Senior Project: Chimera, 3 headed beast

[Blazin]

Update: 2/25/08

Finished Enclosure Design:



Finished Crossover Design here

Enclosure BOM:

2 - 3/4" x 48" x 96" MDF
1 - 1/4" x 48" x 96" Plywood(back of front Baffles for t-nuts)

[ThomasW]

Physically time aligning requires the acoustic centers to be matched within 1mm, doing this is a bitch.

And doing this creates problems of it's own, since the solutions are a sloped baffle or steps in the baffle. Sloped baffles are a pain to build, the steps in stepped baffles create diffraction issues.

One can time align using the components in the crossover.

Any amp that supplies a full range signal can be used for measurements.

[Blazin]

Physically time aligning requires the acoustic centers to be matched within 1mm, doing this is a bitch.

And doing this creates problems of it's own, since the solutions are a sloped baffle or steps in the baffle. Sloped baffles are a pain to build, the steps in stepped baffles create diffraction issues.

One can time align using the components in the crossover.

Any amp that supplies a full range signal can be used for measurements.

Alright so, basically the solution to time alignment issues is measuring the response of the speakers again once mounted in the enclosure and tweaking the crossovers?

Basically, doing the correct measuring, calc, build, and measure procedure should compensate for some of the issues?

[ThomasW]

There are more important things to worry about than having the acoustic centers physically aligned...you can test this by putting the drivers on different baffles, move them back and forth and listen to what happens...

[Dennis H]

There's no need to build fancy baffles. Time alignment is just another way of saying phase alignment. With software like Speaker Workshop it's pretty easy to adjust crossover components until the phase matches through the crossover region. That's how commercial speakers do it and I imagine your prof would like that as it's an electronic solution to a mechanical (distance) problem. In your final paper, you can wax eloquent about how you had to juggle phase matching, flat on-axis response and vertical lobing with the crossover.

[Dennis H]

2.) Can I use any type of amplifier to power the speakers for various measurments? I'd really prefer not to unhook my Reciever to do the measurements, there is a shelf system in my lab, should have sufficient power (more than the PC onboard card).

Any amp will work as long as it has a low output impedance -- a small fraction of an ohm. If the output impedance is high, it will muck up your measurements as it interacts with the varying impedance curve of the drivers. Most of the measurements are at a watt or so, so it doesn't need a lot of power.

[Blazin]

I'm going to pic up the parts for the Wallin's jig today, looks to be about $40 at RS as most of the resistors are available in my lab.

I am realizing that the test components are going to eat up a decent amount of my budget.


Is there any way to save myself $100+ on the Microphone/PreAmp?

I saw a demonstration of MatLab measuring the frequency response of a small speaker the other day using a standard PC Microphone.. Would this work for SW as well?


My budget is around $700; however, I'm going to be coming close to that already without the addition of one time use(this project) Microphones/Amplifiers/Preamps.

[ThomasW]

Is there any way to save myself $100+ on the Microphone/PreAmp?

I don't think so....

I saw a demonstration of MatLab measuring the frequency response of a small speaker the other day using a standard PC Microphone.. Would this work for SW as well?

Use can use any mic you want, but don't expect the frequency response to be flat unless you use a dedicated measurement mic.

My budget is around $700; however, I'm going to be coming close to that already without the addition of one time use(this project) Microphones/Amplifiers/Preamps.

Budgets should be created after researching what's needed to accomplish the task.....

[Blazin]

I think I am going to build a preamp/microphone using one of the Panasonic capsule microphones..

Linkwitz, Wallin, and others have guides to do so..

Budgetwise, I thought I had figured everything out; however, I suppose I am missing a few things here and there.

[ThomasW]

IMO the amount of time/labor needed to build mic and preamp is such that one could work at a job and earn enough to buy the commercial units faster. There's no guarantee of the assembled accuracy of the DIY mics and the capsules are fragile....

The Behringer or the $10 cheaper Nady CM-100 mics are a known entities, so no guessing as to their accuracy...

[cjd]

If you're using Speaker Workshop, I prefer the cable setup outlined by Claudio Negro far better than the Wallin. Cheaper too.

