Leakage Inductance - Page 4 - diyAudio
Go Back   Home > Forums > Design & Build > Equipment & Tools
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Equipment & Tools From test equipment to hand tools

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
Thread Tools Search this Thread
Old 2nd March 2013, 02:01 PM   #31
diyAudio Member
Join Date: Aug 2012
Hi again,

just made a measurement of the secondary inductance using the following methods/values:

-Cx value of 0.033uF
-both secondaries connected in series
-signal generator voltage of 1v
-signal generator freq.@ resonance: 1268 Hz
-primaries opened

Calc. result of: 0.477405938 H

Seems to agree quite nicely with his corresponding meas. of 0.42H per winding! However that value was det. from his by dividing his parallel measurements by 4.

What i did differently:
-connecting both secondaries in series
-setting the sig. generators voltage to about 1v

What is a little strange is that his measurements were for both secondaries, e.g. why didn't i calc. 4x 0.47 or 1.9096!?

Instead i got directly the value per winding??

So the mystery continues.....

  Reply With Quote
Old 3rd March 2013, 01:43 PM   #32
diyAudio Member
Join Date: Aug 2012
Can anyone familiar with Hagerman's snubber Paper verify the logic of my calculations here?

I'm a little confused about how to calculate the snubber capacitor Cs and the snubber resistor Rs, when using the "extra" Cx cap.

1. Rs = sqrt(Lt / Ceq),
where Lt = sec. leakage inductance. For my transformer, I'm using 74.01 uH

2. Ceq = Ct + Cd + Cx,
where Ct = sec. interwinding capacitance, Cd diode capacitance and Cx, snubber capacitor.

3. Cs = 2 * pie * sqrt(Lt * Ceq) / Rs

According to Hagerman, if Cx is much much larger than Ct + Cd, Ceq = Cx. For my Cx value of 0.1uF this should hold, since Cd is approx. 50 pF and Ct for my transformer is about 3578 pF

If i put in my values, i get this:

Rs = 27.20 ohms
Cs = 0.628 uF
Cx = 0.1 uF

So i would use Cx across the supply rails followed by Rs + Cs in series across the supply rails.

  Reply With Quote
Old 3rd March 2013, 04:20 PM   #33
diyAudio Member
Join Date: May 2011
Location: Silicon Valley
You could verify it yourself using LTSPICE. You could simulate the transformer + rectifier + filter + load, twice: once without the snubber, and again with the snubber. Then you could eyeball the waveforms and ask yourself, does this look OK?

Of course, simulations can fall victim to the garbage-in, garbage-out syndrome. If the circuit you simulate is incorrect, the simulation result will be incorrect too. In this case, if the transformer measurements were incorrect, or if you have interpreted the measurements incorrectly, or if you incorrectly translate them into a circuit model for simulation, then you've got garbage-in. You will naturally receive garbage-out.

I've attached a simulation schematic which shows one way you can test your snubber in LTSPICE. I deliberately left out the hard part -- interpreting the measurements and forming a simulation model -- so as not to rob you of the opportunity to learn by doing, and to learn more by doing-over. Fill in the yellow box and simulate. You may get results that match mine, which are also attached.

However, keep several points in mind: (1) the measured data might be WRONG. (2) your interpretation of the measured data might be WRONG. my interpretation might be wrong too. (3) when you convert the measured data into an LTSPICE simulation model, you might do that WRONG. so might I.

If any link in the chain is WRONG, from measurements, to interpretation of measurements, to circuit modeling; then you've given garbage-in to LTSPICE. You should expect garbage-out.

Also, don't neglect the possibility that I and others on this website may be agents of Satan, placed on earth to mislead you with falsehoods, half-truths, irrelevancies, and non sequitirs. Don't blindly trust everything you see on websites.
Attached Images
File Type: png schema.png (14.5 KB, 181 views)
File Type: png without.png (17.0 KB, 176 views)
File Type: png with.png (17.0 KB, 172 views)
  Reply With Quote
Old 3rd March 2013, 06:26 PM   #34
diyAudio Member
Join Date: Aug 2012
Thanx again transistormarkj.

Do I really sound that gullible?

I'm just a novice whose interested in deepening his electronics knowledge.

For all i know, you could be some evil sales man wanting to sell me stuff i don't need like the wrong capacitors for my snubber network....

Looking at Hagerman's equivalent circuit, i would fill in your little yellow box with.
-an inductor with my leakage value
-a cap in series with it using my interwinding cap.

