|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
76.90.39.24
.
Have Fun and Enjoy the Music
"Still Working the Problem"
Follow Ups:
Are you after phase shift or phase angle? If this is a take off on the discussion the single ended forum--- as in load lines--- then it is phase angle.
MSL
Builder of MagneQuest & Peerless transformers since 1989
Tre, can you explain your question in a little more detail? I looked at it with two other people, and none of use was sure exactly what you meant.
--------------------------
Buy Chinese. Bury freedom.
I have some output transformers that the maker rates as -1db at 28Hz.
I'm trying to work an equation backwards to calculate the inductance.
If I knew the -3db point I would just plug that into L = R/(6.2834*F)
Thanks
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"I have some output transformers that the maker rates as -1db at 28Hz... work an equation backwards to calculate the inductance."
I'm trying to visualize this process in consideration of all the variables involved. Personally, if I was about to design an amplifier using those transformers, I would want to measure them myself.
--------------------------
Buy Chinese. Bury freedom.
Phase shift is arccos(10^(-1/20)).
Look at my diagram. You can work it out from there, except you use L/R in place of 1/RC.
-Henry
Hey,
Assuming a first order filter the -1dB point is an octave above the -3dB point so you can just use L = R/(pi*F) (or you could use 14hz and 2piFL).
The problem you are going to have is finding what value was used for R in the transformer test. A possibly better test would be to apply an AC voltage and measure the current through the primary then by using ohms law you can calculate the inductance. If you do try this approach be sure to use a sizable voltage (say 10V @ 60hz) to ensure enough excitation.
dave
.
Have Fun and Enjoy the Music
"Still Working the Problem"
...aren't Tre's initial #s the result of a simple RC HP circuit? I suppose RL works too although the sign of the angle changes. Or did I assume too much?
It's been such a long time since I've done this stuff; I'm a bit rusty. Anyway...
To plot the magnitude and phase response of a single-pole filter, draw the complex plane as shown. This is a real pole, so it sits on the real axis at σ = -1/RC.
For a sinusoidal input, you plot a point at the frequency of interest on the jω axis. In this case, I picked ω=1/RC. Draw a line from this point to the pole. The magnitude of response is inversely proportional to the length of the line. The phase shift is -θ.
So, for example, in this figure the phase shift is -45 degrees. The length of the line is sqrt(2) times its length at DC. So the magnitude of response is 20 * log(1/sqrt(2)) = -3.01dB.
To find the phase at any level of attenuation, just work the equations backwards for whatever value you want.
This picture gives you a nice visual indication of why the ultimate phase shift is 90 degrees, since, evidently, as frequency goes to infinity, the angle of the line approaches straight up and down.
It also explains why the phase shift is -45 degrees when ω=1/RC.
FWIW.
-Henry
...and let Excel do the heavy lifting. And now seeing what Tre' is really attempting to do (not a simple RC hipass), it seems Dave Slagle's method is probably best. Never Assume!
If you want an accurate measure of a transformer's L then there is no replacement for using an inductance bridge.The question I have in using Dave's suggested method... is all of the current so measured the result of just the load presented by the L? Does it isolate the other transformer losses (loads)?
Also Dave's suggestion of using say 10 volts at 60 hz is a good start.
If measuring L you want to excercise the iron pretty well--- the L will generally keep increasing as the emf force is increased up to the point of it hitting the knee of the BH curve. So you might also want to try
some larger drive voltages (again keeping the frequency at 60 hertz or less).If you want to get toward what is often called OCL (open circuit L) then you might want to measure the L at the transformers full rated power. Using ohms law you can derive the volts necessary to drive the nominal primary impedance to full rated power. Doing this at 60 hertz should not overload the primary since most audio transformers are going to rated at frequencies lower than 60 hertz.
But, again, an inductance bridge is the best tool for these measurements.
MSL
MSL
Builder of MagneQuest & Peerless transformers since 1989
Edits: 01/30/15
Nothing to argue with in what you say but now that we better understand Tre's objective, I would think Dave's proposal ought to serve the purpose. It would be another matter if Tre' was attempting to thoroughly characterize his OPT but I don't think he is. Interestingly, the method Dave proposes is virtually identical to the method D.T.N. Williamson suggested to verify the inductance of his OPT design for the "Williamson Amplifier" when it was constructed by DIYers in the late 40s of the previous century.
Yeah. My feeble attempts to use simple models to extrapolate transformer parameters have never met with much success. I agree with your comment about Dave's method.
Of course, the politically correct response is that Tre should just lash it together and see how it sounds...
-Henry
...but not before grafting on triple leads of Kimber TCCS for each stock lead. ;-)
...but since you're working with -3dB and 45deg, I'll assume 1st order RC: I cranked out ~ 27deg.
Edits: 01/29/15
.
Have Fun and Enjoy the Music
"Still Working the Problem"
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: