Single sheet TH challenge

Hi david,

I have been thinking about it and maybe I have a more simple solution that is still user friendly I think and hopefully more easy to imply :)

Make for Zmin a separate box in the "Input Parameters" page (see figure 1). This way the user is responsible for manually searching for the Zmin in the "Electrical Impedance" chart.

After the Zmin is filled in the box the corresponding Watts appear in the "P" box beneath it. By changing the Voltage in "Eg" the power will automatically change (based on Zmin). When you change the power in the "P" box, the Voltage in "Eg" will change with it. If Zmin is not used there is also no value in "P".

This way the user can have full control over Power and Voltage.

In figure 4, I made a suggestion for a new screen page that shows the "POWER response" with all valuable information. If Zmin is not used this page is not available.

What do you think?

Power_Response_01.jpg
 
What do you think?

Hi Djim,

I thought that you wanted a chart showing SPL response calculated using the value of Eg that delivers 1 watt into Zmin. Your suggested power response chart has the same general form as the electrical impedance chart, only inverted - I’m not sure how much value it would really add in practice.

If you are prepared to manually search for Zmin, then why not just use the Calculate Parameter tool to find the value of Eg required to deliver 1 watt (or any other desired power level) into the chosen Zmin value, and to then calculate the standard SPL response chart using that value of Eg, giving you the sensitivity result you want.

The Calculate Parameter tool can be selected by double-clicking on the Eg input box in Edit mode - see attachment.

As a general comment - this whole process of measuring sensitivity seems to be very subjective :). It all depends on the value of Zmin within a passband that appears to be arbitrarily chosen. If the passband is not arbitrary, then what are the rules (applicable to any speaker) that can be used to determine the upper and lower frequencies. Comparing the performances of different speakers in this way seems to be rather questionable...

Kind regards,

David
 

Attachments

  • CalcEg.png
    CalcEg.png
    6 KB · Views: 378
Last edited:
I don't know, but I do know that the method proposed for power sensitivity modelling is not right.
As a general comment - this whole process of measuring sensitivity seems to be very subjective :). It all depends on the value of Zmin within a passband that appears to be arbitrarily chosen. If the passband is not arbitrary, then what are the rules (applicable to any speaker) that can be used to determine the upper and lower frequencies. Comparing the performances of different speakers in this way seems to be rather questionable...
For what it is worth I whole heartedly agree with these two quotes. Not much is to be gained. And secondly the means and methods are not different enough from what is already being done in Hornresp. The ambiguity in the industry is the main culprit. And trying to tailor a program for a very narrow interest group is not what seems to make the most sense.

As Andrew pointed out not one of us has demonstrated a universally accepted method of calculating sensitivity. Not because the methods are wrong. But because the wide acceptance is the problem.


I think David has given a very balanced means to do this. And some tools to delve deeper like the Calculate Parameter tool.


There is a saying in english. No use beating a dead horse.

Lets let David work on other improvements. He has some up his sleeve for sure!



 
Hi jbell,

Thanks to david, and all for the discussion. I've enjoyed reading and watching.

I also, would like to acknowledge those who have contributed to the loudspeaker sensitivity discussion - it has been a most interesting collaboration. Thanks to Djim in particular, for raising the issue in the first place. If nothing else, we have established that there is no universally-accepted single method for measuring loudspeaker sensitivity. Hopefully an agreed standard can be developed sometime in the future.

Meanwhile, some "food for thought":

The SPL response measured using the voltage that delivers 1 watt into Zmin is simply the SPL response obtained using any value for Eg plus an offset of 20 x Log10((Zmin ^ 0.5) / Eg).

For example, if the SPL at 100 hertz is 105 dB when Eg equals 2.83 volts, then the SPL at 100 hertz using the voltage that delivers 1 watt into Zmin = 7 ohms will be:

105 + 20 x Log10((7 ^ 0.5) / 2.83) = 105 - 0.58 = 104.42 dB

It's been a very valuable discussion for me, and I'm honored that it was associated with this thread.

And here's me thinking that you would be upset because we had hijacked your thread... :).

Kind regards,

David
 
I'm honored that it was associated with this thread
Hi Jbell,

Your single sheet thread seems to have become a standard on its own! Anyway, sorry for the "occupy movement" around here Jim :)

Maybe it didn’t lead to something concrete but it was an interesting read with quiet some useful info from everybody. To be honest I didn’t realise there were that many different interpretations besides the sensitivity based on Zmin and an average sensitivity figure. Thanks everybody and especially for you David, for your patients and willingness.
 
