In Car Measurements For Passive Design

[dawaro]

Ok I have asked this once before but got everything but an answer so I am going to try one more time. I am hoping some of the guys like Augerpro, Jed, and Thomas will chime in.
I have a friend that is building a show car and he wants to do a 3-way component set in the front doors. We have looked at retail sets anywhere from $500-$2500 for the set. I find it hard to believe that we cannot exceede the quality of either the sound or the drivers for far less.
Before all the recomendations start, going active is not an option, that is why I am asking the question. I have 20yrs of car audio experience and I know all the active options.
My question is would it be best to take the measurements with the mic in the center of the car at ear level and how would the gating have to be adjusted due to the space constraints? I am not sure if I will be doing the measurements with Arta or JustLMS at this point but it will be one of the two unless I can find a way to make my Linear X PCRTA work.

[Winter]

Accurate direct fequency response measurements, justMLS and others, in a car interior are extremely difficult due to early reflections from the numerous close surfaces. For initial crossover design I usually mount the drivers on a large flat baffle, say 3 feet by 4 ft, in a driver spacing similar to the planned car interior layout. Measures are taken to absorb most of the rear wave behind the panel. The crossover is then designed, with less, if any, baffle step compensation than for a normal house interior placement.

The drivers are then mounted in the car and pink noise (pink noise test CD) is played through the crossover to the drivers. Measurements are taken with a real time analyzer, with the drivers both in and out of relative phase for best result. The mic is placed in the listening position among other places. Approximante tweeter attenuation and best relative phase are determined with the pink noise measurements.

I really don't see a need for more than a two way system in a car, but I guess it is a show car.

[augerpro]

That's tough to say David. Experimenting would be your best bet, although at the very least I would probably put the mic in middle at head height. Short gates would be pretty much pointless, so I would say go with 500m or ungated and use 1/3 octave smoothing. Just do the best you can on the XO model, no need for perfection like you would see in projects here. Considering how compromised this situation is just try to make the best of it-the results can't be any worse than a stock active XO used with no measurement.

Winter's suggestions about using several mic positions and using pink noise to find the best compromise are right on.

[John_E_Janowitz]

The first thing I do is try to pick driver positions with the least obstructions to the listener. After I do that and get drivers mounted I'll place the mic at the listening position. For most cases this is always at the head position of the driver, as that is the primary listener. I actually use either a dBX Driverack 260 or a Behringer DCX2496 for processing as I haven't found anything that works as well in the car audio realm. I take an impulse response of the farthest driver from the listening position, usually the subwoofer. Then I take an impulse response of all the other drivers, one at a time and adjust the time alignment until all drivers show an impulse at the same point. From then on you can continue with your EQ, crossover, etc. I like to get rid of any large response issue prior to doing the crossover. As stated, a lot of this is hard as you get extra reflections in a vehicle. You can time gate to get rid of the longer reflections, but it's difficult to remove very close reflections.

If doing a passive Xover, you don't get the benefit of the time alignment, parametric EQ, etc so you largely have to take what you get. For $260 or so, the Behringer is a good option and there are sites where you can get plans to make a 12V power supply for it and avoid the power inverter. Doing a multi-way active system in a vehicle is quite cheap as there are good 4 and 6 channel amps that can be had quite inexpensively. I used a 5 channel amp to do 2way actives in the door and a single subwoofer originally with great results.

John

[dawaro]

Accurate direct fequency response measurements, justMLS and others, in a car interior are extremely difficult due to early reflections from the numerous close surfaces. For initial crossover design I usually mount the drivers on a large flat baffle, say 3 feet by 4 ft, in a driver spacing similar to the planned car interior layout. Measures are taken to absorb most of the rear wave behind the panel. The crossover is then designed, with less, if any, baffle step compensation than for a normal house interior placement.

The drivers are then mounted in the car and pink noise (pink noise test CD) is played through the crossover to the drivers. Measurements are taken with a real time analyzer, with the drivers both in and out of relative phase for best result. The mic is placed in the listening position among other places. Approximante tweeter attenuation and best relative phase are determined with the pink noise measurements.

I really don't see a need for more than a two way system in a car, but I guess it is a show car.

Yes it is mostly going to be just for show. With 850hp under the hood and gas at its current rate it wont be driven much :E

The main reason for the 3-way is the owners love of midbass. The front stage of the car will probably be crossed at 60hz. From what you describe it sounds like I could almost use the mfg measurements from the IEC baffle and run it through one of the sim programs to simulate the off axis position. I may give that a shot before I start investing in more measurement equipment.

Man the troubles we go through for our buddies...

[Preacher]

There are plenty of DIY setups with great midbass in cars. I've run seas RNX in car doors, seas re 8's as a two way. Both have nice midbass. Another option is to use a two way whether custom or stock passive and use the xover on the amp to run a midbass for a hybrid threeway.

[dawaro]

right now what we have are the new JL Audio HD amplifiers. One mono for the subs and one 4 channel for the front. It is possible that we may go active between the midbass and passive between the mid & tweet. I just have a hard time believing we cannot exceed the quality of xover that is built as a generic fit everything application. I have seen quite a few component set xovers and most leave a lot to be desired.
I am almost tempted to take something like Zaph's new design and try it in the doors just to see how it sounds. Normaly I would use compression drivers and horns under the dash (since I know the guys at Image Dynamics) with an 8" midbass in the kick panel but he wants to avoid any interference with the clutch so we are stuck to the doors.

[Hdale85]

Well you could use any drivers and just go full active. Thats how must of the DIY car audio guys do it if not using a name brand component set.