C

[Blazin]

IMO the amount of time/labor needed to build mic and preamp is such that one could work at a job and earn enough to buy the commercial units faster. There's no guarantee of the assembled accuracy of the DIY mics and the capsules are fragile....

The Behringer or the $10 cheaper Nady CM-100 mics are a known entities, so no guessing as to their accuracy...

True; however, its not only the cost of a microphone, but also the preamp..

If the WM-60A capsule + basic preamp circuit is good enough for Linkwitz, I think it's good enough for me..

It's also only about an hour of labor to construct the preamp.

[Blazin]

If you're using Speaker Workshop, I prefer the cable setup outlined by Claudio Negro far better than the Wallin. Cheaper too.

C

I saw that, It was a little hard to follow.. perhaps I'll try to figure it out from his MSPaint pictures :lol:

[Blazin]

Any amp will work as long as it has a low output impedance -- a small fraction of an ohm. If the output impedance is high, it will muck up your measurements as it interacts with the varying impedance curve of the drivers. Most of the measurements are at a watt or so, so it doesn't need a lot of power.

Perhaps I'm thinking about this in the wrong way, after looking at some specs, what kind of amplifier has an output impedance of <1ohm?

[Dennis H]

Perhaps I'm thinking about this in the wrong way, after looking at some specs, what kind of amplifier has an output impedance of <1ohm?

Most any solid state audio power amp. A receiver would be fine. The spec is damping factor which is defined as 8-ohms/output-impedance and it should be at least 50. Sound cards or audio preamps aren't suitable as they have output imedances in the 100s of ohms and expect to feed into an amp with an input imedance of at least 10K. Tube power amps also aren't so good as they usually have an output impedance of 4 or 8 ohms due to the transformer. So you need to find out what kind of amp they have in the lab. If it's a low-current preamp kind of amp, you'll have to use your receiver. If it's an audio power amp, it should be fine.

Oh, and about the cost of the mic, it may not be a one-time deal. You may find you enjoy designing speakers and keep doing it as a hobby after you get out of school.

[Blazin]

Most any solid state audio power amp. A receiver would be fine. The spec is damping factor which is defined as 8-ohms/output-impedance and it should be at least 50. Sound cards or audio preamps aren't suitable as they have output imedances in the 100s of ohms and expect to feed into an amp with an input imedance of at least 10K. Tube power amps also aren't so good as they usually have an output impedance of 4 or 8 ohms due to the transformer. So you need to find out what kind of amp they have in the lab. If it's a low-current preamp kind of amp, you'll have to use your receiver. If it's an audio power amp, it should be fine.

Oh, and about the cost of the mic, it may not be a one-time deal. You may find you enjoy designing speakers and keep doing it as a hobby after you get out of school.

The "amp" in the lab is a shelf system, so I'm sure its pretty terrible.

my reciever has damping factor of 60... and you're slowly convincing me to go with a mic/preamp.. PartsExpress has that stuff I'm assuming, any combos to look for?

[ThomasW]

If you have a duplex soundcard and only need a preamp get the Rolls MP13 mini mic preamp

[Blazin]

So I guess I'm going to start a CBOM here

Drivers:

Tweeter:

http://www.partsexpress.com/pe/showdetl.cfm?&Partnumber=275-075

Midrange:

http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=295-362

Woofer:

http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=295-490

Behringer ECM8000 + XENYX802 Mixer

http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=248-625

http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=248-576

Crossover Components:
$250-300???

3/4" MDF:
$100

Various other build supplies:
$50


Total: Approx $750

[Dennis H]

Yeah, the Rolls is a good preamp for the Behringer mic. But, if money is really tight (I remember how it was as a student), there's nothing wrong with building the mic and preamp on SL's page. You can include that as part of your final report of what you did. The Behringer uses the Panasonic capsules so the quality should be about the same.

[JonP]

True; however, its not only the cost of a microphone, but also the preamp..

If the WM-60A capsule + basic preamp circuit is good enough for Linkwitz, I think it's good enough for me..

