Simulation software Pspice etc.

[rmeyn]

I don't know if there is much interest out there with this forum's members, or perhaps with new members, in a thread discussing successes with simulation programs. I use a student version of Pspice. In the beginning I found it difficult to learn. With perserverance I am amazed how it has helped me to design circuits and to analyze circuits already published. I am not an expert, but I can share some of the experiences I've had with those who are interested....
Richard

[lamazoid]

Yep, it's interesting. Would you give us some example of your simulations?

[rmeyn]

I think the best thing for you to do is first download a copy of the student version of Pspice. try www.electronics-lab.com/downloads/schematic/013/ The download button is near the bottom of that page. Then I would recommend you try a simple tutorial like the following tuttle.merc.iastate.edu/ee203/class/notes/spice/diodetutorial.pdf You will be amazed what you can simulate if you begin by studying tutorials and demonstrating the same results. For me to show my simulations will be difficult if you do not have a version of Pspice already installed. My latest simulation was a series regulated 9VDC power supply. Years ago I built such a supply to power my experimental circuits. (Mine actually was a variable 12VDC). It works perfectly and I always wanted to know how it worked. I simulated it with Pspice and I have a much better understanding of how it works and how to troubleshoot it when it does not. Good luck and keep us informed. Richard p.s. It is perfectly legal to download this software. Orcad, the company distributing it, encourages the sharing of this file. They know someday you, or one of your buddies will purchase the full blown version.

[lamazoid]

I'm not interesting in actual understanding of this exact software. I have many similar programs: CircuitMaker, SuperSpice, Multisim 9, MicroCap and others. Here, in Russia, we dont worry about is it legal or not, because it's all illegal anyway ;-) I have bought a dvd with a bunch of industry-leading electronics software for only ~$10 and i'm really happy

The thing that matters is WHAT you are simulating and HOW you do it. I say regulated power supply is easy stuff, it does not need real simulation. It's well described in many textbooks for beginners.
The REAL troubles come when you're trying to simulate oscillators, transformers and other high-frequency analogue stuff. In simple programs like PSpice you can't just draw the circuit and see how it works in a right way. You should first take into account physical model of each component, its properties, the differencies between ideal and real component. And you need all this stuff in your BRAIN , not a book or something, so you must understand the theory of operation BEFORE you try to simulate. I think that's not very suitable for educational purposes. Modern systems like Multisim 9 partially solved this problem, but its still far from perfect....

[Gary Lecomte]

lamazoid said:

I'm not interesting in actual understanding of this exact software. I have many similar programs: CircuitMaker, SuperSpice, Multisim 9, MicroCap and others. Here, in Russia, we dont worry about is it legal or not, because it's all illegal anyway ;-) I have bought a dvd with a bunch of industry-leading electronics software for only ~$10 and i'm really happy

The thing that matters is WHAT you are simulating and HOW you do it. I say regulated power supply is easy stuff, it does not need real simulation. It's well described in many textbooks for beginners.
The REAL troubles come when you're trying to simulate oscillators, transformers and other high-frequency analogue stuff. In simple programs like PSpice you can't just draw the circuit and see how it works in a right way. You should first take into account physical model of each component, its properties, the differencies between ideal and real component. And you need all this stuff in your BRAIN , not a book or something, so you must understand the theory of operation BEFORE you try to simulate. I think that's not very suitable for educational purposes. Modern systems like Multisim 9 partially solved this problem, but its still far from perfect....

I Totally Agree. They are FAR FROM PERFECT.
Too many times there are a lot of circuits that will work in Simulation, but Not in real life or Visa Versa.

Most often, For the time it takes for me to enter it for simulation, I can Breadboard it using real parts and do real tests on it.

A Few Years ago, I bought the Full Version of "Electronic WorkBench".
What a Bunch of CRAP that one is. And for the $1000.00 it Cost, I can buy a LOT of Parts to make real operational circuits.

