Perhaps. If for example, the impedance of a driver at 1Khz changes as a result of bass content, the delay caused by a very high Z cable may cause midrange delay change. With two identical cables, unknown if it will be heard. If you split one zip to make it 300-500 ohm RFZ, the settling time and hence the delay will become far more dependent on the driver impedance.
That is why I say, just go 4 by zip, 25 0hm Z, move on to next problem.
It seems nobody wants to spend the exorbitant amount of money two lengths of #12 zip entails..
jn
Probably 1 sample. But what Im getting from the discussion is that the problem only exists when certain parts of the spectrum are delayed, not the whole signal. So if 200 to 500 hz is delayed on one side, otherwise its like moving one speaker forward 2mm which I would think is an inaudable change. Someone correct me if Im wrong. So a test signal is not easy to make, thats why I asked if theres any already out there.
How much does speaker impedance change due to signal? This is new to me though it makes perfect sense. So a speakers freq response will be different for a sweeped sine than white noise, and different again for pink noise, and level dependent? Anybody measure this? Seems like more of a problem than the cable.
What are you trying to discover? In reality the whole spectrum would arrive at each ear at a different time, except for the parts that don't.
Also one the reason different measuring sw will yield different T/S parametres on the same driver.
dave
A long time ago, it would not have made sense to me. But as I learned how to use the Delta Tau motion system software and started playing, I found that I could run the PID loop hotter and hotter, distortion (following error) would reduce, but then I noticed that too hot and the system would start oscillation in the constant velocity portion of a sine move. That is the point where the system has very little intertia fighting the motion. The looser system bode plot would drop the phase margin and the system oscillated. During the acceleration phase it would be very stable because the inertia raised the magnetic forces and increased the phase margin. That was about 6 years ago. In addition, the gap devices we work with have magnets which attract like you can't believe, the top one pulls down harder on it's motor, the bottom one can actually go weightless. The result is, I can tune the top one much harder and keep phase margin, but the bottom one I have to reduce the gain so it doesn't oscillate at zero force.
The motion people are now starting to understand that the magnetic cushion, or force that is applied to the wires, can change the bode plot response during acceleration. One basic issue is that the magnetic forces in motors are in general a 1/r force relation, and during motion it can become even more complex. Solenoids and voice coil actuators are no different.
So now I can't see how the impedance can be the same during all phases of the voice coil motion within the gap as a response to acceleration forces as well as gap field variations caused by overhang front or back of the vc, or saturation effects in the pole piece and fringes, as well as eddy current issues. Lots of stuff going on.
By dropping the wire Z smack into the middle of the driver range, 25 ohms giver or take, my intent is to minimize the effect a varying impedance can have on the cable to load settling time. As, I've not a very good idea on how we can alter the drivers to eliminate the variation. And the 4 cable set is just so darn cheap..
jn
Last edited:
Is there any measuring software that can precondition the VC using a large low frequency signal, but actually measure the speaker at other frequencies simultaneously. Run 50 hz for example, scan the driver through full response. Also, be able to spot impedance change through the entire 360 degree portion of the sine, either the LF or the scanned F.
jn
In fact, Kunchur didn't wrote incorrectly about it, but used the term in a different way and that gave reason to the "bs-bingo-routine" by the usual suspect.
Afair, Kunchur thought his participants could detect intrachannel time differences down to the ~5us level, but it couldn't be ruled out that the detection was mainly triggered by a slight level difference instead.
Has it not crossed your mind that you are probably wasting your time?planet10, I would welcome a real answer to one of those lots of sound people can hear but cannot be measured as you claimed. It's been years and still no real answer from you other than dodging replies. I think we all would welcome your real straight answer.
@Evenharmonics,
it would help if you could cite some relevant literature/studies in which the effects of head motion were examined.
The studies, I'm aware of, so far provided evidence for a unconscious self compensating effect wrt to our head movements.
Beside any illness it is pretty normal for humans to move their heads nearly all the time and if you think about it, this self compensation mechanism makes a lot of sense.
it would help if you could cite some relevant literature/studies in which the effects of head motion were examined.
The studies, I'm aware of, so far provided evidence for a unconscious self compensating effect wrt to our head movements.
Beside any illness it is pretty normal for humans to move their heads nearly all the time and if you think about it, this self compensation mechanism makes a lot of sense.
Yes, I could prove that what he assumed to be a time difference was based on a mathematical error and was in fact a level difference of 0.7dB, exactly the level threshold of our hearing system.
I communicated this with him in several letters and he admitted that this was not his discipline.
Hans
j,
The most accurate T/S software directly measures Fs (phase is zero), then the 2 -3 dB points either side (phase min or max). The sw that guesses the parameters like WT3 might be able to do what you suggest but they ar enot as good at the job of generating useful T/S.
T/S measured on any day is a function of the weather (mostly temp/pressure) and change with drive level. Factory data is usually generated at higher voltages and in my experience is a better starting point for any design (at least with the better quality stuff i typically work with).
dave
The most accurate T/S software directly measures Fs (phase is zero), then the 2 -3 dB points either side (phase min or max). The sw that guesses the parameters like WT3 might be able to do what you suggest but they ar enot as good at the job of generating useful T/S.
T/S measured on any day is a function of the weather (mostly temp/pressure) and change with drive level. Factory data is usually generated at higher voltages and in my experience is a better starting point for any design (at least with the better quality stuff i typically work with).
dave
Ha! Ain't that the truth. I've got 8pr of drivers from 'a certain' well respected (and rightly so) company in for testing at work. Repeatedly measured, usual variation with graduated changes in voltage / current drive and break in. And continuing fluctuations caused by changes in ambient temperature, pressure, humidity. Not vast, but, say, a few Hz on Fs, and a couple of percentage points on Q etc., with fully calibrated gear.
Some measurements I had posted in the past
Speaker's Fr variation with signal level
Speaker's Fr variation with signal level
George
- Status
- Not open for further replies.