Hi Djim,

Just curious - what power is shown in the Figure 4 Power Response chart, and how is it calculated? Thanks.

Kind regards,

David

The “Flip Horizontal” function calculates that curve quite nicely ;).
It provides all any sound engineer needs to know, one can put a narrow Q boost at 56 Hz on the kick drum and not burn the voice coil in this cabinet because an amplifier won’t deliver significant power in to a 58 ohm load:spin:.

I consider Djim a friend, but this pursuit of his is a 58 ohm load of haufen mist :D.

Art
 

Attachments

  • Flip,Flop&Fly.png
    Flip,Flop&Fly.png
    152.2 KB · Views: 545
Last edited:
Hi David,

Figure 4 is like my other examples screens just an example and haven't paid much attention to its accuracy. But figere 4 is based on a Voltage constant of 2.9V (figure 1) at an power of 1 Watt based on the lowest impedance of 8.4409 Ohm in the bandpass at 88.67Hz.

Hi Art,

A model should be theoretically correct and doesn't 'see' reality. In other words in reality the impedance peak would be much lower and wouldn't end in a perfect pointy peak. Even for friends, in theory it would be difficult to burn out the VC at that specific point since it would go way beyond its Xmax much sooner ;)
 
Figure 4 is like my other examples screens just an example and haven't paid much attention to its accuracy. But figere 4 is based on a Voltage constant of 2.9V (figure 1) at an power of 1 Watt based on the lowest impedance of 8.4409 Ohm in the bandpass at 88.67Hz.

Hi Djim,

Thanks for clarifying that the Power Response chart shown in Figure 4 is for illustrative purposes only, and that the results are not meant to be an accurate representation of what actually happens in practice.

Just for the record - using your Figure 1 input parameter values, the predicted electrical input power will be as shown in attachment 1, and the predicted acoustical output power will be as shown in attachment 2.

The information used to generate the two attachments was taken from Hornresp exported chart data.

Kind regards,

David
 

Attachments

  • Power_In.png
    Power_In.png
    71.5 KB · Views: 463
  • Power_Out.png
    Power_Out.png
    68.9 KB · Views: 465
Last edited:
Out of curiosity, since all the data is already there in the export file why haven't you put it in a chart within HornResp before?

Hi Djim,

While the variation of electrical input power with frequency for a constant input voltage is certainly of some interest, I didn't think that it was really necessary to provide the information in a chart in Hornresp. Power values can be readily checked at particular frequencies if required using the Sample tool with the SPL response chart, or by scanning the exported chart data.

Incidentally, it is interesting to note the difference between the power calculated by Hornresp - which takes into account the resistive and reactive components of the complex impedance Ze (red trace), and the power calculated using simply the magnitude of Ze as given in the Hornresp electrical impedance chart (blue trace).

Kind regards,

David
 

Attachments

  • Power_Compare.png
    Power_Compare.png
    74.3 KB · Views: 88
Last edited:
Hi David,

Ooohhh I like these charts very much! The last chart raises the question if the complex calculation method is always similar to the Zmin calculation within the 1/4WL and full WL?

An example of how these charts can be used :)
Instead of looking to the SPL Response for setting out the bandpass, you can also use the Power chart for a more balanced power bandpass, or to show where the dynamic compression points can be found. Lots of extra information...
I use the 1 Watt horizontal as reference line in both examples (this could also be your max AES power or the power based on Xmax for instance).

Input_Vs_output_03.jpg
 
Last edited:
David,

Have you given any thought towards allowing the ability for complex inductance effects to be included into HR modeling? Many of the modern high power bass drivers on the market cannot be accurately simulated with parameters not including inductance or even using a simple inductance model when that information is actually given.

For example see the attachment. This is a simulation of a driver in a sealed enclosure using the measured specification without LE (Red response curve), using the simple LE model (Orange response curve), which is what capability HR currently has and finally using complex inductance (Yellow response curve).

q18 wt2 simulation.PNG

I have also attached a real measurement of the driver in question in the enclosure to show how closely the complex model matches and how progressively more inaccurate the simpler models are without this information. Note this is not an outlier driver either, this type of response and sensitivity variation due to inductance is very common with large bass drivers. Some are much worse than this.


close mic in room q18.jpg