Setting delay between tops and subs...

[Brian Steele]

Just a quick one - what technique(s) do you use to set the appropriate time delay (if necessary) between tops and subs in a PA system?

[billfitzmaurice]

Just a quick one - what technique(s) do you use to set the appropriate time delay (if necessary) between tops and subs in a PA system?

By and large, none. What matters most is time align between midrange sources, so I align the mains to the backline amps and instruments. And since the subs tend to be wall loaded and placed closer to the plane of the backline than to that of the mains they end up aligned anyway. A difference of as little as 5ms between the backline and mains in the midrange can be heard easily, while a difference as great as 30ms between the mains and subs is almost impossible to hear.

[Brian Steele]

By and large, none. What matters most is time align between midrange sources, so I align the mains to the backline amps and instruments. And since the subs tend to be wall loaded and placed closer to the plane of the backline than to that of the mains they end up aligned anyway. A difference of as little as 5ms between the backline and mains in the midrange can be heard easily, while a difference as great as 30ms between the mains and subs is almost impossible to hear.

Thanks Bill. I heard mention of delaying the mains to match the subs (see link for example: http://www.diyaudio.com/forums/subwo...h-drivers.html), hence the question.

This past weekend I was again called upon to provide an impromptu small PA system using some borrowed CV PSX-153s and my POC TH, all driven by the amps in my car (having detachable speaker connecters on the amps made it easier than expected). The car's system includes time alignment capability, but measured in cm (distance from listening position) rather than in ms. I found that setting the sub to 130cm seemed to provide the "best" results, but this was done solely by ear however, not by any measurements. The PSX-153s were high-passed @ 80 Hz and the sub was low-passed @ 80 Hz as well.

I think I'll repeat the config this weekend and use an RTA to see if there's any noticeable and measurable difference when adding delay to the mains.

[billfitzmaurice]

I found that setting the sub to 130cm seemed to provide the "best" results, but this was done solely by ear however, not by any measurements. The PSX-153s were high-passed @ 80 Hz and the sub was low-passed @ 80 Hz as well..

130cm/4.3 feet is about 5ms, which is inaudible at 80Hz. But if the sub is direct radiating and pushed hard it could have significant harmonic content well above 80Hz, which could be audible at 5ms. Then there's the matter of phase response glitches in the overlap between the tops and subs output, which I very much doubt is well controlled with an auto sound crossover, and that might be improved with as little as 5ms delay.

[Brian Steele]

130cm/4.3 feet is about 5ms, which is inaudible at 80Hz. But if the sub is direct radiating and pushed hard it could have significant harmonic content well above 80Hz, which could be audible at 5ms. Then there's the matter of phase response glitches in the overlap between the tops and subs output, which I very much doubt is well controlled with an auto sound crossover, and that might be improved with as little as 5ms delay.

I wasn't playing at any significant level, so I don't think distortion contributed to what I was hearing. The filtering was done by a Premier 980BT, which was a top of the line deck when it was released a few years ago, so I don't think it's glitches in the phase response either. For all I know, the "improvement" I heard could be all in my head. The RTA will confirm what the actual situation is in a few days .

[billfitzmaurice]

The filtering was done by a Premier 980BT, which was a top of the line deck when it was released a few years ago, so I don't think it's glitches in the phase response either.

The specs show 3rd order LP on the sub, but only 2nd order HP on the mains, so there could be quite a lot of bandwidth overlap. Users of pro-sound crossovers with 4th order/4th order filtering notice quite an improvement over lower order filtering.

[Brian Steele]

The specs show 3rd order LP on the sub, but only 2nd order HP on the mains, so there could be quite a lot of bandwidth overlap. Users of pro-sound crossovers with 4th order/4th order filtering notice quite an improvement over lower order filtering.

The 980BT can do 6dB/octave or 12dB/octave on the mains and 6, 12 or 18dB/octave on the sub. The subwoofer amplifier also has a built-in undefeatable 24dB/octave lowpass filter that I have set at 200 Hz. I used the 980BT's 12dB/octave settings for both the mains and the subs, so perhaps there was even more overlap.

I'm looking forward to checking the results with an RTA to see if it can pick up what was really going on.