How accurately do I need to build the volume of my bass cabinet?
If I am aiming for 29 litres, say, what is the level of give and take that is allowable?
I know that I need to deduct the volume of the bass driver, and the reflex port.
This brings me to my next question: How do I measure the right length of reflex tube? If they were just tubes, this would be simple, but most have flared ends. Do I ignore the flares and just measure the parallel sided dimension? Again, what accuracy is needed?
If I am 10% out in cab volume or tube length for example, what would be the affect to the tuning?
I will be using a ScanSpeak 22W8851 driver. Strassacker recommends a 29 litre cabinet with a HP70 reflex tube 28cm long. This would achieve 40/30Hz (-3dB/-8dB).
If I am aiming for 29 litres, say, what is the level of give and take that is allowable?
I know that I need to deduct the volume of the bass driver, and the reflex port.
This brings me to my next question: How do I measure the right length of reflex tube? If they were just tubes, this would be simple, but most have flared ends. Do I ignore the flares and just measure the parallel sided dimension? Again, what accuracy is needed?
If I am 10% out in cab volume or tube length for example, what would be the affect to the tuning?
I will be using a ScanSpeak 22W8851 driver. Strassacker recommends a 29 litre cabinet with a HP70 reflex tube 28cm long. This would achieve 40/30Hz (-3dB/-8dB).
Try it? Have you got WinISD? Try the different volumes and lengths and see how they affect performance. I would also say that the room conditions will change the response more than the odd centimeter on the tube length.
My first reflex (ported box) did not perform all that well. I built it exactly to the design manual that came with the driver.
Since then I have always "tuned" the reflex using adjustable length ports.
I have come to believe that the "ideal" response that is predicted by the graphs (whether by software or by manual calculation) is not exactly achievable with a preset fixed length port.
In addition the amplifier and the cables and the room interact with the response and this requires further "tuning" to get a sensible response that gives a realistic output arriving at your ears.
Since then I have always "tuned" the reflex using adjustable length ports.
I have come to believe that the "ideal" response that is predicted by the graphs (whether by software or by manual calculation) is not exactly achievable with a preset fixed length port.
In addition the amplifier and the cables and the room interact with the response and this requires further "tuning" to get a sensible response that gives a realistic output arriving at your ears.
10% is fine , no audible differences , but measurable for sure . room temperature , barometric pressure and humidity can change your driver's T/S parameters up and down for at least the same percentage .
Cheers ,
Rens
Cheers ,
Rens
To be accurate, you should measure the impedance and see and listing !
The tuning depends on what is in the box ...
The tuning depends on what is in the box ...
To be accurate, you should measure the impedance and see and listing !
The tuning depends on what is in the box ...
This is what I did.
I used an AC ammeter (Fluke multimeter), and used a frequency generator to find where current was maximum. At that point, impedance is minimum, which is indicative of port tuning.
But that tells you or the room nothing about what Qbox is. Is Qbox ~0.6, or 0.7, or 0.8, or 0.9, or 1.0, or good God, I hope you don't aim for Qbox>1.0
I hate high Q boxes. They never sound "right" on a wide range of music and voice.
I prefer lower Q and fairly recently discovered that low Q boxes can also sound quite accurate.
I hate high Q boxes. They never sound "right" on a wide range of music and voice.
I prefer lower Q and fairly recently discovered that low Q boxes can also sound quite accurate.
So how do you find the 'Q' of a box?But that tells you or the room nothing about what Qbox is. Is Qbox ~0.6, or 0.7, or 0.8, or 0.9, or 1.0, or good God, I hope you don't aim for Qbox>1.0
I hate high Q boxes. They never sound "right" on a wide range of music and voice.
I prefer lower Q and fairly recently discovered that low Q boxes can also sound quite accurate.
A bit like testing for Q when doing the T/S parameter measurement.
The spread of the voltage around the peak gives the height to width ratio of the peak. That is roughly the Q of the port output. It's similar for a sealed box. No idea for a TH.
The spread of the voltage around the peak gives the height to width ratio of the peak. That is roughly the Q of the port output. It's similar for a sealed box. No idea for a TH.
Better than 20% you are kidding to yourself. Voice coil temp, surround variation, humidity, BP and so on. Of course, we try for better. We try for 5%. BTW, I too like low Q. .6 seems about right. WT-2 is an easy measurement tool.
Not to be too pedantic but box Q is meant to mean the Q related to box losses (along with vent Q). Q loss is usually down to leakage and also stuffing losses. Small did his paper with an assumption of a box Q of 7 (not 0.7).
You are referring to a system Q n the region of 1 or less. Unfortunately that isn't correct either as we would be considering a 4th order system, so the Qs that would apply to a 2nd order system (Butterworth = Q of 0.7) don't apply either. A 4th order Butterworth has 2 Qs, one of 0.541 and one of 1.31.
We discussed this previously, but I don't think you grasped it then.
http://www.diyaudio.com/forums/multi-way/176992-4th-order-butterworth-filter-q-2.html#post2360525
David S.
You are referring to a system Q n the region of 1 or less. Unfortunately that isn't correct either as we would be considering a 4th order system, so the Qs that would apply to a 2nd order system (Butterworth = Q of 0.7) don't apply either. A 4th order Butterworth has 2 Qs, one of 0.541 and one of 1.31.
