How to construct a cube louver (Acoustat) - Page 14 - diyAudio
 How to construct a cube louver (Acoustat)
 User Name Stay logged in? Password
 Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read Search

 Planars & Exotics ESL's, planars, and alternative technologies

 Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
diyAudio Member

Join Date: Jun 2005
Location: Fort Worth, Texas
Quote:
 Originally Posted by jelanier Inductance varies as to the square of turns. L2 and L3 share the same field. L2 represents half the number of turns. Twice the number of turns would yield 4 times the inductance. The load is connected at 2X. Since L2 is half the number of turns it represents a step up ratio of 25:1. So .288H x 25^2=180H If you model this as a single ended design (a single winding of 720 with 15mh leakage) the result is the same. Jim
Aaah. I misunderstood one of your previous posts. I thought you had said you were reflecting the secondary inductance to the primary. I jumped to the conclusion that you were double-dipping on your modeling of the secondary inductance; having it modeled on the primary and secondary sides of the transformer. In fact you were just calculating the value of the primary inductance based on the known secondary inductance and the square of the turns ratio.

You are probably aware that the specified 720H(and calculated 288mH primary inductance) are the maximum values when the core permeability has reached its peak just before the onset of saturation. At lower input voltages, the inductances values will be much lower. Even so, those are pretty high inductance values, which hints at a very high quality grain oriented core material.

Last edited by bolserst; 6th May 2010 at 02:54 PM.

diyAudio Member

Join Date: Jun 2005
Location: Fort Worth, Texas
Quote:
 Originally Posted by jelanier I would recommend Terman's engineering handbook if you wish to gain more knowledge on winding transformers.
Thanks for the book recommendation. I will have to check it out.
Most of my transformer design knowledge has come from the "Radiotron Designers Handbook (4th edition)" and more recently "Transformer and Inductor Design Handbook (3rd edition)"

For those who were not aware, the copyright on the 4th edition Radiotron Designers Handbook has expired.
You can download free copies several places on the web. Here is one:
RDH4 mirror

 6th May 2010, 04:18 PM #133 diyAudio Member   Join Date: Apr 2010 I think Terman's Radio Engineering Handbook is available free download as well. I have a really nice hard cover version Jim
 6th May 2010, 07:15 PM #134 diyAudio Member   Join Date: Feb 2008 Location: Jackson,michigan I just got playing with circuitmaker 2000 and using your values in a single ended configuration I got similar results. Now all I have to do is figure out my real world parameters of my core and different configurations and plug them in and then I can see whats really happening. Thanks again Jim and Steve! This helps me out tremendously. I super appereciate your help,you don't even know how much! jer
 8th May 2010, 12:58 PM #135 diyAudio Member   Join Date: Apr 2010 Transformer Measurements The major parameters that are needed for tuning the Q of the high frequency resonance are: Stray Capacitance, Leakage Inductance and Winding Resistance. The measurement system I use is not conventional, but is more useful in determining these values in the frequency range and loading in which we are interested. The resonance is determined by the value of leakage reactance and total capacitance. The total capacitance is the value of internal stray capacitance plus the external capacitance which will ultimately be the ESL load. The frequency of resonance is: F=1/(2pi(LC)^.5) where C= Cstray+Cext We will make 2 frequency measurements. Each measurement is made with a different value of external capacitance. From this we can determine the product LC. It is then possible to determine Cstray by substitution. L is then determined by having a known frequency, Cstray and Cext. Cext should be values near what will be used as the ESL load. The test setup has a 4 ohm resistor in series with the primary of the transformer. This limits the impedance as seen from the amplifier. The amplifier is driven by a signal generator. Set the gain so that the amplifier has an output of about 1 volt. The probe is then connected at the primary of the transformer as shown in the test circuit. We will find the frequency which has the minimum voltage by sweeping the generator. The frequency of minimum voltage indicates resonance. Resistance can be determined by direct DC resistance measurements of the primary and secondary windings. Remember that the effective resistance of the RLC secondary circuit is Rsec + Rprimary(Voltage Ratio)^2. One could also determine total effective resistance by measuring the magnitude of the voltage at the minimum and calculating the secondary value. The 4 ohm resistor and the resistance at resonance make up a voltage divider. Another way is to measure amplitude at key frequencies around the resonance to calculate the Q. I use a spreadsheet to do the LC calculations. Be sure to protect the transformer by either using a low power (1watt) 4 ohm resistor or series fusing in case the amplifier breaks into oscillation. Always be aware that the secondary voltages can be quite high. Take care not to make contact! Jim
diyAudio Member

Join Date: Jun 2005
Location: Fort Worth, Texas
Quote:
 Originally Posted by jelanier The major parameters that are needed for tuning the Q of the high frequency resonance are: Stray Capacitance, Leakage Inductance and Winding Resistance. The measurement system I use is not conventional, but is more useful in determining these values in the frequency range and loading in which we are interested. The resonance is determined by the value of leakage reactance and total capacitance. The total capacitance is the value of internal stray capacitance plus the external capacitance which will ultimately be the ESL load. The frequency of resonance is: F=1/(2pi(LC)^.5) where C= Cstray+Cext We will make 2 frequency measurements. Each measurement is made with a different value of external capacitance. From this we can determine the product LC. It is then possible to determine Cstray by substitution. L is then determined by having a known frequency, Cstray and Cext.
Hello Jim,

How many times do we get to say we used something we learned in highschool Algebra.
Jim's technique adds known external capacitive loads to the transformer to define two unique equations
which can then be used to solve for the two unknowns: Cstray & Lleakage.

