Line Array Element testing plot... smaart?

T

Thomas Petrusky

Freshman
Mar 15, 2012
64
0
6
Here is the dilemma, I have a pile of line array elements and I know there are some issues, as one side just didn't sound as bright as the other side. I flopped L/R drive cables from amp racks "unchanged". Settings on the boxes were correct. So I numbered the boxes as they came down to investigate.

Opened a few I was sure should have blown out hi drivers, everything tested out ok both sweeping with the generator and Multi Meter.

Here is my next goal, hook up a box after box in the same location and sweep or pink noise it under full power on the bench with an RTA mic remaining in the same location and somehow record the amplitude along the freq. Then take the results with them plotted out and know which boxes have drivers not performing properly...

Ok I have a copy of smaart 7 but I am not sure it is the ideal software for this project. Ideas?
 
Smaart will work fairly well if you want to see if there are response deviations. Just play pink noise through them until the average stabilizes, save the trace, and move on to the next cabinet. I'd play them ata reasonable level, not clipping or anything.

It is a well known phenomenon that drivers change over time. Depending on the drivers, it may be time to replace a few of them, if not all of them as a batch.
 
Riley Casey said:
Boxes that test well in the shop but don't sound right when arrayed on a show would have me checking for damaged or miss-wired cables. A speaker jumper with pins swapped or partially shorted would have peculiar effects.
Click to expand...

Yep.

The problem does not appear to be with the speaker drivers or crossovers (but could be on the loop-thru connectors). I'd conduct a quick continutiy check of the loop-thru connectors on the boxes and then move on to the other components of the system (e.g. improper configuration of your system processor or amplifier racks or 2-wire NL4 mixed in with the 4-wire variety (for biamped boxes)).
 
Rob Timmerman said:
Yep.

The problem does not appear to be with the speaker drivers or crossovers (but could be on the loop-thru connectors). I'd conduct a quick continutiy check of the loop-thru connectors on the boxes and then move on to the other components of the system (e.g. improper configuration of your system processor or amplifier racks or 2-wire NL4 mixed in with the 4-wire variety (for biamped boxes)).
Click to expand...

You can also start grouping them together for measurements. ie measure 1-2, 3-4, 5-6...if they are wired in pairs. (1-2, then 1-2-3, and 4-5 then 4-5-6 if they are wired three per) The response should be very predictable, and consistent between all the elements.
 
Thomas Petrusky said:
Here is the dilemma, I have a pile of line array elements and I know there are some issues, as one side just didn't sound as bright as the other side. I flopped L/R drive cables from amp racks "unchanged". Settings on the boxes were correct. So I numbered the boxes as they came down to investigate.

Opened a few I was sure should have blown out hi drivers, everything tested out ok both sweeping with the generator and Multi Meter.

Here is my next goal, hook up a box after box in the same location and sweep or pink noise it under full power on the bench with an RTA mic remaining in the same location and somehow record the amplitude along the freq. Then take the results with them plotted out and know which boxes have drivers not performing properly...

Ok I have a copy of smaart 7 but I am not sure it is the ideal software for this project. Ideas?
Click to expand...

Any measurement is void if you do not or can not get the same results with a repeat of the measurement.

Measure a few boxes then go back to the first unit tested. If the trace is different.... it's pointles.

You need a testing protocol and technique that results in 100% repeatability. Much of that is environmental.
 
Tim McCulloch said:
Any measurement is void if you do not or can not get the same results with a repeat of the measurement.

Measure a few boxes then go back to the first unit tested. If the trace is different.... it's pointless.

You need a testing protocol and technique that results in 100% repeatability. Much of that is environmental.
Click to expand...

Yes, but it's also important to remember that all measurement equipment has some amount of imprecision (measurement tolerance) that will result in minor differences in readings even without changes in the equipment or measurement setup.

Likewise, all parts have some tolerance, so some variation between instances of a design (e.g. between different array boxes of the same make and model) should also be expected.
 
No disagreement, Rob, but with modern equipment the same box or pass band component should measure within 0.5dB of a previous measurement taken only a hour or so before.

My point is that the failure is in the making of the measurements and often for nonintuitive reasons, including *any* changes in the measurement environment. This can be as simple as having another object in a different place during subsequent remeasurement. The physical testing environment must be 100% consistent, every time.

That's hard enough when you have lots of space and for most of us (we have 6000 sf of working space) it's not possible to ensure a 100% consistent environment.
 
The good news here is that this entire back and forth is irrelevant to the OP's situation, because all of the boxes traces were within a dB or two of each other, which for an imprecise measurement in the near field is more than enough to rule out driver failures. In addition, all of the sharp dips in the traces are at the same points, which means that any reflections were about the same, from either the cabinet or the environment. Point is - sound comes out, and it's plenty close enough to test for his application.
 
Top Bottom