Rational Acoustics



Danny Dagher
October 29th, 2008, 12:02 PM
So here's the issue. I've got two Meyer UPA-1C's without the M1-A processor. And I'm trying to compensate for the lack with a Driverack PA. Is it worth it?

I'm not familiar with the response of a 1C but it seems to roll off fast after 8k and it does not have anything after 80. The high freq. driver does not seem very well behaved with a 6db difference between 2 and 3k, which makes me wonder if there is driver damage.

Either way I know Meyer has declared a 1.6k crossover point for the box and so I was wondering what the best way would be to phase align these drivers if possible.

Harry Brill Jr.
October 29th, 2008, 03:01 PM
So here's the issue. I've got two Meyer UPA-1C's without the M1-A processor. And I'm trying to compensate for the lack with a Driverack PA. Is it worth it?

I'm not familiar with the response of a 1C but it seems to roll off fast after 8k and it does not have anything after 80. The high freq. driver does not seem very well behaved with a 6db difference between 2 and 3k, which makes me wonder if there is driver damage.

Either way I know Meyer has declared a 1.6k crossover point for the box and so I was wondering what the best way would be to phase align these drivers if possible.

First you want to EQ your individual drivers on axis for the flattest response practical.
I think Meyer used an Impulse Alignment on these. Basically you put your mic at the geometric center between the drivers. Turn on the HF only. Set your delay in smaart. Adjust the smaart delay to get the flattest overall phase trace through the center of the passband. Lock it in! Do not touch the Smaart delay again. Do not move the mic. Capture the HF trace. Turn off the HF and turn on the LF. Adjust the delay on your driverack driving the LF until the phase trace of the LF is as FLAT as possible through the bandbass. Adjust the levels so most of the LF passband is equal to the level of the saved HF trace. Check the phase AT the acoustic crossover frequency. If they are touching, then go to the next step, if they are 180 degrees apart, then invert the polarity of the LF so that they are touching. The next step would be to turn the HF back on, look at the crossover, and EQ at the input, pre crossover to clean up any ripple near the crossover.

You want to make sure you aren't coloring your measurements with the room. If you don't have a large expensive anechoic chamber, you might want to do this outside away from any buildings. Put your speaker as high as you can on a speaker stand (pole), and keep the mic reasonably close. Wind is more of an issue than the background noise up to a point.

PaulTucci
October 29th, 2008, 09:55 PM
Harry,

I believe Meyer to be all about constructive-addition at the crossover point.

I'm with you until this point. Instead of getting the phase trace of the low frequency driver as flat as possible in the middle of its passband, purposefully misalign the time so that at the crossover area, the two phase traces are close and parrallel, 180 degrees out and parallell, or atop one another. When they combine, they will combine the most constructively. As you said polarity reversal may be necessary.

The crossover filters will skew the phase response more at the upper end of the LF driver than in its midpoint of the passband. Get the two drivers to play nice where they are of equal amplitude and of the same signal (acoustic crossover area) and one will have achieved the poster's goal.
Tucc

Jamie
October 30th, 2008, 01:28 PM
Danny,

A couple of points about the UPA-1c and its pal the M1-A CEU (Control Electronics Unit).

First, regarding the crossover in the M1. The Meyer CEU's were all purely analog devices. This means that they did not employ signal delays for driver alignment and generally utilized lower-order crossover filters (with coorespondingly lower amounts of phase shift). When necessary, Meyer did employ analog all-pass filters to address timing/phase issues in their crossovers.

The impact of these restrictions/choices was a flatter over-all phase response with a low driver that effectively led its high driver in the crossover region. Why??? Well, if you assume a symetrical crossover, both the high and low drivers would recieve the same amount of phase shift at x-over = same amount of signal delay. Therefore, the only time arrival difference would (mainly) be due to the physical position offset of the drivers - and the low driver is nearer (front-loaded) than the high driver (horn-loaded).

Effectively, the UPA's alignment with the M1 resembles the "impulse" type alignment we have discussed in previous threads.

So, if your intent is to recreate the original M1 alignment for your UPA, best choice would be to find an M1-A and measure/copy its HF and LF out responses. If that is not an option, remember to utilize low order HP/LP filters and make the crossover work without using delays.

A second point to make about the Meyer CEU's is that they employ a pretty effective driver protection scheme (Speaker Sense) that monitors the amplifier outputs directly, and limits based on the true drive voltage for the speaker. A bit cumbersome to hook up, but it certainly does its job.

-j

FILO4PRES
October 31st, 2008, 01:07 PM
http://www.youtube.com/watch?v=ViE_0hq4wGc

This isn't perfect, but I think it elaborates what Harry is talking about.