPL and the 10NDA610 driven by the tube amp. The system would then have 3-way active channels where the top channel would in turn have a passive crossover within. Active crossover generated on Acourate, applying room correction, time alignment, etc.
What kind of 2-way passive xo?
I'm starting with a 1st-order in an attempt to keep it simple and keep it time-aligned by sloping the front baffle.
From my reading, the general guideline seems to be the drivers should work well 2 octaves beyond the crossover point (2kHz).
- TPL-150H: Beyma recommends 1kHz xo with min 12 dB/oct slope. So by 500Hz the recommendation is to be -12dB. If I use a 6dB/oct at 2kHz it would also be -12dB by 500Hz, but of course below 500Hz the 1st order would attenuate less than the recommended 2nd order. Beyma also rates this driver 80W AES above 1kHz. At over 100dB 1W @ 1m I would hardly put over 1W to it.
- 10NDA610:
- The reasons I'm inclined for this driver: said to sound very well, designed as midrange, extremely low Le for a pro driver (0.06mH with AIC engaged), high sensitivity (good for SET), very non-reactive impedance with AIC engaged (also good for SET). The downside is the 10" size and 2kHz xo frequency. I guess I need to try it.
- Not many measurements available about this driver, but OscarS recently measured it here, see post #6 and scroll down to the 4th graph past the pictures. Further down he has impedance graphed too. It seems to me 2kHz would work well and in fact the natural decay of the driver would help. A 10" driver past 2kHz will likely beam and how audible that will be remains a question.
Do you think this pair would be well suited for 1st order 2-way passive xo?
Given the steep breakup rise of this driver and 10dB at 2kHz difference between on axis and 45 deg off axis. I think you will struggle to get a good intergration between these drivers especially with a first order crossover.
Its likely you will need to make on axis compromises to acomodate the off axis variation. This pretty much has to be done by ear as its very difficult to predict how it will sound. However the very flat impedance should at least give you minimum error due to the SET output impedance.
Personally I would go at least 2nd order electrical and cross at 1KHz this reduces the off axis problem and will give you much greater control over the filter shape to correct the frequency response. A second order electrical gives you adjustable Q which I think you will need to intergrate the two drivers.
Regards,
Andy
Its likely you will need to make on axis compromises to acomodate the off axis variation. This pretty much has to be done by ear as its very difficult to predict how it will sound. However the very flat impedance should at least give you minimum error due to the SET output impedance.
Personally I would go at least 2nd order electrical and cross at 1KHz this reduces the off axis problem and will give you much greater control over the filter shape to correct the frequency response. A second order electrical gives you adjustable Q which I think you will need to intergrate the two drivers.
Regards,
Andy
Thanks for the input, Andy.
The 10dB off axis at 2kHz you mean on the 10NDA610 I imagine, but where are you seeing those 10dB? Not sure if you are looking at the same graphs I am and I didn't figure it out or you have an alternative source of info - on which I would be very interested.
I'm not as versed on passive crossovers so bear with me please. I will look into 2nd order. I must admit 1kHz seems low, though.
This is the data I found on the midrange which shows 10dB variance.
http://www.eighteensound.com/Portals/0/PDFs/10NDA610.PDF
1kHz is low but if you are driving this from a SET I am assuming the power is bery low and the tweeter is rated 80W. I would be supprised if you have more than 10W from your SET. So the tweeter should be able to cope especially if you use a 2nd order electrical as propose in the tweeter data sheet.
http://www.eighteensound.com/Portals/0/PDFs/10NDA610.PDF
1kHz is low but if you are driving this from a SET I am assuming the power is bery low and the tweeter is rated 80W. I would be supprised if you have more than 10W from your SET. So the tweeter should be able to cope especially if you use a 2nd order electrical as propose in the tweeter data sheet.
The only variance that can't be dealt with a XO level is the relation between on- and off- axis response. And this one would allow an xov point higher than 1 kHz IMO.
Can you explain further? Are you saying all neo drivers are 'worse' than their ferrite equivalent?
