|
|
|
|
| |
 |
| |
Calibration
If you have not run a 1-20 kHz arbitrary sweep using the calibration resistor, you may be surprised to find that,depending on
how well you made the connections, the resistance is typically flat to 5 milliohms and phase to 0.05 degrees or better. This is
far better than what is required to test a woofer, but it does come in handy when testing inductors and capacitors.
Using the Simulator
Sealed and closed box simulations are variations of a common theme. To the simulator, a sealed box is nothing more than a
vented box with a very low vent tuning frequency. For example, a vent tuning frequency of Fb=0.1Hz is equal to a sealed
box. Similarly, an infinite baffle response is created by increasing the box volume to something very large. When
configured for open air, or a sealed box,the simulated impedance and phases should closely match the data taken in the
Q/Fs and Vas tests.
Triple Action Buttons
Buttons that have the form "< xxxx >" are known as Triple Action Buttons. If you click in the center of the button, a
dialog box will open allowing you to enter any arbitrary value. If you click on the left or right hand side of the button,
the value will go up or down by 10%.
Setting the Output Frequency
If you click in the center of the frequency control (or current set) button, a dialog box will open allowing you enter an
arbitrary frequency. If you click on the left or right hand side of the button, the value will go up or down by 10%. If
enabled in the Options pulldown menu, you can also set the frequency using the left mouse button by simply clicking on the
frequency point you desire while the mouse is over the T/S control window. Clicking on the right mouse button then adds the
current test value to the current data buffer. This however does does not effect previous calculations. It is mostly used
for data curve 'beautification'.
The Sweep Step and Minimum Search Ratio
Changing the sweep step ratio will also affect speed at the expense of a choppier looking graph. You can however go back and
add in points after the initial test is completed using the mouse.
During each test the WT begins its search using the sweep step ratio. As it crosses a point of interest
(such as when finding Fms), the algorithm will reverse its progression of testing frequency and iterate in ever smaller steps.
Ultimately, a value is found and the test resumes. The default value of 1.001 tells the software to stop narrowing its search
when the change ratio drops below0.1% change. A value of 1.01 (+/-1%), is probably more than sufficient, but you can also zero-in
on data by using a smaller ratio.
Bad or Noisy Data
If your data does not appear to be correct or noisy, check if the output drive current is too low, or the output may be muted.
The default action is to have the output sine wave enabled, but you can set the levels lower to measure higher impedances.
If the level is too low or off, the meter will simply try to lock onto noise and return gibberish. Another problem might be
that test leads need repair.
Increasing the WT2 update rate
Speed, accuracy and stability under Windows are all tradeoffs with the Woofer Tester. In general updates and sweeps are faster
when a smaller buffer size or number of buffers is specified. However at some point the data may become unstable or Windows
may not have enough time to process each buffer in which case a data drop out occurs. In Windows programming it is impractical
to process data on a sample by sample basis, so it is collected up into 'frames' of data usually consisting of many tens of
thousands of samples. Further complicating things, frames can also be shuffled around in frame buffer memory adding up to quite
a bit of delay, and this occurs seperately on both the input and the output! This is all fine and dandy for playing back or
recording a wave file, where many seconds of delay can be tolerated, but it is not at all well suited to realtime applications.
Though great care has been taken to minimize the CPU requirements, the software does rely on the PC to perform all user
interface and signal processing computations. Ultimately, this makes the tester sensitive to other things that may be going on
in the PC. If you are having trouble with consistency, it is advisable to shut down as many running processes as possible when
performing critical tests. The bare minimum for usability is about a 450 MHz Pentium class machine running Win98Se and
650 MHz for WinXP and Win2K. If dropouts do occur, there can be several other reasons.
First off other Windows multimedia programs can consume a large amount of processor bandwidth. The bandwidth usage however
comes in bursts as Windows prefers to process data in large chunks rather than on a sample by sample basis. This means there
could be large gaps where data gets clogged up. Not surprisingly this leads to large delays as well. The tester, being
being a Windows multimedia type of application also falls into this category. The solution is to increase the number of
buffers and the size of each buffer at the expense of update rate. If you have a speedy machine without much other stuff
other stuff contending for resources you can decrease these settings for a faster update.
The wave buffersize, number of buffer frames and data averaging are all set from the options pulldown menu when the
TS Control Window is active (these options are not available for other child windows). Please note that not all PCs will
be stable with a smaller buffer size or lower frames. The buffer size should be kept to a power of 2. 8192, 16384 (the default),
32768 and 65536 are powers of 2. The number of averaging buffers can also be configured in this same menu. Fewer buffers will
result in less accuracy while more will improve accuracy. You should only need to change the number (not the size) of wave
buffers if you are experiencing dropouts. This is highly unlikely, so we suggest keeping the number of buffers at the default
value of 3.
When the test frequency is below 10 Hz, the sample rate is automatically dropped from 44 khz to 11 khz, resulting in a 4:1
decrease in update rates. If you do not need to test at these low frequencies, simply go into the options menu and change
the lower sweep start point.
Conflicts with Other Audio Programs Trying to Use the WT2
Not surprisingly, it is possible for other programs or the operating system to use the Woofer Tester 2 for audio input and
output. In generally the tester should not be set as the default audio device for the system as this will cause disruptions
in the testers data stream, and sounds from the operating system will not be heard.
|
|
|
