hi,
time correction and units alignment;
i am trying to set up a 3 way spk and i have to decide how to place the units .
how do i calculate the distance between the units (3d view) in order to get a time delay correction; i use 3 way elec. xover and 3 way amplification. I know that tweeter has to be place back wards some how, but how much, if I incline the spk box back wards this will do it?
best regards
Williams
time correction and units alignment;
i am trying to set up a 3 way spk and i have to decide how to place the units .
how do i calculate the distance between the units (3d view) in order to get a time delay correction; i use 3 way elec. xover and 3 way amplification. I know that tweeter has to be place back wards some how, but how much, if I incline the spk box back wards this will do it?
best regards
Williams
Hi,
I had a long email discussion with Ashley James, MD of AVI HiFi on this subject. AVI are a very well respected manufacture of amps, cd players and speakers. (www.avihifi.co.uk)
Ashley maintains that there is no need to compensate for acoustic delay between drive units, but it is vital to ensure the electrical signals are in phase.
He said they carried out a test, where the tweeter and bass/mid units were in different enclosures, one on top of the other. They could move the tweeter boxes by +/-1m forward and backwards, before there was any noticable difference !!
I had a long email discussion with Ashley James, MD of AVI HiFi on this subject. AVI are a very well respected manufacture of amps, cd players and speakers. (www.avihifi.co.uk)
Ashley maintains that there is no need to compensate for acoustic delay between drive units, but it is vital to ensure the electrical signals are in phase.
He said they carried out a test, where the tweeter and bass/mid units were in different enclosures, one on top of the other. They could move the tweeter boxes by +/-1m forward and backwards, before there was any noticable difference !!
cs said:
I would really like to know what drivers and type of XO they used during the test, and the exact components and ratings used. XO frequency is also important.
Measurements in the lab allowed me in identify exactly how much I needed to move a super tweeter to get the response I was looking for. Although initial listening did not show much difference with an XO at 10KHz, if the XO was at a lower point where normal 2-way speakers cross, the difference would be significant.
Member
Joined 2004
Hi there
I tried to time align my tweeters electrically with an alpass network some years ago. To align them with my mid. I could not hear an difference. (1.2cm back in time...)
If you time align a tweeter with 2 inches, how do the on and off axis measurements ever are going to add up?
Best Regards
Roland
I tried to time align my tweeters electrically with an alpass network some years ago. To align them with my mid. I could not hear an difference. (1.2cm back in time...)
If you time align a tweeter with 2 inches, how do the on and off axis measurements ever are going to add up?
Best Regards
Roland
phase_accurate said:
A pretty safe assumption, I would have thought.
Unless you are in the habit of listening to your system, either with your ear pressed against the speakers, or in an anechoic chamber !
I have tried delay compensation with active speakers in the past, and agree with AVI - it makes absolutely no difference !
Your room must be very reflective in the mid and high ranges. Normal living areas have lots more absorbing stuff so that most difference you will hear is in the bass. For timne lignment it is mor necessary to listen for changes in the percussion instruments with mid and high frequency content.
Time alignment only makes a remarkable difference in systems suffering from time/distance mismatches longer than a quarter of the period/wavelength of the crossover frequency.
For a two way system crossed over at 2Khz, those "worth-the-effort" time and distance figures are 4.3cm and 125us respectively.
So time correction may not be woth the effort at all for small hi-fi speakers, but it's absolutely mandatory for proper summing when large drivers/enclosures or horns are involved (it's a must for any good sounding PA system, like modern line arrays).
Note that in conventional small/medium sized well-engineered hi-fi speakers, the small time, distance and phase mismatches introduced by the placement of each driver and it's own phase response are alreeady compensated in the crossover (by moving poles and zeros up and down as required), so there is no room for further correction.
Also, the reverberation argument is false because a system where full-spectrum on-axis radiation dominates in the listening position sounds completely different from a system where on axis radiation dominates at some frequencies and reverberated junk dominates at others (because on-axis stuff is cancelling itself due to improper alignment).
The difference is easily demonstrated by toggling tweeter polarity in a properly aligned system. When the wrong polarity is chosen, reverberant sound (thrash!!) will dominate in the crossover region (subjectively it may be perceived as a big loss of definition).
For a two way system crossed over at 2Khz, those "worth-the-effort" time and distance figures are 4.3cm and 125us respectively.
So time correction may not be woth the effort at all for small hi-fi speakers, but it's absolutely mandatory for proper summing when large drivers/enclosures or horns are involved (it's a must for any good sounding PA system, like modern line arrays).
Note that in conventional small/medium sized well-engineered hi-fi speakers, the small time, distance and phase mismatches introduced by the placement of each driver and it's own phase response are alreeady compensated in the crossover (by moving poles and zeros up and down as required), so there is no room for further correction.
Also, the reverberation argument is false because a system where full-spectrum on-axis radiation dominates in the listening position sounds completely different from a system where on axis radiation dominates at some frequencies and reverberated junk dominates at others (because on-axis stuff is cancelling itself due to improper alignment).
The difference is easily demonstrated by toggling tweeter polarity in a properly aligned system. When the wrong polarity is chosen, reverberant sound (thrash!!) will dominate in the crossover region (subjectively it may be perceived as a big loss of definition).
Eva said:
That's because you are effectively inverting the electrical phase between the two drive units. The so-called 'time alignment' is varying the group delay difference, which is irrelevant.
AVI say that you must ensure accurate electrical phase matching. They actually use 4th order Linkwitz-Riley, where the phase difference between highpass and lowpass is 360 deg at all freqs.
If acoustic (group delay) alignment was important, how do you explain their test of moving the tweeter forwards or backwards 100cm, without causing any noticable change ?
Personally, I think their design approach is exemplary. Take a look at their website - they give a great deal of information - far more than any other loudspeaker manufacturer I can think of !
My experience agrees with their findings.
Electrical phase matching is important, but group delay matching is just a red herring !
My experience agrees with their findings.
Electrical phase matching is important, but group delay matching is just a red herring !
cs said:
When it's misaligned, it does not matter how "much" misaligned it is, it sums (and sounds) equally badly whether you move the tweeter 10cm, 30cm or 50cm (imagine that the correct offset is, say, 6cm). It's all a matter of proper on-axis summing (and also proper summing of the transients coming from both drivers).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.