Strange TSP values

[Radian]

Hi gang.
I am trying to find the problem of my setup because I get weird values when measuring TSP.

ESI U22xt USB sound card fed only by usb power.
Phone out, L R line in.
All three bars (out, ref in, normal in) show 12 db at 1khz test tone.
109 Ohm sens resistor made of 2x 220 ohm 0.25watts resistors in parallel.
Open leads calibration shows a nice straight line.
Closed leads is a little ragged past 10khz.
8 Ohm calibration again really nice.
Results measuring speaker impedance
shows a peak at 95 Hz although 12 inch speaker has definitely it's resonance below 50hz.
Same strange impedance results with little 3 inch speaker at about double the resonance.

Any help is greatly appreciated.
Radian

 

[sm52]

109 Ohm sens resistor made of 2x 220 ohm 0.25watts resistors in parallel.

Try a smaller resistor, less than 50 ohms, preferably 30 to 40 ohms.

Closed leads is a little ragged past 10khz.

What is the short circuit resistance?

 

[Radian]

Thanks I will try to lower sense resistor.
Don't know the short circuit resistance.
Open circuit shows 15db headroom
Short circuit 65db.
Is the short circuit resistance to be measured including sense resistor or after?

 

[sm52]

Is the short circuit resistance to be measured including sense resistor or after?

It can be seen on the graph on the left, vertical axis.

Open circuit shows 15db headroom

An open circuit should show resistance (ohms) on the vertical axis.

 

[Radian]

Thank you so much.
I will check once I get back home.

 

[Radian]

I just read in a forum that for the open circuit measurement the Rsens value is set to zero and the Rsens resistor bypassed. Is this correct?
If so will this bypassing of the Rsens resistor be repeated for the short circuit measurement?

 

[sm52]

Rsens is always present in the circuit, is not bypassed, its value is entered during calibration steps and is not reset.

 

[Radian]

Went down to 55 Ohm rSense and no change.
Blue impedance line starts at 3.4 ohms peaks to 26 ohms at 220hz and goes back to 4 ohms....,
Speaker is a Faital 3FE25 with 110hz resonance.
Tried Asio 96khz instead of java 96khz
sounds not clean and the impedance line is a bit ragged.

 

[Radian]

I get the same results with Arta Limp.
Resonant peak at 230 Hz although chassis has 110 Hz resonance.
Really strange.
I think I need to invest in DATS.

Thanks for trying to help.
Radian

 

[sm52]

REW does a good job of measuring impedance. These are graphs of three calibration steps and midbass impedance. Rsense 34 Ohm. Open circuit resistance 100 Ohms, short circuit resistance 4 Ohms (almost).

 

[John Mulcahy]

short circuit resistance 4 Ohms (almost).

That's really, really high. How could your test leads have such high resistance?

 

[sm52]

How could your test leads have such high resistance?

During the failures with the second step it was 18 ohms. So I was glad that somehow it became 4 ohms. But since you say that this is too much, I will try to find out why. Thank you.

 

[John Mulcahy]

Tried Asio 96khz instead of java 96khz
sounds not clean and the impedance line is a bit ragged.

Typically need to use a large buffer setting with ASIO (1024 or 2048 samples). With REW Java drivers best choosing the device names that start "EXCL" to use WASAPI Exclusive drivers. Use 48 kHz for your measurements in case there is some odd OS or driver resampling going on, 48 kHz is plenty for this purpose in any case.

 

[sm52]

But since you say that this is too much, I will try to find out why

I found the reason. Poor contact TRS male - TRS female on the measuring wire. Maybe some grease got on the contact. Now the short circuit resistance is equal to the resistance of the test leads. 0.6 Ohm. Thanks, John.

 

[Radian]

Typically need to use a large buffer setting with ASIO (1024 or 2048 samples). With REW Java drivers best choosing the device names that start "EXCL" to use WASAPI Exclusive drivers. Use 48 kHz for your measurements in case there is some odd OS or driver resampling going on, 48 kHz is plenty for this purpose in any case.

I got some error when trying to use EXCL. I get the feeling that my Windows 10 has a problem.
DATS is on its way but I'd rather save the money.