It's also only about an hour of labor to construct the preamp.

Well, if that's not scaring you off, (good for you) a wand/capsule antenna and small preamp might be the way to go... Mic preamps are a bit tricky, lots of gain, very small signals. Use a low noise op amp, good construction practices, and put it in a metal box for noise shielding. The capsules have the advantage to only needing a few volts rather than the +48v phantom supply like the Behringer and similar. The original type of capsule everyone liked is NLA, but you can get a WM-61 that is fairly similar. Most testing won't require the high SPL mod that Linkwitz mentions... so that's a little less a soldering challenge.

A thought might be, look up Kim Giraden (sp?) who provides a service of calibrating mics... or more correctly, measuring your mic and providing you a calibration file. He'll calibrate your mic or mic/preamp, which you probably want done, but he also provides mic capsules and I think, mic assemblies, that are pre cal'ed... Don't have his info handy but it's out there, on the Wallin site as I remember, too... He charges about $40 for the service, a bit more for a pre cal'ed mice capsule. You probably want to build your mic and preamp, and send the whole thing to get comprehensive cal data, but if you can test your preamp, and it's pretty flat... then you wouldn't need to.

[JonP]

Another note regarding Speaker Workshop... I would suggest making the cables that have the various test setups. I built the Wallin Jig, and a large amount of my learning curve problems were with accidentally leaving the switches in the last test configuration, bumping them and not knowing it, etc...

It's nice to have everything together in one box, but it was a big source of additional confusion. And, believe me, there is enough already, with discovering what settings work with your card, knowing when the inputs and the outputs are in clipping, and the various quirks of the program, etc... I'd say start simple, till you get your feet wet.

You might want to look for Arta, as another software alternative for testing... might be less "quirky"....

[Blazin]

Update:

Apparently, since my school is lacking the measurement tools to perform frequency response and impedance measurements for audio equip they told me they would be willing to invest money in products to do so based on my recommendations(what I'll be using).

With that in mind..

I will be using a computer with full duplex integrated sound card (the school's PC)

Money is not as large a factor now as test equipment will not be coming out of my pocket.

Are the wallin's jig and the Behringer/Rolls combo the best tools for the job in conjunction with SW?

I want to make sure that in the future students who undertake some form of acoustical measurment and design have quality tools.

Thanks all!

[Dennis H]

Well, it depends on how much they are willing to spend. If they want the BEST, they should order Praxis for measurements and LspCAD Pro for crossover design. In the neighborhood of $1K each. Also budget for a better sound card. The integrated ones are usually rather noisy and high distortion.

However, if they are on a typical cheap school budget, the Behringer/Rolls combo with Speaker Workshop will do the job if not as elegantly. You still need a sound card. Send the Behringer/Rolls off to Kim Girardin to get a calibration file for about $40.

[Blazin]

You still need a sound card. Send the Behringer/Rolls off to Kim Girardin to get a calibration file for about $40.

The school is probably going to stick with the least costly solution, until there is a lot of interest in this specific area..

I have an old sound blaster live value I could dedicate to one of the machines just to use something better..

What is the general calibration cost through Kim?

Got this info off trueaudio.com:

Kim Girardin
Wadenhome Sound
1400 Homer Rd. Suite 2
Winona, MN, 55987, USA
Phone: 507-454-8844
kmgrdn-at-luminet-dot-net

That still up to date?

Also, the Wallin's speaker jig, is there something equivelant the school could purchase?

[Dennis H]

I think Kim moved recently but the email address should still be good and probably the phone number. You don't really need a Wallin jig. Just build a couple of probes with series/parallel resistors so a speaker-level signal won't blow up your soundcard. And get a precision resistor that will handle a few watts to use as a reference for impedance sweeps. Clip leads are your friends.

All that said, I haven't used SW so I'm not sure what it needs in the way of probes, etc. Some of the measurement systems do a 2-channel measurement and compare the SPL from the mic to the voltage measured at the speaker terminals but I don't know if SW has that capability.

[ThomasW]

Info from the grapevine says that Kim K. hasn't has his mic calibrated in a long time.

