Thiele-Small Parameters measurement using dual added mass method.

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Aug 25, 2023

• #76

jcandy​

Hello. There is a question how to fix the driver during the measurement of parameters. The driver must be positioned vertically in the air. The question is how to fix it correctly? Intuitively, it seems that the driver should be fixed through the holes in the basket that the manufacturer made for this, to some kind of structure that can hold the driver so that there are no obstacles around the driver. But on your videos and pictures, the driver is fixed by a magnet. In some cases, there are solutions when the driver is fixed to the housing by a magnet, and the basket is fixed in the hole by pressing. But in most cases, the drivers are fixed to the case through holes around the circumference of the basket. So how to properly fix the driver so that there are no parasitic vibrations during the measurement?

 

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Dec 16, 2024

• #77

John Mulcahy​

I measured a driver that has very low sensitivity. About 80 dB. During measurements, the sweep was very quiet. It's usually louder. Can the measurements be trusted in this case? Should there be a dependence of the level of the measuring signal on the sensitivity of the driver being measured?

 

John Mulcahy

REW Author

Joined

Apr 3, 2017

Posts

8,220

• Dec 17, 2024

• #78

Dual added mass measurements include a Bℓ curve, it should be close to horizontal around the resonant frequency if the measurements are good.

 

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Dec 17, 2024

• #79

Thank you. The parameters include Lp. In my case it is 80.24 dB (1W/1m). The help doesn't describe what it shows. I couldn't find a description of it on the Internet. Does it have anything to do with driver sensitivity?

 

John Mulcahy

REW Author

Joined

Apr 3, 2017

Posts

8,220

• Dec 19, 2024

• #80

It is the calculated power sensitivity, but it is generally better to use the manufacturer's figure.

 

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Dec 19, 2024

• #81

Thank you.

 

jcandy

New Member

Joined

Dec 3, 2017

Posts

13

Location

San Diego

• Dec 31, 2024

• #82

@sm52

Just FYI if you have a Speakerbench-related question or documentation request you are welcome to open a new issue HERE

 

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Jan 1, 2025

• #83

jcandy said:

if you have a Speakerbench-related question

Click to expand...

Hello. Thanks for the invitation.

 

jcandy

New Member

Joined

Dec 3, 2017

Posts

13

Location

San Diego

• Jan 1, 2025

• #84

To answer your question in post 77 about measurement level, the usual advice is to say that for good comparison with linear theory, the level needs to be high enough to have a good S/N ratio, but low enough to avoid nonlinearity. Practically this means something between 10mV and 500mV for 3in-8in drivers. For larger drivers and for modeling high-power applications, you will of course want higher voltages. Is there a specific question about Speakerbench?

 

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Jan 5, 2025

• #85

jcandy said:

To answer your question in post 77 about measurement level

Click to expand...

Thanks for your clarification.

 

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Sunday at 7:19 AM

• #86

John Mulcahy,​

Does it matter at what signal level I calibrate the equipment before measuring the impedance? If I plan to measure impedance at levels of 100-700 mV, at what level should I calibrate? And if at the levels of 1.4-3 V, at what level should the calibration be done?

 

John Mulcahy

REW Author

Joined

Apr 3, 2017

Posts

8,220

• Sunday at 8:05 AM

• #87

sm52 said:

Does it matter at what signal level I calibrate the equipment before measuring the impedance?

Click to expand...

No.

 

S

sm52

Member

Joined

Mar 14, 2019

Posts

1,027

• Sunday at 2:09 PM

• #88

Thank you. The picture shows two tables of TS parameters created based on the same data. Left from REW, right from Speakerbench. I highlighted a small discrepancy in blue, and a large discrepancy in red. Is there anything that can be done to ensure that the parameters marked in red match closely?

 

Attachments

• 
  TS GHXamp 4 Ohm.png

  108.1 KB · Views: 14

trobbins

Member

VIP Supporter

Joined

Nov 26, 2017

Posts

137

Location

Australia

• Sunday at 4:25 PM

• #89

A lot of parameters have significantly more resolution than seems appropriate. Some measured parameters likely have a significant accuracy tolerance, depending on how the measurement was taken, and with what means, and whether the measurement was done at an appropriate location (eg. the voice coil resistance). It almost suggests the need to use a monte carlo simulation to appreciate what range a calculated parameter could end up with.