Anyone have SPICE models for PSU Toroidal Xfmr ??

Status
Not open for further replies.
Re: MEASURING AND SPICE-MODELING A POWER TRANSFORMER

gootee said:
Transformer modeled in this example: Triad Magnetics VPP36-1560 (same as Hammond 183K36), 56VA, dual 115V primaries, dual secondaries (36VCT @ 1.56A or 18V @ 3.12A), PCB mounted.

My split-bobbin transformer has dual primaries and dual secondaries. I was under the impression that I could treat them as two separate single-pri/single-sec transformers, for modeling (and measuring) purposes. But now I'm not sure if it can be modeled well, that way, and, even if so, I don't know if my measurement methods were valid (probably not), especially for the leakage inductance, since I only shorted ONE of the secondaries, to measure what I thought was the "corresponding" primary.

Hi Tom,

It sounds like you'll want to hook both the primaries and secondaries in parallel for the measurement. I'm assuming the primaries will be hooked up in parallel in actual use anyway per the standard use of international transformers. So it seems that the primary inductance you measure this way will be the correct one. But even if the secondaries are hooked up in series in the actual design, it seems to me that, as long as they're identical, you could still get the secondary inductance from the paralleled case. Then you could just double the measured paralleled value to get the individual secondary inductance.
 
(CORRECTED) POWER TRANSFORMER LT-SPICE MODEL (ATTACHED)

Everyone,

ATTACHED is a new version of the LT-Spice Power Transformer Model.

It is a corrected and improved version of the one that I had attached to Post #34, of this thread.

This newer model requires ONLY the voltage and resistance measurements (and the frequency) to be entered, and calculates everything else, automagically.

It also includes instructions for making the measurements.

Thanks again, andy_c and powerbecker, for doing most of the work, for this power transfomer modeling project.

----------------------

LIST OF CHANGES

from the previous LT-Spice power transformer model (from Post #34, which was named "pwr_xfmr.asc.txt"):

- Added more .param statements to do calculations, so only actual measurements need to be entered, and everything else is calculated automatically.

- Added instructions for taking measurements. (NOTE: The procedures are very dangerous, and possibly wrong, and could kill you! Use them at your own risk!)

- Used new measurements for the included "example" 50VA transformer's parameters, taken with both primaries and secondaries in parallel, and without as many blunders as before, I hope. (The leakage inductance is now more reasonable, at .08835 H, and primary inductance is 6.258 H.)

- Changed AC Mains' impedance model's inductance from 800 uH to 400 uH, hoping that that might better-model it for 115V/60Hz. Use it at your own risk.

- Corrected a syntax error in the param statement that calculated the coupling factor, k. (I don't think the error caused any ill effect, before, though.)

------------------ END OF CHANGES LIST

Remove the ".txt" from the end of attached file's filename, in order to use it with LT-Spice (which is free, from http://www.linear.com ).


- Tom Gootee

http://www.fullnet.com/~tomg/index.html

-
 

Attachments

Re: Re: MEASURING AND SPICE-MODELING A POWER TRANSFORMER

andy_c said:




Hi Tom,

It sounds like you'll want to hook both the primaries and secondaries in parallel for the measurement. I'm assuming the primaries will be hooked up in parallel in actual use anyway per the standard use of international transformers. So it seems that the primary inductance you measure this way will be the correct one. But even if the secondaries are hooked up in series in the actual design, it seems to me that, as long as they're identical, you could still get the secondary inductance from the paralleled case. Then you could just double the measured paralleled value to get the individual secondary inductance.


Andy,

I did all of the measurements again, with both primaries and both secondaries in parallel, which would be the case when it's used with 115V AC (for my application, at least), which should give the rated 18V @ 3.12A from the secondaries. (I attached the updated and improved LT-Spice model to my previous post.)

When used with 230V AC, the mains voltage is applied across the series primaries, and the secondaries are still in parallel, so that the secondaries will still give the rated 18V @ 3.12A. In my units, I use a hefty DPDT slide switch, for selecting 115V or 230V, which reconfigures the primary connections.

ASSUMING, for the moment, that the present model is more-or-less valid, for 115v across the paralleled primaries, then it seems like it should also be valid with 230V across series primaries, since that would be basically equivalent to using two 115v sources with one across each series primary, which would be basically equivalent to using one 115v source across the primaries in parallel.

At any rate, the updated LT-Spice model looks MUCH better than the previous one, when simulated with my power supply. At least it no-longer appears to be grossly-underpowered. Everything is much more in-line with what I was expecting, now.

It's also easier to use, now that you ONLY need to enter the few measured voltages and resistances from the transformer testing. There is one sort-of-minor thing that I'd still like to try to do with it, though, and that is to find out if I can somehow display the values of the auto-calculated parameters.

- Tom Gootee

http://www.fullnet.com/~tomg/index.html

-
 
Hi Tom,

Thank you very much for your efforts. You're the one that's done the bulk of the work here. I do appreciate it.