We discussed this previously, but I don't think you grasped it then.
http://www.diyaudio.com/forums/multi-way/176992-4th-order-butterworth-filter-q-2.html#post2360525
David S.
Hi,
The first aspect of a vented box is the equivalent
sealed box Q, which varies with the box volume.
The second aspect is tuning of the port frequency.
In electrical circuits its known as "sensitivity",
how critical are values to cause a given change.
Lower Q boxes with low vent tunings for overdamped
alignments are relatively "insensitive" to minor variations.
Higher Q boxes with higher tunings are more "sensitive",
a given variation will have a lot more effect than the above.
In electronics you can design low sensitivity circuits, and
high ones, for the latter you need better specified parts.
There is no simple answer to the question, sensitivity
varies with the nature of the chosen solution, EOS.
rgds, sreten.
The first aspect of a vented box is the equivalent
sealed box Q, which varies with the box volume.
The second aspect is tuning of the port frequency.
In electrical circuits its known as "sensitivity",
how critical are values to cause a given change.
Lower Q boxes with low vent tunings for overdamped
alignments are relatively "insensitive" to minor variations.
Higher Q boxes with higher tunings are more "sensitive",
a given variation will have a lot more effect than the above.
In electronics you can design low sensitivity circuits, and
high ones, for the latter you need better specified parts.
There is no simple answer to the question, sensitivity
varies with the nature of the chosen solution, EOS.
rgds, sreten.
My first reflex (ported box) did not perform all that well. I built it exactly to the design manual that came with the driver.
Since then I have always "tuned" the reflex using adjustable length ports.
I have come to believe that the "ideal" response that is predicted by the graphs (whether by software or by manual calculation) is not exactly achievable with a preset fixed length port.
In addition the amplifier and the cables and the room interact with the response and this requires further "tuning" to get a sensible response that gives a realistic output arriving at your ears.
so in a vented cabinet tuned to say 25hz, does that tuning frequency atually change depending on where the speaker is in a room or what room it is in? or is the tuning frequency 100% denoted by the box itself, and the room just decides what bits of the range get boasted? (by room nodes etc)
download Unibox & play with variations in box size, tuning, & filling - you can overlay the graphs to see how much difference will result, this will give a feel for it. It won't tell you how audible this will be though... I doubt whether 10% variation will make much difference in most scenarios.
Unless you can find a driver with horizontal TS curves, there is no such thing as accurate tuning of a BR.
Further driver to driver variation on really good drivers typically varies +/- 10% making life even harder.
To me it makes sense to do vented systems that have a high tolerance to dynamically changing parameters.
dave
Further driver to driver variation on really good drivers typically varies +/- 10% making life even harder.
To me it makes sense to do vented systems that have a high tolerance to dynamically changing parameters.
dave
Dublin 78,if you wish to obtain the maximum performance in the bass region with the 'Revelator' then from the published T&S parameters an internal working volume of 120L is required for a matched vented enclosure,with Fb=20Hz.It is handy to be able to confirm the parameters after running in the speakers.The equivalent sealed box, as Streten infers, can be the cause of bass hangover if you use a compromised compact design.Although many designers would think that a Qts of 0.25 is marginally low for a classic B4 alignment,in practise it works out in your favour,as the effects of mutual coupling(for the usual pair of speakers) and room gain are not part of the normal calculation.
so in a vented cabinet tuned to say 25hz, does that tuning frequency atually change depending on where the speaker is in a room or what room it is in? or is the tuning frequency 100% denoted by the box itself, and the room just decides what bits of the range get boasted? (by room nodes etc)
Its very much the later. The speaker has pretty high mechanical impedance and is barely aware of the room and its modes. The tuning isn't changed by the room but the room has its response impressed on top of the speaker.
David
the link to a previous discussion is useful in showing that at that time, there were many opposing views from various Members on what the Q meant, or on what the result from a combination of different Qs did to the overall result. I was certainly not convinced by your argument that I misunderstood the subject. I have not gone back over that previous discussion to discover if any of my beliefs have changed since then........................We discussed this previously, but I don't think you grasped it then.
http://www.diyaudio.com/forums/multi-way/176992-4th-order-butterworth-filter-q-2.html#post2360525.........
I think we are all agreed that the conditions around the port do affect the tuning.......... The speaker .............. is barely aware of the room ............... The tuning isn't changed by the room but the room has its response impressed on top of the speaker...........
Flat plates on the ends of the ports are different from ports that are open ended and different again from flared ends on those ports. Similarly drivers and boundaries that are close to the internal end of the port/s affect tuning.
It is not a long stretch from that agreed set of design conditions to assume that similar tuning variations will affect the external end of a port. Am I wrong in that assumption?
Let's keep going. Placing the external end of the port near the floor, or near the back wall will affect the tuning of the port. Again I ask, is this true?
And extending the argument further, these box locations within the room do affect the tuning of the box.
I will finish by saying that I can hear changes in the sound output as the box is moved around the room and in particular when near boundaries. Some of this change must be due to reflections affecting the "spacial loading" on the speaker, But I suspect that some is due to tuning changes.
For these reasons, I believe the tuning should remain "not fixed" to allow adjustment until a satisfactory sound has been achieved. One could then choose to "fix" the tuning to suit that particular set up.
I was certainly not convinced by your argument that I misunderstood the subject.
That was very clear...both then and now.
- Status
- Not open for further replies.
- Home
- Loudspeakers
- Multi-Way
- How accurate does bass reflex tuning need to be?