As a result of that and the relative consistency of the ECM-8000, leads me to believe one should just save the $40.

[Blazin]

OP update.

[Blazin]

Used an online calculator to get a basic idea for XO components..

Looking at approx $280 in components for the pair


Anybody have an experience using the Dayton WT3 woofer tester with SW? If the school's going to purchase the components, might as well get something more fool proof than a plastic project box built by some dumb kid

[Blazin]

Update:

Built my Wallin's jig, loading SW on lab computer and installing an old SoundBlaster Live! Value 5.1.. Calibration tonight, possibly even a driver test - oh yes.


A question that popped into my head..

If I'm measuring the impedance/freq response of the drivers, should I do all of them (same type) and average the results.. or literally design individual crossovers based on the driver values I get in each network..

I guess it boils down to:

Am I measuring the freq/impedance in free air ONLY to get the values accurately into SW for box/crossover build or is it also because the speakers vary greatly in regards to their spec sheet.

[JonMarsh]

You should find individual drivers of the same type very close in their operating pass band. Breakup modes may look different, but their frequency location shouldn't be...

You should be measuring your drivers in the cabinet, not freee air. The enclosure/baffle mounting affects frequency response due to boundary loading and diffraction effects form the cabinet, and will influence the impedance curve, too.

Factory specs are usually done assuming infinite plan baffle for measurement.

~Jon

[Blazin]

i was under the impression i need free air to calc T/S to design an enclosure, to measure in the enclosure

[Dennis H]

i was under the impression i need free air to calc T/S to design an enclosure, to measure in the enclosure

Right. Depending on the drivers, the factory specs may be off so it doesn't hurt to check the T/S. Dayton specs are usually pretty good though. Then build the box and measure again in the box for the crossover work.

Edit: the bass driver should be the only one you need T/S for. The midrange should have its own sealed enclosure big enough to damp the backwave above the XO when you add some stuffing. Hopefully you'll be crossing well above the mid's box tuning.

[Blazin]

Right. Depending on the drivers, the factory specs may be off so it doesn't hurt to check the T/S. Dayton specs are usually pretty good though. Then build the box and measure again in the box for the crossover work.

Edit: the bass driver should be the only one you need T/S for. The midrange should have its own sealed enclosure big enough to damp the backwave above the XO when you add some stuffing. Hopefully you'll be crossing well above the mid's box tuning.

Right, thats what I thought.. If I have the tools to validate the spec sheets, why wouldn't I?

So last night I was attempting to calibrate SW, when I realized that one of my 1/8" > RCA cables was not working -damnit- so I stopped; however, When following Wallin's walkthrough, I found that I was unable to achieve his VRMS of approx 2.88Vrms.. I was using an Oscilloscope to monitor for clipping, and the Maximum volume output produced a Vrms of 540mV, with excessive clipping obviously. I suppose that the sound card I'm using cannot output that much voltage due to its very limited pre-amp, I plan on using a volume of approx 60 in SW as I could not visibly see any clipping at that output -- I know I will require an amplifier for sure to measure Freq Response, but will I be able to perform Impedance measurements still without 2.8Vrms?

[Dennis H]

You need an amp with low output impedance for all measurements.

[Blazin]

I went back to the lab with a working cable, read the SW manual -RTFM- duh.. and I was able to achieve Vrms of approx 1.14V without clipping, I set all of my volume levels, and calibrated SW finally.. I made it to the point of Impedance Measurements, and I had to take a break as I currently do not have a jig to hold my woofers vertically.

I noticed a comment on claudio's site about testing drivers in a generally non-reflective area, I am going to be testing my drivers in a lab consisting that is generally reflective, should I bring something like a blanket to place behind/in front of the driver when performing tests?

I'm ordering my Mic/Preamp sometime this week and I'll be looking for a cheap amplifier to perform the Freq Res tests.. any suggestions?

[cjd]

check the zaph audio blog for a great little test amp that seems to have a somewhat reasonable price, all things considered. I'm not sure if the T-amp would be good for this type of work or not, but it's a good price...