I actually get a great deal of satisfaction(sad I know) being able to determine two of the most difficult to measure properties of a transformer using nothing other than a signal generator and a multimeter. No fancy scope, computer, or LCR meter.

If solving equations isn't your thing let me(or Jim) know and I can email you a spreadsheet to do the calculations for you.
Step by step solutions also available upon request.

Graphically what you are doing when you solve for two unknowns with two unique equations, you are solving for the point where the plotted curves of the two equations cross.

Plot #1
Plotting up the two unique equations for all possible values of Llkg and Cstray that fit the equations, and you get plot#1 below.
Notice the curves cross at the calculated solution for Cstray & Lleakage.
Notice also that the slopes of the two curves are very similar.

Plot #2
With such similar slopes to the curves, if there is just 2% error in the measured value for F2(32,147 instead of 31,517), the calculated value for Cstray = 882pF is now off by 26%.

It helps to understand the possible error band when using this technique

Obviously, this technique can't be used to determine what portion of Cstray is due to secondary winding capacitance, primary winding-to-secondary winding capacitance, and other stray capacitance, like to the core. This breakdown may be important if you are trying to optimize a transformer design.
Attached Images
 LC1.GIF (18.3 KB, 64 views) LC2.GIF (19.2 KB, 60 views)

Last edited by bolserst; 10th May 2010 at 04:14 AM.

diyAudio Member

Join Date: Jun 2005
Location: Fort Worth, Texas
Quote:
 Originally Posted by bolserst Plot #2 With such similar slopes to the curves, if there is just 2% error in the measured value for F2(32,147 instead of 31,517), the calculated value for Cstray = 882pF is now off by 26%. It helps to understand the possible error band when using this technique
Was tired last night and forgot to mention that in the past I had always gotten better results using Cext = 0 pF (no added load capacitance) for one of the equations.

Here are Plot #1 and Plot #2 replotted with Cext = 0pF for the 1st case instead of 500pF.
Note that the slopes of the curves show a bit more separation than in the last post.
The result is that 2% error in the determination of F for the 2nd case results in only 7% error in the calculated value of Cstray; a nice improvement over 26% from the previous example.
Attached Images
 LC3.GIF (28.2 KB, 48 views) LC4.GIF (29.5 KB, 37 views)

 11th May 2010, 11:36 AM #138 diyAudio Member   Join Date: Apr 2010 It used to do that by using Cext as zero for one of the frequencies. It turns out that numbers don't compute with some transformers because leakage inductance changes with frequency. That is, coupling changes with frequency. For that reason, I have been using values that are closer to the load Cesl. Maybe use one Cext above and one below. Leakage inductance is a fictitious inductance that is a function of coupling. The core acts differently at different frequencies. As an example, if I remove the .015H leakage inductance and replace the K_linear coupling value with 0.99998958349609092434983098030463 instead of 1, the results are the same. (720/(720+.015))^.5=0.99998958349609092434983098030463 I have had some transformers that were so bad, that I simply measured Q from key frequencies with Cext=Cesl and tuned as needed.
 11th May 2010, 01:44 PM #139 diyAudio Member   Join Date: Aug 2008 Time to back up with a simple question. The spacer used will yield a d/s less than the thickness of the spacer because of the height of the wire. When someone is claiming a separation of say 2mm. are you using a spacer thicker than 2mm. by the amont of the thickness of the wire? Is there a common convention on what everyone is sayin? Paul
diyAudio Member

Join Date: Jun 2005
Location: Fort Worth, Texas
Quote:
 Originally Posted by pforeman Time to back up with a simple question. The spacer used will yield a d/s less than the thickness of the spacer because of the height of the wire. When someone is claiming a separation of say 2mm. are you using a spacer thicker than 2mm. by the amont of the thickness of the wire? Is there a common convention on what everyone is sayin? Paul
Hi Paul,

For comparison between ESLs, the d/s spacing is the distance between diaphragm and outer surface of stator insulation.
With perforated sheet metal stators this is usually just equal to the thickness of the spacer plus any adhesive used.
As you correctly pointed out, for most wire ESLs the spacers used are thicker than the desired d/s spacing by OD of the wire.

In attached pic from the Acoustat white paper d1 = d/s spacing, and d2 = spacer thickness.
Attached Images
 ESL_spacing2.gif (24.5 KB, 79 views)

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is OffTrackbacks are Off Pingbacks are Off Refbacks are Off Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post Paswa Chip Amps 17 17th February 2013 10:00 AM kalmara Solid State 23 20th May 2008 12:36 PM C Schelp Multi-Way 24 8th September 2002 03:39 AM

 New To Site? Need Help?

All times are GMT. The time now is 12:34 AM.