I’m aiming for a maximum capability of about 110 dB peaks, in an active 3 way (open baffle). I’ve done the bass baffles and am about to cut the mid-tweeter mounting frames.
The mids will be Neo10’s, the tweeters open back Neo3’s. Both chosen as they are dipole and very low distortion; already purchased.
The sensitivity of the Neo10’s is about 2 dB higher; and the much bigger Neo10’s have power handling of 100 watts, while the tweeters are only 20 watts. In simple terms then the maximum capability of the tweeters is about 8 db less, about 103 dB.
How to get the tweeters to about the same maximum capability of c 110 dB?
fyi specs :Bohlender Graebener Neo3-PDRW Planar Tweeter w/Back Cup
The nominal impedance of the tweeters is 3.5 ohms, so paralleling two of these squarish drivers in a two tweeter “line array” ie on top each other would be a tough load for most amps. Series connection wont do the job.
I have a few spare good amps: so if two tweeter drivers are used per side, each tweeter with it's own channel of amplification - it’s not a normal + 3 dB parallel arrangement with one amp shared . .
So what would be the maximum SPL of this set up? is it + 6 dB(?)
The mids will be Neo10’s, the tweeters open back Neo3’s. Both chosen as they are dipole and very low distortion; already purchased.
The sensitivity of the Neo10’s is about 2 dB higher; and the much bigger Neo10’s have power handling of 100 watts, while the tweeters are only 20 watts. In simple terms then the maximum capability of the tweeters is about 8 db less, about 103 dB.
How to get the tweeters to about the same maximum capability of c 110 dB?
fyi specs :Bohlender Graebener Neo3-PDRW Planar Tweeter w/Back Cup
The nominal impedance of the tweeters is 3.5 ohms, so paralleling two of these squarish drivers in a two tweeter “line array” ie on top each other would be a tough load for most amps. Series connection wont do the job.
I have a few spare good amps: so if two tweeter drivers are used per side, each tweeter with it's own channel of amplification - it’s not a normal + 3 dB parallel arrangement with one amp shared . .
So what would be the maximum SPL of this set up? is it + 6 dB(?)
Last edited:
High Frequency Driver Power Handling | Eminence Speaker
Seems reasonable enough. Might negate the need for dual tweeters.
The reason there's so little power in the HF (so far as I can work out)...
We don't hear the peak amplitude of a waveform. We hear the steepness of the waveform (the gradient of the wave when it passes through zero). For a given frequency, a larger amplitude has a steeper gradient, so it sounds louder. Now play a signal of higher frequency. (Don't change the volume control of the signal generator), and it's a much smaller amplitude, yet sounds as loud. Because its a higher frequency, the gradient as it passes through 0v is much steeper, but the amplitude (and therefore power input) is much smaller.
Its easy to see with a scope and a signal generator.
Chris
Seems reasonable enough. Might negate the need for dual tweeters.
The reason there's so little power in the HF (so far as I can work out)...
We don't hear the peak amplitude of a waveform. We hear the steepness of the waveform (the gradient of the wave when it passes through zero). For a given frequency, a larger amplitude has a steeper gradient, so it sounds louder. Now play a signal of higher frequency. (Don't change the volume control of the signal generator), and it's a much smaller amplitude, yet sounds as loud. Because its a higher frequency, the gradient as it passes through 0v is much steeper, but the amplitude (and therefore power input) is much smaller.
Its easy to see with a scope and a signal generator.
Chris
Chris
Yes power handling is related to XO frequency and slope. And more sensitive and/ or higher power handling drivers could solve the maximum SPL issue.
However I will use the - already purchased - open back Neo3’s because of their combination of dipole radiation pattern and very low measured distortion (eg by Zaph), which will bring benefits at any volume level. I intend to cross them low and steep.
Being active crossovers, I could easily go 6th even 8th order.
That still leaves the issue of boosting the chosen Neo3’s maximum capability by about 7 dB . .
Yes power handling is related to XO frequency and slope. And more sensitive and/ or higher power handling drivers could solve the maximum SPL issue.
However I will use the - already purchased - open back Neo3’s because of their combination of dipole radiation pattern and very low measured distortion (eg by Zaph), which will bring benefits at any volume level. I intend to cross them low and steep.
