first time build HM170C0-810921 MTM need help

[natfin]

Hello all,

This is my first post. I've been reading and researching for awhile and hope someone will be interested in this.

I have four Audax HM170C0 6.5" drivers that I'd like to turn into a vented MTM with the peerless HDS 810921 Tweeter. I am thinking around 2K for the crossover, I was thinking a 4th order L-R filter for both the low and high pass in parallel.

The boxes are built. I oversized a little for adjustment later. Internal volume is about 30L. The baffle is 9 1/4" W x 21 7/8" H. Tweeter is located 11 1/2" from bottom of baffle and centered between the edges. The ports are 2 5"x2" pvc per box. These will have to be adjusted as well.

I originally had Madisound do a LEAP design but that turned out very bad. It had a serious dip at 2.8K, the crossover section. Also the tweeter was about 8db down from the level of the wooofers. Plus they sounded terrible. I don't have my measurements from that assembly otherwise I would post them.

I've posted some FR and IMP measurements of the drivers on the baffle.

One of the caveats for me in this process is that I am using a mac. I have virtual PC running Windows 2000 and am able to use software like Speaker Workshop and Spl tools and what not but I haven't been able to make an accurate trace of the impedance of the drivers using SPL Trace. The ohm scale is always not accurate no matter which scale I set in the Tracing options and hence I am unable to produce usefull .zma files. I am able to export .frd for the drviers from Fuzzmeasure, but it does not have an ablility to export zma files. I guess I am asking for help in where to go from here. I would love to be able to model the crossover myself but I can't get the correct .zma files that would be needed in a program like Speaker Workshop. I could go ahead with pencil and paper and a couple of years but it would be nice to some optimization before I start buying parts.

If anyone has some suggestions I would greatly appreciate it. Please pardon such a long post. I have been a reader of the forum for greater that a year now and I'm ready to go out on a limb and start designing for myself-diy style.

Nathan

[natfin]

TS and internal box dimensions so far

Hello again,

I just wanted to post some TS and box dimensions to go along with my original post.

HM170C0 impedance mounted in baffle
TS

6.9 DC Resistance
Fs= 47.1
Qms= 7.832
Qes= 0.334
Qts= 0.320
Res= 168.771
Le= 0.013
n= 0.686

VAS= give or take= 26.7

Box Internal Dimensions 30.06 L 1834.175 cu in
Height: 20 3/4" Width: 7 3/4" Depth: 11 1/2"
this is as close as I could get to golden ratio for box dimensions.

I would post the sketch up file of the layout but I lost those diagrams with a hard drive crash over the summer
-Nathan\

[ThomasW]

Unfortunately there are some situations where one needs the ability to run native Windows XP and the programs associated with it.

If you can post a schematic of the Madisound crossover. Might be a good idea to see if it can be reworked.

What were the conditions of the measurements? Distance of mic to speaker?, distance of speaker to boundries, gating?, etc.

[natfin]

Hi there,

The conditions are not ideal. Measuring in my room / listening space.
Mic ECM 8000. 32" inches from center of the driver that is being measured on the baffle. 6ms window from the beginning of the impulse viewed in Fuzzmeasure.

[natfin]

Oh, the baffles are hanging from a ceiling fan in the middle of the room. There nearest reflection would have to be from that which is about 3 1/2' from the top of the suspended baffle. The nearest wall is about 7' away.
I could take them outside but that is a bit difficult, plus the city is not always so quiet.

Nathan

[cjd]

I'm pretty sure Fuzzmeasure is capable of doing the required measurements - At any rate, I thought Jon had used that for some work. Perhaps not, and my memory is incorrect.

C

[natfin]

Fuzzmeasure seems to working for me, but I guess I'm a little confused about how to calculate values for the MTM set up. Do I need to divide by 2 whatever results I get from an equation-ie. an impedance compensation. For example:
C=Le/Rc^2 then divide by two? Or should I just wire the drivers in parallel and make some impedance measurements as if they were one driver?

[cjd]

Yeah, wire 'em up, measure in-baffle. If you set everything up and measure both the tweeter and the mid-woofers without moving a thing (at 1-2M) you have time delay and the rest pretty much all rolled up in your measurements.

Calculating values would be based on the net impedance/inductance. Crossover values would be similarly based on the impedance at (and around) the crossover point. Which isn't flat (though a zobel sure helps that).

[natfin]

That's what I was hoping. That I could measure in baffle and go from those measurements based on the logic that this is what the circuit sees. I've put a zobel on the mid-woofs measured imp and seems that so far it flattens the rising impedance. 20uf-4.7ohm. It's a start anyway. I'm going to try a 4th LR at 2k- see what happens.