Measuring multi-driver systems (MTM, MMTMM, whatever)

[Dirk]

When measuring a multi-driver speaker like an MTM, do you take measurements of both woofers simultaneously? Or do you unplug one woofer and measure the other? This is before designing a crossover, of course.

[Paul Carmody]

When measuring a multi-driver speaker like an MTM, do you take measurements of both woofers simultaneously? Or do you unplug one woofer and measure the other? This is before designing a crossover, of course.

Personally, I measure both woofers hooked up, keeping the mic at the same listening axis I used for the tweeter. The only "catch" is that you cannot tell the simulation software that you're running an MTM... just treat the measurement as if it were of single woofer.

Yes, there are more "technical" ways to measure it, but I like it this way because I can trust the "time of flight" for the acoustic phase to be correct for all drivers (rather than using minimum phase and specifying X, Y and Z offsets)

[Dirk]

Thanks Paul.

Do you then measure at 1m, 2m, as back as first reflections allow? I'd have to think the summation would "look" different to the mic based on distance.

[dthomas]

Thanks Paul.

Do you then measure at 1m, 2m, as back as first reflections allow? I'd have to think the summation would "look" different to the mic based on distance.

Personally I prefer to measure a single woofer at 1m then allow the software to calculate the differences in summed output at varying listening distances. For the reason you pointed out. The summation will be different at different listening distances and if the software is allowed to it should give you a clearer picture of what will happen at your chosen listening distance. If you measure them both the software will assume that the woofer is a single point source and won't account for the changing angle at varying distances.

[Paul Carmody]

Do you then measure at 1m, 2m, as back as first reflections allow? I'd have to think the summation would "look" different to the mic based on distance.

yes

Seriously, though, I have no "set" distance from where I measure. Instead, I take into consideration the application of the speaker. If it's for a small room (or nearfield listening), then 1m works well (I have sometimes, but not always, had good luck with .5m measurements as well). For speakers in large rooms, I usually start at 1m, work on the XO, then do some sanity checks at 2m or more doing system FR measurements--and sometimes even new driver measurements at the farther distance to see if it still "looks right" in the simulation.

Also, I think a lot about listening axis when I measure. However, if you're doing an MTM, then you'd better be on-axis with the tweeter (vertically, at least)

[djarchow]

The answer is that it depends. If the woofers have seperate enclosures you can measure just one.

If the woofers share a common enclosure you need to measure both at the same time, wired as they will be in the final configuration. Otherwise if you just wire and measure the one woofer, the other will work like a passive radiator and mess up your measurements.

If you measure this way, you can simply scale your measurement SPL down by the appropriate amount for one driver, and increase the impedance by the appropriate amount. So for a pair of identical drivers in parallel reduce the mesured SPL by 6db at all frequencies and double the impedance values at all frequencies. Then you can use these values to model your speaker as an MTM instead of an MT.

Regards,

Dennis