Being active crossovers, I could easily go 6th even 8th order.
That still leaves the issue of boosting the chosen Neo3’s maximum capability by about 7 dB . .
Last edited:
paralleling drivers, each with their own amplifier will give +3dB if you feed in double the total power.
There is a further gain due to twin sources. For Bass drivers and lower Mids, there is an extra +3dB available if the spacing (centre to centre of the sources) is less than 1/4 wavelength of the frequency being reproduced. This extra gain falls gradually as the spacing is increased to 1/2 a wavelength of the frequency being reproduced.
Another factor comes in as the reproduced wavelength is reduced cf. the source spacing.
The output is no longer a wide lobe of sound but becomes an interference pattern where the lobes narrow with nulls and peaks as one travels around the lobe pattern. Multi source Treble suffers badly from this "lobing". Line arrays are an attempt to remove this lobing but I think they become effective when the ratio of spacing : frequency remains very low and the overall length : spacing is kept very large. Two drivers does not make a line array.
How is the 20W specified? The peak transient power handling may be 20W or slightly higher/lower or very much greater.
Continuous narrow band HF is very low in avearge power level. Peak transient HF can be very high, I contend from my interpretation of various discussions that the treble peaks can be just as severe as the Mid and LF peaks.
If power compression does not set in, you may be able to achieve much higher than 103dB pk with this driver.
There is a further gain due to twin sources. For Bass drivers and lower Mids, there is an extra +3dB available if the spacing (centre to centre of the sources) is less than 1/4 wavelength of the frequency being reproduced. This extra gain falls gradually as the spacing is increased to 1/2 a wavelength of the frequency being reproduced.
Another factor comes in as the reproduced wavelength is reduced cf. the source spacing.
The output is no longer a wide lobe of sound but becomes an interference pattern where the lobes narrow with nulls and peaks as one travels around the lobe pattern. Multi source Treble suffers badly from this "lobing". Line arrays are an attempt to remove this lobing but I think they become effective when the ratio of spacing : frequency remains very low and the overall length : spacing is kept very large. Two drivers does not make a line array.
How is the 20W specified? The peak transient power handling may be 20W or slightly higher/lower or very much greater.
Continuous narrow band HF is very low in avearge power level. Peak transient HF can be very high, I contend from my interpretation of various discussions that the treble peaks can be just as severe as the Mid and LF peaks.
If power compression does not set in, you may be able to achieve much higher than 103dB pk with this driver.
Last edited:
Andrew,
Thanks, * Peak power handling is actually 50 watts. So that would get it to around 105 dB.
> paralleling drivers, each with their own amplifier will give +3dB if you feed in double the total power.
So using separate amps is no different to using one amp of twice the power . .
Though that’s if the amp can handle 1.8 ohms – only above 2 kHz. What might that approximate to in a full amplifier range sense, where the minimum impedance might be in the midbass, where much more sound energy is present?
a) Maybe 3 ohms, ie marginal for many amplifiers; or
b) 4 ohms, ie acceptable for many amplifiers
If b) is right, in this case where drivers handling 50 watts peaks, with a minimum of 2 ohms: one amplifier per tweeter capable of 100 watts into 8 ohms, 50 watts into 2 ohms would equal the output of a single paralleled amp capable of 200 watts into 8 ohms, 100 watts into 2 ohms?
Centre to centre of the 2 tweeters would be 90 mm, which I think equates to 3810 Hz. The XO would be at about half that frequency, 2000 Hz
I don’t know the effect of the two driver Line array arrangement – only two drivers, hence not much?
I want to be below power compression, even on the peaks – sounds much better, I think
Thanks, * Peak power handling is actually 50 watts. So that would get it to around 105 dB.
> paralleling drivers, each with their own amplifier will give +3dB if you feed in double the total power.
So using separate amps is no different to using one amp of twice the power . .
Though that’s if the amp can handle 1.8 ohms – only above 2 kHz. What might that approximate to in a full amplifier range sense, where the minimum impedance might be in the midbass, where much more sound energy is present?
a) Maybe 3 ohms, ie marginal for many amplifiers; or
b) 4 ohms, ie acceptable for many amplifiers
If b) is right, in this case where drivers handling 50 watts peaks, with a minimum of 2 ohms: one amplifier per tweeter capable of 100 watts into 8 ohms, 50 watts into 2 ohms would equal the output of a single paralleled amp capable of 200 watts into 8 ohms, 100 watts into 2 ohms?