I will def. try that out.

Btw. your results with my snubber values look quite ok, assuming you modeled/ interp. everything correctly.

  Reply With Quote
Old 11th March 2013, 04:51 PM   #35
diyAudio Member
Join Date: Aug 2012
I would like to measure the ringing across my power supply diodes using the test setup proposed in Bob Cordell's Audio Amplifiers Book. For the test setup circuit, see the attached schematic.

Could anyone familiar with that method, explain to me what is meant by that earth connection, which i have circled.

If i were to connect it to GND, this would be directly shorting the positive rail and would end up cooking my diode(s).

Is this earth simply the earth on the pwr chord?

Attached Images
File Type: jpg bob_cordells_diode_perf_meas_setup.jpg (42.6 KB, 155 views)
  Reply With Quote
Old 11th March 2013, 04:59 PM   #36
DF96 is offline DF96  England
diyAudio Member
Join Date: May 2007
To measure the voltage across a component using a 'scope, it can be very convenient to ground one end of the component.

A PSU is normally grounded at the output end of the 0V rail. Provided the whole circuit is otherwise floating then, apart from stray capacitance issues, it can be grounded anywhere. Remember, the 0V rail is only 0V because someone decided to ground it. Apart from that, there is nothing 0V about it; it is merely Vrail above the 'negative' rail and Vrail below the 'positive' rail. Voltage always is a comparison between two points.
  Reply With Quote
Old 11th March 2013, 11:46 PM   #37
diyAudio Member
Join Date: May 2011
Location: Silicon Valley
This little test jig is adapted from Bob Cordell's Figure 16.10 on page 353. It lets me observe oscillatory ringing of the power transformer's secondary, and to observe the damping effects of the snubbing network (Rs, Cs, Cx) as I adjust Rs. I like to dial Rs until I get a damping factor (zeta) of approximately 0.8. Then I simply turn off the power, remove the (socketed) trimpot from the jig, and measure Rs on my ohmmeter. Snubber design for this transformer is now complete! Without performing a single inductance or capacitance measurement.

The snubber's parallel capacitor Cx swamps out the capacitance of the transformer secondary, plus the junction capacitance of any rectifier I could possibly select, so the transformer + snubber work together well in the final application, even with different layout and lead lengths. Even with different rectifiers.

I like to use extremely "slow" (long tRR) diodes in the jig; they seem to give the largest amplitude of oscillatory ringing and the longest duration of ringing.

Sharp eyed readers may note that I've modified the output network (R1, D5, D6) a little bit, compared to Bob Cordell's original. I'm using two low capacitance signal diodes instead of one high capacitance rectifier. I've also cut the resistance by 10X, to be less sensitive to fixture and probe capacitance. The good news is, you can simulate all of this in LTSPICE and customize the output network however YOU like.

To avoid overheating the 30 ohm, 50 watt load resistor, keep the transformer secondary voltage below { sqrt(30*50)/sqrt(2) } = 27.4 volts RMS. If your secondary voltage is higher, either feed the primary from a variac, or change to a different load resistor arrangement. Also be mindful of the filter capacitor voltage rating.

I chose this particular load value so I could measure an Antek AS3218, which is a 300VA toroid with two 18 volt secondaries. It runs the secondary at about 0.8 amps, plenty enough to get the diodes switching on and off hard, so they'll sing their RF tune.

In case anyone wondered, "Dprot" is supposed to protect the surprisingly expensive, high ripple-current, low ESR filter capacitor from reverse bias, in case I accidentally install one or more of the bridge diodes backwards. I'd rather blow the fuse than explode a large electrolytic capacitor.
Attached Images
File Type: png jig_schem.png (12.8 KB, 159 views)
  Reply With Quote


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

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 Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

Similar Threads
Thread Thread Starter Forum Replies Last Post
Leakage inductance value is realistic? treez Power Supplies 3 17th September 2012 07:51 AM
How to measure leakage inductance percy Power Supplies 17 29th August 2008 09:57 PM
Electrolytic leakage lndm Power Supplies 5 26th July 2006 07:17 AM
Designing transformer leakage inductance DaBit Power Supplies 3 17th January 2006 08:08 PM
Transistor leakage? Solid Snake Solid State 1 1st October 2003 08:48 AM

New To Site? Need Help?

All times are GMT. The time now is 09:06 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2