Well, 10Hz wavelengths are fairly long and will quickly generate a lot of heat even at low power, and that will change the Fs.
The basic principle is to move the driver as far as possible with as little current. One of the things you can do if you are really concerned about heat is to play a breakin tone at the Fs of the driver. With the resistance being extremely high, the current is very low and as a result heat is almost non-existent. For all practical purposes though, the 10hz tone having high driver excursion allows for good air movement to cool the coil. We monitor DCR as it is does affect Bl when measuring the rest of the parameters. If DCR has risen, of course we will not be measuring Bl until the DCR goes back down to the rest value. In a driver with a copper sleeve on the pole like we have, this happens very quickly as the copper wicks the heat from the coil much faster than a steel pole alone. Until the entire steel pole begins to have a significant rise in temperature, the heat travels from the coil to the copper and into the pole quite quickly. Again though, we are talking a 30 sec tone with 1-2 amp current flow through the coil in the case of the IB15. This is not going to significantly heat the coil.
As far as heat changing the Fs of the driver, the only way that can happen is in cases where the compliance of the spider would change due to the temperature change. This will not happen in 30sec on these drivers.
Yep, I believe that. That is about right. However, the change you see after 30 seconds is NOT a result in the change in suspension compliance. At least not all of it. It is mostly from the heat change. Run your 30 second blast then let it cool for several hours and you'll see the difference in compliance change and change from heat. My guess is that you'll see the Fs drop back into the 19Hz range.
We used to exercise the driver suspension by hand when verifying prototype units. In reality just physically moving the suspension to it's limits back and forth 20-30 times creates the exact same effect. The nice thing about the 10hz tone is that I don't have to sit there and move the driver back and forth myself. It's kind of a pain in the butt to do. Other ways to do it are to take an air actuator with short throw and attach it to the cone and adjust position back and forth with that, or in a sealed enclosure use a vacuum line to increase and decrease the air pressure to move the driver. These physical displacement methods introduce no heat. In reality though the 30sec tone has the same result. Again, you can tell if heat is doing anything by monitoring DCR. If you do not notice a change in the DCR, you won't have enough heat to change the spider compliance either.
No, there are two things changing. There is a mechanical and an electrical change. Both will change over time.
I'm curious as to what you are referring to that electrically changes over time. The only drivers with electrical change over time are older Alnico drivers that do demagnetize over years or some of the original Neo magnets that also demagnetize with heat at much lower temperatures than ceramic. As a result in those drivers BL goes down and Q goes up. Ceramic magnets do not demagnetize under any real world use as we can never get them physically hot enough to reach a demagnetization point. I have drivers that I have had in operation for over 10 years now. Bl and Re are the same as they were when the drivers were made. There are of course the old wives tales about the properties of copper changing over time, etc. Some state change in properties due to heat over time and others state due to the direction of current flow. This has of course been proven to be false in both cases. The amount of heat created in the process to forging, drawing and annealing process is much higher than any temperature the VC can ever see. Heat during operation has no affect on the property of the wire. Signals going through the wire are alternating current, so they don't create any "polarizing effect" on the wire as the high end cable guys would like to believe.
No, that is not true either. Part of the damping is the field strength around the voice coil. Just increase motor strength and and Fs will come up. Decrease it and Fs will drop. Heat it up and Fs will drop. Cool it off and Fs will come up. You might also note that the DCR will change after your 30 seconds of hard play. Try measuring the DCR before and after heating it. Also, if you don't adjust for the change in DCR after heating it then it will also alter the parameters.
The field strength around the coil is an issue of the electrical damping. This is your Qes. Changing Qes does not change Fs. Fs is only a function of mass vs compliance. It is a simple physics problem and models as any other mass on a spring problem.
fs = 1/(2*pi*sqrt(Cms*Mms))
There is no Qes or Bl factor in there. The only way to shift Fs it to change Mms, which is a physical amount that can't change, or to change Cms. Changing Cms as the suspension warms up will change Fs accordingly, but that is not happening with a 30sec test tone at low power. As a check to validate what the laws of physics are stating, I can do the same driver with one or two magnets. This changes Bl by about 15% in most cases. However, Mms, Cms, and Fs do not change. The laws of physics don't allow them to.
I wouldn't expect it to show much else, would you? If you look real close you might see the slight change in the minimum impedance area due to the changed DCR. It may be pretty hard to see on an impedance sweep though.
If there is heat, yes, there would be a change in the DCR of the driver. I can zoom in as close as i need to verify details like these. I also have a cursor that tells me what the resistance at 1hz is. By looking at that number I can get a pretty good idea of any change in DCR
But you are going to see more differences in parameters by having to allow for and calculate the air loss of the box or the differences between different drivers than you will see from simply using the added mass method to get the measured parameters. So I wouldn't be to hard on the added mass method. If done correctly it too can be very accurate.
The benefit we have is that we can physically weigh each part going into the driver. We often cut drivers apart to measure the whole Mmd just to double check we are still using consistent amounts of glue, etc. We can take the known Mmd and calculate the amount of mass to add to get Mms using the following formula.
m = (8/3)*rho*r^3
where
rho = density of air
r = piston radius
So we have a known Mms, known DCR, known Sd. We need just the simple impedance curve to calculate Fs, Qms, Qes, and Qts as a result. In reality when starting with the known Mms, we don't even need to do anything for delta compliance method. Cms and Bl calculate already. The impedance curve tells me everything I need to know if something is out of line or not. The key as I wrote in the post on our site is getting an accurate and repeatable impedance curve. We just do the closed box test to verify and double check what we already know. Using multiple methods to verify the same results guarantees the accuracy.
John