Centre to centre of the 2 tweeters would be 90 mm, which I think equates to 3810 Hz. The XO would be at about half that frequency, 2000 Hz
I don’t know the effect of the two driver Line array arrangement – only two drivers, hence not much?
I want to be below power compression, even on the peaks – sounds much better, I think
Last edited:
I think I’m at around 108 dB.
c. 105 dB at Peak power handling + 3dB from paralleling drivers
Personally I have more and better quality options for 2 amplifiers capable of 100 watts into 8 ohms, say 80 W into 4, and 50 W into 2 ohms ~ rather than a single amp of twice that power.
I've quoted driver Watts at Peak handling; and amplifier Watts at RMS. But that should be a reasonable buffer against amplifier clipping.
c. 105 dB at Peak power handling + 3dB from paralleling drivers
Personally I have more and better quality options for 2 amplifiers capable of 100 watts into 8 ohms, say 80 W into 4, and 50 W into 2 ohms ~ rather than a single amp of twice that power.
I've quoted driver Watts at Peak handling; and amplifier Watts at RMS. But that should be a reasonable buffer against amplifier clipping.
Okay, from the link I posted:
The HF drivers will take 20w continuous. A crossover at 3.5 - 5kHz means they see 15% of the total power.
So, 20w = 15%
133.333w = 100% power. Work to that power handling to compare to woofers.
One per side is more than enough: 90dB@2w will take you to 107dB@100w
A higher crossover point (>5kHz) will give you effectively 200w power handling, which is another 3dB. 110dB is enough.
The HF drivers will take 20w continuous. A crossover at 3.5 - 5kHz means they see 15% of the total power.
So, 20w = 15%
133.333w = 100% power. Work to that power handling to compare to woofers.
One per side is more than enough: 90dB@2w will take you to 107dB@100w
A higher crossover point (>5kHz) will give you effectively 200w power handling, which is another 3dB. 110dB is enough.
This is an average figure for most types of music, played without clipping.
In effect this 15% allows one to size the heatsinks of the amplifiers for playing average levels in your home/PA environment.
That 15% has nothing to do with comparing peak signal levels, nor of transient levels.
Yes, assuming they are adding in phase the pressure will double and you will get +6dB.
Not to confuse you but if they are seperated in distance you would get double the radiated power, or +3dB when averaged on all axies. If the units are placed close together then they will add in phase everywhere for the +6dB. Although +6dB seems like 4 times the power, it isn't. 3dB comes from power doubling and 3 dB more from directivity doubling.
David S.
David,
That’s good to hear, Thank you.
So at the power handling maximum of 20 watts “Program level” (rather than the maximum of 50 watts “Peak level” - Ie hopefully that will be a reasonable ‘buffer’ against driver compression and distortion) the max is c. 103 dB.
+ Double the drivers (with appropriate amplification) for adjacent drivers for + 6 dB, gets us to c. 109 dB,
Close enough to the target, with the drivers hopefully unstressed/ not compressed; and with enough power so the amps don't clip.
I just recalled Linkwitz writing that double eg a crossover frequency, and the driver excursion requirements (for a cone - Neo's are planar magnetic, I’d think the same applies).
So if the XO frequency were raised by a quarter (from say 1600 Hz to 2000 Hz) driver excursion would reduce by 1 - 1/ 1.25^2, ie 36%. So moving the XO frequency up just marginally should reduce any driver ‘stress’ that occurs.
That’s good to hear, Thank you.
So at the power handling maximum of 20 watts “Program level” (rather than the maximum of 50 watts “Peak level” - Ie hopefully that will be a reasonable ‘buffer’ against driver compression and distortion) the max is c. 103 dB.
+ Double the drivers (with appropriate amplification) for adjacent drivers for + 6 dB, gets us to c. 109 dB,
Close enough to the target, with the drivers hopefully unstressed/ not compressed; and with enough power so the amps don't clip.
I just recalled Linkwitz writing that double eg a crossover frequency, and the driver excursion requirements (for a cone - Neo's are planar magnetic, I’d think the same applies).