To conclude that neo is generally harsher than ferrite from this comparison of 12" drivers in a guitar combo is definitely flawed ! You compare a woofer with higher Qts and less upper end extension to a midbass with lower Qts, higher upper Extension and lower Le. This is definitely not a pure ferrite to neo comparison !
Neo can - but does not necessarily have to - sound better than ferrite. Keep in mind that neo is an electrical conductor and it also allows a higher saturation of the pole-Piece and pole plate.
Regarding crossover: I would cross around 1.5 kHz becaue the driver is not too directional at that frequency yet. And that half octave leeway from 1 kHz makes the AMT work less hard. A first order (or better approximated first order) crossover isn't possible with that driver because the rising response of the bare driver is more or less compensating a first order rollof. I.e. it would more or less equalise the response flat up to almost 4 kHz. I would use a 2nd order lowpass with a very low Q wirth a subsequent early rolloff that compensates for the rising response of the driver.
Regards
Charles
Neo can - but does not necessarily have to - sound better than ferrite. Keep in mind that neo is an electrical conductor and it also allows a higher saturation of the pole-Piece and pole plate.
Regarding crossover: I would cross around 1.5 kHz becaue the driver is not too directional at that frequency yet. And that half octave leeway from 1 kHz makes the AMT work less hard. A first order (or better approximated first order) crossover isn't possible with that driver because the rising response of the bare driver is more or less compensating a first order rollof. I.e. it would more or less equalise the response flat up to almost 4 kHz. I would use a 2nd order lowpass with a very low Q wirth a subsequent early rolloff that compensates for the rising response of the driver.
Regards
Charles
They are in Russian in various forums reviews. In any case, if you think that there is no deterioration in quality then there is nothing to worry about.
You see, manufacturers do not make the same Thiele/Small parameters for comparing the sound of ferrite and neodymium speakers, so I had to choose from what was. In this forum, I did not find a discussion where ferrite sounded better than neodymium, even on the internet, so it seems that this is a feature of Russians, but there is no total rejection of the neodymium, they are easily placed at home.
Thank you. OK, so 2nd order at 1.5kHz it will be. You mention low Q. Would that be a 2nd order Linkwitz-Riley with a Q=0.49?
The horisontal beamwidth of TPL 150H is said to be 80 degrees between about 1,2 kHz to 10 kHz. A 10” driver has about 21 cm cone diameter, so a 90 degree / - 6dB beamwidth for it should be around 1620 Hz before beaming starts. If one aims at 80 degree beamwidth as crossover frequency between the 2 drivers, it should be around 1780 Hz for a 10” driver with Ø 21 cm / 345 cm² cone area. Having a good margin downwards, I believe would be beneficial for the TPL 150H.
(Now, I’m not very proficient when it comes to speakers or crossovers, so take above with a pinch of salt. There are certainly others at the forum who could chime in and chop me down with some good arguments against 1780 Hz as crossover frequency.)
(Now, I’m not very proficient when it comes to speakers or crossovers, so take above with a pinch of salt. There are certainly others at the forum who could chime in and chop me down with some good arguments against 1780 Hz as crossover frequency.)
I agree. With the vast 10s that I'm testing, I've found that keeping the XO point dial (on my dbx 231xs) right near the ~1.4kHz marker (or the next dial indent above it) creates a very nice transition to the horn-loaded Hygeia RT-62041 tweeter, which is not rated near as low as the TPL-150H, but I make it operate low by way of techniques, which can also be applied to the TPL-150H since it has such a high 2.83V sensitivity. I would say it all sort of depends how much beaming we are willing to accept from the larger cone-diameter low-frequency driver (due to which ever factors one chooses to consider).
The beauty of high sensitivity tweeters is that the output can be notched (in multiple areas) where needed to artificially extend the low-end operating point, so long as one is willing to accept the typically higher distortion% that sometimes occurs when using it below it's rated pass-band. I'm almost 99.999999% positive that this can be applied using several parallel notch filters in passive XO as well, as opposed to attenuating the entire response down to "meet" the low-frequency driver. Has anyone else played with this kind of technique?
Here are the pictures from a book.
Attachments
- Status
- Not open for further replies.