This thread has been inactive for a while, but it is close to what I came to ask/talk about today.
Enclosure simulators predict frequency response (and other things), but do not predict impedance plots.
While FR of ported (to include back horn and the new style of TL) enclosure is difficult to measure, the impedance curve of the enclosure and transducer combination is easy to measure.
An easy check of simulation to build would be to have a predicted impedance curve.
I don't know of a simulator application that does this. Any suggestions?
If you are a programmer able to work in C++ and linux, let me know. I would love to have some help developing such a simulator if none already exists.
Mark
Enclosure simulators predict frequency response (and other things), but do not predict impedance plots.
While FR of ported (to include back horn and the new style of TL) enclosure is difficult to measure, the impedance curve of the enclosure and transducer combination is easy to measure.
An easy check of simulation to build would be to have a predicted impedance curve.
I don't know of a simulator application that does this. Any suggestions?
If you are a programmer able to work in C++ and linux, let me know. I would love to have some help developing such a simulator if none already exists.
Mark
MarkMcK said:
I believe both Unibox and WinISD do this. My own sim does, too
It is not that difficult. Get yourself a copy of "Theory and Design of Loudspeaker Enclosures" by JE Benson.The problem with Impedance is that fiber reactance changes somewhat with frequency, and fiber in proximity to the cone alters the effective parameters.
MarkMcK said:
Have a look at Basta in my signature...
An externally hosted image should be here but it was not working when we last tested it.
Thanks for the suggestions guys. I am glad to know that there are simulators providing this data.
Regardless of the fiber problem, this will be an easy check of a built simulation. If the imped curve is way off of prediction (not unusual for extreme alignments), then the frequency response of the build is not going to follow prediction either.
I would like to see more publication of predicted and measured imped curves by builders.
Mark
Regardless of the fiber problem, this will be an easy check of a built simulation. If the imped curve is way off of prediction (not unusual for extreme alignments), then the frequency response of the build is not going to follow prediction either.
I would like to see more publication of predicted and measured imped curves by builders.
Mark
Yup, Fiber fill materials. Do you have a model for this in Basta, Svante? Is it purely resistive or more complex?
IIRC, Leach created a lumped model for fiber effects and wrote an AES paper on it ~1988. I seem to recall some comparisons to measurements, but the memory is dim. I need to do some more digging in this area...
I have seen some measured results where impedance peaks where at about the same frequency, but higher and sharper than predicted even for a lossless box - that really threw me for a loop. I wasn't interested enough at the time to do anything with them
Regards,
Ron
IIRC, Leach created a lumped model for fiber effects and wrote an AES paper on it ~1988. I seem to recall some comparisons to measurements, but the memory is dim. I need to do some more digging in this area...
I have seen some measured results where impedance peaks where at about the same frequency, but higher and sharper than predicted even for a lossless box - that really threw me for a loop. I wasn't interested enough at the time to do anything with them
Regards,
Ron
I am not sure why calculating the impedance plot of a driver in an enclosure is so hard with or without fibers. My MathCad models have been fairly accurate at calculating impedance curves which are indeed good indicators of the driver's motion in an enclosure. I have measured, or seen other people's measurements, that correlate well with calculated impedance curves for BR, TL's, and BLH's. The peaks are in the correct location and the magnitude and phase seem to be very close between calculated and measurements with and without fibers.
Ron E said:
Basta! uses a purely resistive model. I vaguely recall this article too, I think. My impression is, however, that the resistance of the filling is only important near resonance, and that a resistive model is ok in almost any case. I do not predict the resisitance from an amount of filling, but let the user enter a Q value.
A sharper peak than predicted in the impedance curve would IMHO not be explained by the filling, but a lower than expected Rms value. Pity that you did not follow this through, it would have been interesting to know what happened.
MJK said:
Never really claimed it was, at least not for standard models. Impedance is perhaps more sensitive than amplitude WRT changes in loading between free air and mounted in a box - In other words parameter measurement accuracy.
Motional impedance is just the diaphragm transfer function with Qts in the denominator replaced by Qms.
Svante said:
Do you mean that just the resistance is important near resonance or that resistance only matters near resonance?
I never liked those "fill approximation" methods for calculating Q. I guess if I knew they were based on measured data I would be more OK with them. Otherwise it could just be a naive interpolation.
WRT sharper peaks, the only thing that came to mind at the time was parameter measurement inaccuracy or use of advertised specs, but I know the parameters were measured.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.