C

[Dennis H]

I think the T-amp is bridged so it might not work too well for 2-channel measurements. The (-) terminal is hot, not ground, and I don't think you'd want to connect that to the ground of your sound card input.

[jkrutke]

check the zaph audio blog for a great little test amp that seems to have a somewhat reasonable price, all things considered. I'm not sure if the T-amp would be good for this type of work or not, but it's a good price...

C

Those Teacs sold out right after I posted them on my blog. I just talked to the guy who runs the Dakmart place. I wish I'd bought a couple to have around. Actually, considering what I've seen them sell for on Ebay, I should have bought 10. Very nice desktop testing amp and all around small format amp. You don't see dual mono construction and toroidal transformers in small cheap packages like that often.

The T-amp doesn't work well for testing. It has a conflicting design, not to mention a horrible response curve.

[Dave Bullet]

i was under the impression i need free air to calc T/S to design an enclosure, to measure in the enclosure

Hi there,

If you've read Jay Butterman's SW manual then you probably already know....

http://www.audiodiycentral.com/ntutorials.shtml

1. Take an impedance measurement of the woofer in free air
2. Either using the delta compliance method (which means putting the woofer into a sealed box of the appropriate size for the woofer), or delta mass method (adding mass to the cone to change Fs by approx 25%) - do another impedance sweep
3. In the driver properties - set the impedance curves as per above and the method of VAS measurement from 2. above
4. Ask SW to calculate T/S parameters

I really don't think you need to clamp your woofer for the free air. Unless you are sending it a lot of power and it is a really heavy / big woofer, you can get quite good impedance measurements just placing the speaker with magnet on the edge of a table (vertically of course) holding with a finger or two and doing an impedance measurement (sweep). In my case - I just used the sound card output.

Dave.

[Satansfx]

this is a lil OT... but what is everyone using for a pre-amp for the DIY mics? anyone have a schem? Mine doesnt seem to work right.

[Blazin]

Last night I built my Speaker Stand for free air measuments, I had some 2x4s laying around so Figured I'd put a jig together.. later today impedance measurments will be posted.

[Blazin]

Alright, so I tried to measure free air impedance of my woofer, all I'm getting is pure garbage..

All of my calibrations are in place, I have followed through the manual a couple of times; however, I am left with the same garbage result...



Any help?

[Dennis H]

I don't know what's up with SW. I checked the manual and they do show the impedance being measured without an amp. AFAIK, all the other programs require you to use an amp and I don't really see how it could work very well without one. Here's a sketch from the JustMLS manual including the DIY probes with voltage-divider resistors. R is the precision reference resistor.

[chasw98]

[QUOTE=Dennis H]I don't know what's up with SW. I checked the manual and they do show the impedance being measured without an amp. AFAIK, all the other programs require you to use an amp and I don't really see how it could work very well without one. Here's a sketch from the JustMLS manual including the DIY probes with voltage-divider resistors. R is the precision reference resistor.

[QUOTE]

Dennis:
When I played with SW I kept seeing references to early Soundblaster and Ensoniq sound cards. Back in the day when a 4 watt amp was included with the card and powered speakers were not heard of. You know, when you and I were growing up. :rofl: I believe a lot of the info written about SW assumed a soundcard with a built in amp and that is why it is not called out in the various jig images. Just a hunch.

Chuck

[Blazin]

Yea, that garbage was definitely due to the fact that I was trying to measure the impedance from a DVC Sub..

I went back to the drawing board, tried out my RS150S and voila, an impedance graph that looks very similar to manufacturer specs when 'estimated'.

Lots of noise still, I'm going to attribute this noise to the fact that my soundblaster live value, cannot physically output enough wattage to power the speakers on its own, I guess I'll have to go purchase/find an amplifier for this measurement as well..

The lab 'amp' i spoke of, is broken, broken in the sense that the Volume does not work, it only turns up, not down, probably just a busted POT, but I'd rather not deal with it..

I'm still looking for a cheap Amp solution, I do not want to disassemble and reassemble my home theatre multiple times to test these drivers..