SW question

[kro5998]

oops the wrong file!! here's the right one.

I am starting to use Speaker workshop and i'm having a problem with the frequency measurements.

The impedance measurements are fine and as expected.

When i do frequency measurements i get a 20dB drop from 5k to 20k, both with the tweeter and the midrange measurements.
I know for a fact that the tweeter actually has a slight rise.

The nearfield and farfield measurements show the same result.
I've tried different brands of microphone with the same result.

The amplifier frequency callibration file shows a straight line up to 5k and then drops 10dB to 20k,almost as if a first order filter is applied.

Since both speaker show exactly the same behaviour i suspect a calibration issue.

Any ideas or suggestions??

kro5998

[dick van nierop]

Hi Kees,

Nice to meet again.
What sound card do you use?
What is the sample frequency you're using?
For measurements up to 20kHz you should use the highest frequency you card will support. Normally this would be 48 kHz, on older cards it would be 44 kHz.

You could do a direct measurement of the output of the amplifier (or soundcard if you're not using an amplifier). This should give a 0 dB and 0 degree phase response.

Have you checked and calibrated channel difference?

Can you post some charts?

Mail me if you like. We can also use he phone.

Dick.

[soongsc]

Originally posted by kro5998
oops the wrong file!! here's the right one.

I am starting to use Speaker workshop and i'm having a problem with the frequency measurements.

The impedance measurements are fine and as expected.

When i do frequency measurements i get a 20dB drop from 5k to 20k, both with the tweeter and the midrange measurements.
I know for a fact that the tweeter actually has a slight rise.

The nearfield and farfield measurements show the same result.
I've tried different brands of microphone with the same result.

The amplifier frequency callibration file shows a straight line up to 5k and then drops 10dB to 20k,almost as if a first order filter is applied.

Since both speaker show exactly the same behaviour i suspect a calibration issue.

Any ideas or suggestions??

kro5998

What kind of mic are you using? I once used an unknown 9mm dia mic capsule I got the same results. What kind of speaker are you testing?

[kro5998]

I'll try to post some charts.Haven't figured out how to convert them to JPG yet.

I am building a three way setup with the high and mid in a separate cabinet.The woofer will be active filtered,the mid/high will be passive and i am focussing on this right now.

The cabinet is a 2.3Ltr closed box for the mid speaker.
This is a Vifa MG10MD09/8.
The tweeter is a Peerles WA10/8.
The woofer is a Peerles CSC217R in a TL cabinet.

The mics i have used are a Panasonic MCE100 and a unknown brand with the code M300 stamped on it.(i think its Monacor or something like that).I've build the Wallin mic pre for this.

The souncard is the standard onboard card for the Compaq PC i am using for the measurements.It's a Soundmax3.0 Full duplex chip with 44kHz 16bit capabillity.
It has no output amplifier so i build the Wallin jig with a build in TDA2003 Amplifier for testing purposes.

The sample rate is 44kHz and the ample size is 256k.
I used no pre-emphasis and a repeat count of 10.

Here are the frd. files for the mid/high and amplifier.The driver files are nearfield measurements.

kro5998

ps Thanks Dick i will mail you if i can't figure this out or have other questions.

[kro5998]

this is the calibration file. just rename the file to FRD.

[kro5998]

The measurement files are to big to post here (+ 4 MB) i'll try to convert them to jpg first.

kro5998

[kro5998]

here's the calbrationfile in jpg

[kro5998]

here's the tweeter file in jpg

[kro5998]

and the mid file in jpg

[soongsc]

This looks very much like a mike problem. It looks like your mic was originally intended to be used for normal voice recording. I would recommend changing to a mic using something like a panasonic wm-61. Probably most of those 6mm capsules have very good response.

The calibration file you are showing probably is not a mic calibration file but a sound card calibration "measruement.calb" Normally you should be able to callibrate it to perfectly flat. Attached is an old one that I did without specific knowledge. But if the interchannel latency value is adjusted in muliples fo 1/(sample rate), the at some point it would be perfectly flat. For example for a 48KHz sample rate card, each time adjust the interchannel latency by 0.02083ms, for 44.1KHz sample rate it would be 0.02268ms.