Semi-Le_Calc: Calculator for Advanced Inductance Model Incorporating Semi-Inductance

[bentoronto]

You left out one important step: compare your measurements and calculations to what you'd would expect based on similar drivers and common sense.

I am definitely not suggesting your measurements or math are faulty. But you shouldn't skip that step just in case.

B.

[bolserst]

Originally Posted by Mazza

...I'm thinking I've messed up the Semi Le calc process?

It may be a combination of a couple things. Looking at your imported data in Attachment #1, the resonance frequency for the added-mass measurement is essentially the same as for the free-air measurement. Do you know how much mass you added? You should see the resonance peak in the impedance measurement shift down in frequency. The "Vbox:" entry cell highlighted blue indicates that the added-mass measurement actually has resonance frequency slightly higher than free air, which is indicative of a closed box measurement.

The other item that will need to be addressed is to update the values for cone area and added mass before you click the Vas Calc button. See Attachment #2 for excerpt of the directions and input cells identified.

[Mazza]

Many thanks for your assistance here Bolser.

And I promise I read the instructions twice. Seems I still missed adding the added mass and SD! Apologies.

So I retook my free air and Vas. I used 615 grams added mass. Have reloaded the two text files into the attached and updated SD.

The FS did not seem to budge with added mass. And the workbook now says my Vas is just 0.06L.

I've attached the output from DATS in PDF.

Apologies if I'm still messing this up. Please hang in there with me!

Cheers.

[bolserst]

No worries, we’ll get it figured out.
Looking at the internally saved data from the spreadsheet you posted, both the free-air and added-mass files contained exactly the same impedance data. I would guess that when you use the [Measure Vas] function in DATS, it measures the added-mas impedance, but doesn’t save it to the workbench data memory slot. So, when you click export again you are just saving another copy of the free-air data. I don’t own a DATS system, but I played briefly with the software without the DATS hardware connected. Here is how I would recommend collecting impedance data for use with Semi-Le_Calc.

Step 1: Measure the Free-Air Impedance by clicking on the [Impedance Sweep] button.
Step 2: Export the Free-Air Impedance data to *.txt or *.zma file.
Step 3: Save the Free-Air Impedance data to Memory 1, so you can overlay it with the Added-Mass data.
Step 4: Add your mass to the driver cone, and measure the Added-Mass Impedance by clicking on the [Impedance Sweep] button again. The resonance frequency should shift lower compared to the Free-Air impedance curve you saved. With 610g added, I would expect it to shift down near 15Hz. Except around resonance, the remainder of the curve should overlay the Free-Air curve.
Step 5: Export the Added-Mass Impedance data to *.txt or *.zma file.

I think that should do it.
If anybody familiar with DATS has a better method please let us know.

[Mazza]

Yes that did it nicely - wonderful powers of deduction Bolser!

I did just as you recommended except I created two exports. One for the vanilla free air sweep, the second as another free air sweep but with the added 610 grams of mass.

So I didn't add the first to memory, then, and export the two as one file.

The difference this time was that my second measurement wasn't done using the "Measure V(as)" function in DATS. This doesn't seem to remeasure the impedance, explaining why my first attempts at the "added mass" sweep (which one would normally do using this "Measure V(as)" function) didn't get the desired data for the Semi Le workbook.

My bad entirely then!

Now I have a very reasonable Vas of 121L and you were spot on with your estimate of 15hz with that much mass added.

Thank you!

[bolserst]

Glad to hear you were able to get it sorted out.

[Mazza]

Hi Bolser,
I have rescued an old Celestion AD15H from the boot of my car as I wanted to see how it would model in the tapped horn I will be building for the Alpine SWS15D4.

I have owned this driver for 26 years and pounded on it in my car for all that time. It deserves better!

It sure models nice with Semi-Le enabled...



When I try to calculate Vas I get an FDD error - see attached.

Not sure what I'm doing wrong. Very much hoping for some more help please.

Thank you.

Cheers.

[bolserst]

Originally Posted by Mazza

...When I try to calculate Vas I get an FDD error - see attached

Could be wrong, but my guess is that the mass is not firmly attached to the cone resulting in some decoupling and secondary resonance peaks. It could also be that the woofer was not firmly restrained during the measurement. Can you describe how you attached the mass to the cone and how you restrain the woofer during measurement?

[mwmkravchenko]

Magnets are your friend in adding mass that sticks on a larger woofer. I use them on either side of the cone. The other thing that works well is play dough. Press it in evenly around the dustcap cone junction. I get play dough at the dollar store.

[Mazza]

Gents,
Thanks for your thoughts. I like the magnets and dough ideas Mark. I'll need to source these to try again.

Bolser, again thanks for your guidance. I did try a few different approaches that normally I have no problem with - a stack of different masking tapes. These are usually work well with the tacky side sitting securely against the cone.

With the Celestion AD 15H having a 4" VC the dust cap is large enough for me to take some extraordinary steps. See images. Nothing seemed to solve the FDD issue. I did go all the way out to 2.2 kgs added mass with a 10" driver and still I got the FDD.

Some files attached for your information.

Nevertheless my 4th tapped horn - I'm calling it the BP-MTH-266 - is under construction. And I'm excited!

Many thanks for this wonderful tool. It has really given me that extra level of confidence in my HornResp modeling.