It looks like the transformers for my power amp won't be in until the end of the month. When I get them, I'm going to do your procedure to characterize the leakage inductances and the rest of the parameters. Then I want to design a snubber like the one in Jim Hagerman's article to minimize radiated and conducted RF.

Thanks again!
 
andy_c said:
Hi Tom,

Thank you very much for your efforts. You're the one that's done the bulk of the work here. I do appreciate it.

It looks like the transformers for my power amp won't be in until the end of the month. When I get them, I'm going to do your procedure to characterize the leakage inductances and the rest of the parameters. Then I want to design a snubber like the one in Jim Hagerman's article to minimize radiated and conducted RF.

Thanks again!

Thanks, Andy. I hope your transformers hurry-up and get there, for you.

I did a little research on snubbers, a while back. So I might have some references that you will be interested in:

There's a good paper by a guy from Cornell Dubilier (capacitors, etc), Rudy Severn, called "Design of Snubbers for Power Circuits", which is linked-to at the following link, near the big, red "CDE" logo, about a third of the way down the (long) page:

http://archive.chipcenter.com/circuitcellar/november00/c1100rp58.htm

That link (above) also has some other links to snubber design information.

The following page is extremely interesting (but not mainly for its snubber content, probably):

http://www2.tech.purdue.edu/ecet/courses/eet257/

And here's a short and to-the-point snubber piece:

http://www.ridleyengineering.com/snubber.htm

And yet-another snubber-design paper:

http://www.maxim-ic.com/appnotes.cfm/appnote_number/3835

Hmmm...! Just noticed, in that first link I gave... Maybe I should use one of these in a gainclone power supply:

"General Electric Industrial Systems has capacitors and Surge Protection Devices the sizes of railroad box cars." 😱

There's also this:

http://www.maxim-ic.com/an907

And there's a very interesting paper at calex.com, called "Understanding Power Impedance Supply for
Optimum Decoupling". Sorry, I don't seem to have a direct link.

I hope that one or more of those is helpful.

I only really "absolutely HAD to" use a snubber across the main switch diode, in my 60 kHz boost-mode SMPS. Without a snubber, it looked almost like an HF oscillator for half of every cycle. I currently have 470 Ohms and .0022uF in series across it, and have not even a hint of oscillation.

Oh, regarding causes and solutions for RFI and EMI in general, there's some very good information, here:

http://www.national.com/nationaledge/mar04/article.html

And this one isn't about snubbers, per se, but is interesting (I haven't tried it, yet):

http://www.wenzel.com/documents/finesse.html

Well, I've got lots more. But that's probably enough to wade through, for quite a while.

Good luck!

- Tom Gootee

http://www.fullnet.com/~tomg/index.html

-
 
powerbecker said:
Hello Tom,

thank you for your profound work!

When it later come to the rectifier may be this thread with simulations and measurements can be interesting :

http://www.diyaudio.com/forums/showthread.php?postid=752444#post752444

Heinz!

Thanks, so much, Heinz. But it would probably NOT have gotten this far (not THIS year, anyway), without _your_ help and guidance. I really appreciate it, very, very much. So, thank you, again.

And thanks for the link to that rectifier/snubber thread. Up until now, I have paid attention only to the big schottky diode at the output for the switch & inductor, in my boost-mode SMPS, which definitely DID need a snubber. But now, with the new model for the transformer placed upstream of my rectifier bridge, maybe I can use spice to investigate any oscillation tendencies, there, too.

In fact, I just now AM trying it, with my LT-Spice "more-complete power supply" simulation model. And I actually DO SEE some oscillations in the rectifier bridge currents, already, even during the first few cycles. I did not notice anything, there, when looking at a working unit, with a scope, in past testing. But I wasn't looking for it, there, either. I will investigate this, further.

Thanks again!

Regards,

Tom Gootee

http://www.fullnet.com/~tomg/index.html

-
 
Re: (CORRECTED) POWER TRANSFORMER LT-SPICE MODEL (ATTACHED)

gootee said:

Remove the ".txt" from the end of attached file's filename, in order to use it with LT-Spice (which is free, from http://www.linear.com ).


- Tom Gootee

http://www.fullnet.com/~tomg/index.html

-

With the greatest of respect, I am not sure how to do this. I have the LT Spice Program, but I am unfamiliar with how to remove the ".txt" from this file.

When I click on the attachment, I get a download window which asks me if I want to "Open" or "Save", it does not seem to have the provision to remove the ".txt" at the end.

I tried saving it to program files, but then it didn't list the ".txt" at the end either so I could put it into LTSpice. If I put it directly into LTSpice, I get a text file.

I'm sure the procedure is quick and simple, but I am missing it somehow. Any help would be appreciated.


Here is the url of the attachment in question, I got it from clicking on Properties in the link in post 42.

http://www.diyaudio.com/forums/attachment.php?postid=1153560
 
Got it!

When I saved the file to Program Files, I had a choice to save it as a "txt" file or under "All files". The default was to save as a "txt" files. I moved the selector to "All Files", got a schematic file in Program Files, and then just moved the schematic over to LT Spice files.
 
Status
Not open for further replies.