Weird results by Impedance Measuring

[nader]

Hello
First Thanks for Development of such a useful software.

I need help with measuring Impedance, and the problem is very very small results by measuring, please read the problem description:
Hope someone can help me.

I try using REW to measure Thiele Small Parameters using Added Mass Method.
To do that, I use REW V5.20.13, a PC (using front panel MIC and Headphone interface).

As the first step by reading help file, I started by Measuring Speaker Driver Impedance and I made a jig as following picture (same as suggested in Help File):

The 2nd Step, I did Calibration as the Following settings:

and I did all 3 steps of Calibration process.
After I measure a known speaker, the results on Y-Axis looks so weird, the Impedance show about 20 mOhms to 50 mOhms, that is so low.
(mdat file is attached)

The Driver Impedance graph from manufacturer shown as below:

My Bests
Nader

 

[John Mulcahy]

The soundcard measurement is pretty odd, does it have some sort of tone control applied to it? Those results should usually be very flat. You may need to be careful using a PC mic input, it needs to be mic/line and to be operating in line mode.

The short circuit impedance is pretty high at around 0.6 ohm, are the leads very long?

Are you certain the correct value was used for the ref resistor measurement?

 

[nader]

The soundcard measurement is pretty odd, does it have some sort of tone control applied to it? Those results should usually be very flat. You may need to be careful using a PC mic input, it needs to be mic/line and to be operating in line mode.

View attachment 62418

The short circuit impedance is pretty high at around 0.6 ohm, are the leads very long?

Are you certain the correct value was used for the ref resistor measurement?

Hello John
Thanks for your fast response.

1. About Soundcard, I turned off all filters and effects on both MIC and Outputs (Speaker)
2. I did Changed MIC intput setting to Line-in, but results still same.
3. Total length of Leads about 40cm and leads quality are good.
4. Resistor Value I used is 100.16 ohm, measure with very accurate LCR meter.

Some New Settings I did you can see as following images, and attached mdat file:

Thanks again for helping me

My Bests
Nader

 

[John Mulcahy]

Resistor Value I used is 100.16 ohm, measure with very accurate LCR meter.

I mean the resistor used for the reference cal measurement.

On the Windows audio properties for input and output, Enhancements tab, make sure "Disable all enhancements" is selected.

When you make the soundcard calibration measurement you don't have your impedance jig connected, do you?

In your diagram you show a mono jack for the headphone port, but that will be a stereo output, you should use a stereo jack otherwise one output channel will be shorted to ground.

 

[Veleric]

You may need to be careful using a PC mic input, it needs to be mic/line and to be operating in line mode.

I'm about to make my own impedance rig. Like Nader, I am planning to use a mic/phono jack on my PC for the inputs. But my sound card audio manager doesn't have a place to explicitly choose "Line-in" for the mic input. It only asks me to choose whether the input is a mic or headphones. Does choosing "mic" as the input do the same thing as setting the input to "line in"?
Thanks,
Eric

 

[John Mulcahy]

It only asks me to choose whether the input is a mic or headphones.

If the choice is between mic or headphones it might be a headset connector. Those have stereo outputs for the headphones but only a mono mic input and cannot be used for impedance measurements, which require a stereo input.

 

[nader]

I mean the resistor used for the reference cal measurement.

On the Windows audio properties for input and output, Enhancements tab, make sure "Disable all enhancements" is selected.

When you make the soundcard calibration measurement you don't have your impedance jig connected, do you?

In your diagram you show a mono jack for the headphone port, but that will be a stereo output, you should use a stereo jack otherwise one output channel will be shorted to ground.

Hello John,
First, Thanks a lot for your great support and Help.

Refer to your guidance, I rechecked the setting of soundcard and here is the results:
A) I set input MIC as Line-in.
B) Disabled all Sound Effect on both Speaker and Input.
C) The Measurement Resistor Value is 21.964 ohms (measured by a very accurate LCR Meter).
D) Leads Resistance is about 0.00986 ohms
E) SPL Chart Measured when the circuit was open ( No Speaker Attached, No Leads connected)

Results:
1) Seems The SPL Curve get more closed to zero value and straight line. ( please let me know your opinion).
2) The Impedance Curve still very low and totally out of range.

Please see attached Files:

Wish You The Bests
Nader

 

[John Mulcahy]

The soundcard cal looks good now.

The impedance reference calibration is quite noisy. I still have no explanation for the strange impedance measurement. You could try the 5.20.14 early access build, that might pick up an error that 5.20.13 is not detecting.

 

[Veleric]

If the choice is between mic or headphones it might be a headset connector. Those have stereo outputs for the headphones but only a mono mic input and cannot be used for impedance measurements, which require a stereo input.

John,
When "mic in" is chosen, there is a Left/Right slider for recording volume. That makes me think that the mic input is stereo. What do you think? Am I being over-optimistic?

Also, the specs for the computer list these "three" jacks. The mic and headphone jacks are actually the same jack, and there's no other jack that could be the line-in, unless the one jack is actually all three. Do you think that could be right? Do I still have a chance this PC could work for impedance tests?

Thanks,
Eric

PS, Nader, sorry to hijack your thread, hope you don't mind!

 

[John Mulcahy]

If there is a line in socket you should use that for input and use the headphone socket for output.

 

[Veleric]

If there is a line in socket you should use that for input and use the headphone socket for output.

John,
Thanks, I should have been clearer about my planned setup.
For the output, the PC has an HDMI output that I will run to an old Sherwood RX 4109 stereo receiver, and use the receiver's 6.35 mm TRS Headphone jack as the output. That should be fine right? (It's the same as I already use for frequency sweeps with REW)
For the input, I plan to use the PC soundcard's 3.5 mm TRS "combo" jack, which apparently can serve as a mic jack, or a headpone jack, or a line in. This is the input I normally use for the mic when I do frequency sweeps. But to confirm it could be used as a stereo line in for the impedance measurments, I connected the analog outputs from my CD player to the combo jack and got stereo sound from the PC's speakers. They sounded horrible, but it was definitely in stereo. To be certain it was in stereo, I played Pinball Wizard, which has very distinct left and right channels, and it definitely was. So that input should work for the impedance measurment right?

As for the rig, this is what I am planning. I'm not too familiar with electrical schematics, so I had to try and draw it a little less like a schematic and more like what I actually thought it would look like, Does this look right?
Thanks,
Eric

 

[John Mulcahy]

For the output, the PC has an HDMI output that I will run to an old Sherwood RX 4109 stereo receiver, and use the receiver's 6.35 mm TRS Headphone jack as the output. That should be fine right?

Maybe, it depends whether the input and output will use the same clock, hard to tell. It is much better to use the same audio interface for both input and output.

 

[Veleric]

John,
Thanks.
I actually thought that I was only doing what you suggest in the Help file:

What is it that I am misunderstanding? Sorry I'm so clueless!

Thanks
Eric

 

[John Mulcahy]

The output should be on the device that is also being used for capture, or an amplifier/buffer could be used but that requires special care to avoid possible damage to the inputs if the output level is too high.