Leach Memorial Speakers - Project Start!

[eriksquires]

Hi Everyone,

In honor of the person who gave me most of what I know in this hobby, I'm starting the Professor William Marshal Leach memorial speakers on my blog. I'm calling them the LM-1s. Hopefully I'm not copying a famous speaker name from the 70's or something.

Many are familiar with the small 2-way speakers that came out of the GA Tech lab but not actually designed by Dr. Leach many years ago. These speakers are roughly meant to be a modern take. The ideas of staying within a grad student's budget, simple, and small are key.

I'll be posting my process and all of my measurements as I go. The goal in addition for me to learn is for me to show others how I went about doing something like this in the hopes they find it fun too. There may be existing designs, but the point is not to copy a design, but to go through it step by step.

I hope you'll feel free to comment on my mistakes as I go along.

Best,


Erik

 

[system7]

TBH, Erik, I always read Marshall Leach with interest. He belonged to the old-fashioned school which said that even the lousiest driver could be equalised into shape.

I'm not sure it's true. The Vifa P17WJ is occasionally still available. But I don't much like 6" bass.

Once upon a time Vifa produced a driver (The P13WH) which was flat and musical. Lynn Olson grew a reputation on employing it in an MTM, aka The Ariel.

I give you the FR below. But really, if I lived in the States, I'd use the Aurum Cantus 5 1/4".

We just have to find a tweeter (or two) to match.

 

[eriksquires]

Thanks for the suggestions, Steve, but I'm sitting pretty on 4 woofers and 3 tweeters out of which my new home theater will be born. At least the new parts. We'll see how well I can integreate thes parts with my existing AMT based monitors.

Best,


Erik

 

[eriksquires]

I've finished the first measurements, really just to get the driver distances, but I went ahead and used the 3' measurements for a crossover. The Vifa and Peerless drivers seem pretty sweet so far, flat and with a lot of overlap.

See what I have so far here:

LM-1 First Measurements

 

[ra7]

Erik, did you gate the tweeter measurement? Looks like there is a reflection in there. Also, the woofer measurement looks like it's all far-field. Usually, you have to splice the woofer measurement, i.e., measure 5 mm away from the woofer dust cap (really close) and then measure far-field (maybe 6 ft), and join the two measurements somewhere in between. The far-field measurement is usually gated. This will give you a clean woofer measurement.

Crossover point looks a little low -- the tweeter is only about 10 db down at 1 kHz. You might get distortion. And the woofer level seems a bit high. It probably needs the baffle step compensation. At 100 Hz, it's about 85 db, rising to about 90-91 db at 300 Hz. This is the classic baffle step. The woofer output should be shelved so that it is flat from your bass roll-off corner, otherwise it will sound lean.

Have you read D'Appolitto's book? A must read if you want to do measurements and crossover design.

 

[eriksquires]

ra7, as I mention in the blog, all measurements are far field. I only took data for driver distances, but wanted to play with a crossover anyway.

The box should be 0.3 cubic feet with a port and 50% fill. There's no port or fill yet. In a sealed cabinet this should be 0.15 cubic feet and be -3dB at 100 Hz. Your expectation of a baffle step are natural, but given that the cabinet is either over damped (2x recommended cabinet volume) or still missing the port, I'd wait until my next router jig arrives and I put some fill in before I start to compensate for the baffle step.

I'm actually rather impressed that in a cabinet this large it's doing so well at 100 Hz.

I promise better measurements are coming.

Best,


Erik

 

[eriksquires]

Crossover point looks a little low -- the tweeter is only about 10 db down at 1 kHz. You might get distortion.

For a tweeter, the Vifa ring radiators have very low resonant frequencies.

This particular model is recommended from 1,500 Hz.

I'll look at the distortion again when I'm doing close-mic testing, but I do expect this to go a little lower than normal. Zaph mentions distortion making it a poor candidate below 2kHz, but he's been wrong about other distortion measurements before, I'll wait for my own before making a decision.

Best,


Erik

 

[eriksquires]

ra7, thanks for your helpful comments. I finally ordered a copy of the good Dr.'s book. You point out something I just realized. In the measurements I took of the Mundorf AMT's that comb filtering was never there. I may have to re-measure it to make sure I used similar conditions.

I'm not doubting the physics, I expect that sort of comb filtering at far-field in an un-damped room myself. This would be a good way to show the effects of a large surface area diaphragm compared to a point source tweeter like the Vifa or just find out my measurements were crap to begin with.

Best,

Erik

 

[ra7]

Great. That book is very helpful. There is some math in there, but stick with it. It also comes with his two AudioExpress articles on which measurements matter and how to interpret them. They are wonderful. They should be available online.

 

[eriksquires]

After the quick n dirty tests I've changed my mind about the box alignment and the woofer mounting. I'm going with sealed and surface mounted woofers. Details as always on my blog, here!

Thanks!

Erik

 

[eriksquires]

I just posted a page on measuring driver distances here:

A Speaker Maker's Journey: LM-1 Testing Driver Distances

 

[eriksquires]

Just added the first draft schematic for the LM-1 true bookshelf design here.

 

[eriksquires]

Hi Everyone,

I just realized that I misspelled the subdomain in my blog. Doh!

I've fixed it so all previous links won't work. The correct link to the LM-1 starting post is here.

The correct link to the speaker distance measurements is here.