So if the XO frequency were raised by a quarter (from say 1600 Hz to 2000 Hz) driver excursion would reduce by 1 - 1/ 1.25^2, ie 36%. So moving the XO frequency up just marginally should reduce any driver ‘stress’ that occurs.
This isn't pink noise music doesn't have equal power between octaves, you have to plan for the worst what happens if all the peak in a song came in that 3.5 - 5KHz band? You would effectively have 100% of the power being put through that driver.
We don't hear the peak amplitude of the wave. We hear the rate of change of pressure, ie, the gradient of the curve (as seen on a scope) as it passes through zero.
A 100Hz wave played next to a 1000Hz wave at the same SPL will have different peak amplitude. But the gradient of the wave as it passes zero will be equal. That's why you need huge amounts of power (and air displacement) for low frequencies: to get the gradient as steep at a higher frequency, the amplitude must be massive.
If you're looking for 110dB continuous, I think something a little more sturdy is needed. Running things right on the edge of what they can handle isn't a good idea.
Why do you want that sort of SPL anyway?
A 100Hz wave played next to a 1000Hz wave at the same SPL will have different peak amplitude. But the gradient of the wave as it passes zero will be equal. That's why you need huge amounts of power (and air displacement) for low frequencies: to get the gradient as steep at a higher frequency, the amplitude must be massive.
If you're looking for 110dB continuous, I think something a little more sturdy is needed. Running things right on the edge of what they can handle isn't a good idea.
Why do you want that sort of SPL anyway?
Yes; except that I'd rather use the more conservative power handling figure of 20 w, to ensure sound that is hopefully also clean.
I also ignored the gain from one either side, as I ignored the loss from distance from speaker to the ears.
[/QUOTE] As a background question, does it happen that the mid and woofer would be contributing zero to the SPL with some musical passages? [/QUOTE]
If I understand you correctly, that's the rolloff of the mids as the crossover merges them into the tweeters. A good point.
In my project, the midbass I will use is certainly extended on-axis, up to about 4 kHz. If only a first order is used on that, there will be added SPL on-axis.
> Why do you want that sort of SPL anyway?
Ive been asked that before:
For those rare times when I have the house to myself, playing my favourite albums, and want to move around the house listening "loud and clear"
> Running things right on the edge of what they can handle isn't a good idea.
I fully agree. Hence not using the manufacturer's Peak power handling figure; and its only for 5 - 10 % of usage.
You may buy a car capable of very fast acceleration, but rarely use it all . .
Ive been asked that before:
For those rare times when I have the house to myself, playing my favourite albums, and want to move around the house listening "loud and clear"
> Running things right on the edge of what they can handle isn't a good idea.
I fully agree. Hence not using the manufacturer's Peak power handling figure; and its only for 5 - 10 % of usage.
You may buy a car capable of very fast acceleration, but rarely use it all . .
Last edited:
In other rooms.... I think you need PA equipment. What difference does quality mean if you are down the hall listening?
On the other hand, I'm a big fan of tweeter power. One of my favourite tweeters was made from the best friend egg dome I ever tested (fabulous tone burst performance) - 9 per side.
Say... that makes impedance matching a snap! Not to mention dispersion control.
On the other hand, I'm a big fan of tweeter power. One of my favourite tweeters was made from the best friend egg dome I ever tested (fabulous tone burst performance) - 9 per side.
Say... that makes impedance matching a snap! Not to mention dispersion control.
PA equipment would do the "round the house" job, as would a multi-room system.
Though PA gear would mean some loss of quality - one thing i like about open baffles is their ability to maintain good tonal balance in adjacent rooms. And I'd need that as well as the main "fidelity rig". Still, I do have some some 100 dB/ watt drivers around
A multi-room system would be more neighbour friendly - with the need for a pre & power amp and boxes in each room. If I had the time to build all that . .
Doubling the tweeters seems a simple option
Though PA gear would mean some loss of quality - one thing i like about open baffles is their ability to maintain good tonal balance in adjacent rooms. And I'd need that as well as the main "fidelity rig". Still, I do have some some 100 dB/ watt drivers around
A multi-room system would be more neighbour friendly - with the need for a pre & power amp and boxes in each room. If I had the time to build all that . .
Doubling the tweeters seems a simple option
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.