[rmeyn]

Well boys, I never said they are perfect! And Iamazoid never indicated whether he was a beginner or experienced engineer. I gave him a simple simulation to help in learning. Back in the 1970's when I was doing my Phd work I was involved in computer ecosystem modeling. There was an expression back then that applies now and applies to electronic modelling... "Garbage in Garbage Out" Pspice, Spice, or any other simulation software requires that you have knowledge of the circuit, what components to use, and most of all--an accurate component model! If you blindly select components that you think will work good enough , you will get crap out! 10,000 design engineers, 100's of engineering schools, NASA, and others can not all be wrong. Pspice is NOT a simple program! It is incredibly powerful when used properly. The simulation of oscillators is one application both of you (Gary and Imazoid) should take a second look at. The beauty of Pspice is that you can explore the limits under which the oscillator will operate--and far more quickly and easily than breadboarding alone. I will give one "simple" example. In my experimenting with a water controlled switch, either with an IC switch or transistor switch, I found through simulation that the circuit would only operation with the resistance of water in a certain range, depending upon temperature, salinity, and distance between probes. When I breadboarded my design and tested it--VIOLA! Pspice was correct. I could have spent many hours with different wiring setups to find this out. And Gary, did you check out computer simulation software before you bought it? I'm sure the University of British Columbia would have given recommendations. That's all for now. Richard

[rmeyn]

And Imazoid. Why try to simulate transformers? It can be done, but you know from the manufactures label what the output is! Forget simulation, just use a Voltage source such as VAC and define the parameters, sine wave, frequency, peak to peak voltage etc. Richard

[Gary Lecomte]

rmeyn, At the time, Electronic Workbench was the Leading company in this field.

What was truly disapointing was Even there Support Staff couldn't get the program to respond correctly on the Very Simple analogue circuits I gave them.
One year of Free Support for What?

As to the University of British Columbia, Not really impressed.
Same for BCIT.

Sorry but Can't go into details. Confidentially Problems.

I am Not saying all these programs are Useless.
"TI" has a very useful ones for Oscillators and op-amps. Both are Specialized programs for there products, so the models are More Precise.

Ethier way I prefer Hands on.
There are many things that simulation programs Can't solve or simulate in a practical way.
Besides, there is Much more to a Design than Just the Schematic.
(Heat Considerations, Crosstalk, Layout, and many others)
Wires in a Simulation don't Radiate Energy.

Anyway, I'm off to Vancouver tomorrow for a few days or so.

[rmeyn]

This post I offer, not necessarily for Gary, but for anyone interested. There is a lot of info out there on electronic circuit simulation, but the following is one I believe to the best for offering about as much information as one could desire! zone.ni.com/devzone/cda/tut/p/id/5414 It is provided by National Instruments. One feature is really like is a data base of manufacturers that provide spice models. Remember that spice models are not Pspice models, etc. Make sure you use the correct model for the software you are using.
There are those that do not think much of circuit simulation, but I offer a quote that says it better than I can. "Circuit simulation is an important part of any design process. By simulating your circuits, you can detect errors early in the process, and avoid costly and time consuming prototype reworking. You can also easily swap components to evaluate designs with different bills of materials (BOMs)" National Instruments/NI Developer Zone/Spice Overview Richard (rmeyn)

[lamazoid]

Well, let's take a real example. As you may know i am currently developing a low input voltage inverter. Here is the circuit which gives maximum efficency compared with others i tried:



I designed it, and i don't know is it correct or not. It provides about 30% efficency - not so bad.
The transformer wound on 14mm POT core. 1+1 turn of improvised litz wire 4x0.35mm. Operation frequency is ~180 Khz.
Secondary winding not shown, its configured as flyback converter to charge 0.47 microfarad 800v cap.

Waweform at the drain:



And gate:


In both cases peak amplitude is ~10 volts.

Well, can we now model this thing and optimize its perfomance? I have tried several programs with ABSOLUTELY NO RESULT. It simply refused to "work".

[rmeyn]

Imazoid, I am going to be out of town for a few days. It's Memorial Day Weekend here. I'm not trying to replace the moderator (he's much more qualified than I am), but I wonder about a number of things. The IRF7456 is a synchronous rectifier, and as such, has an integral diode body that is needed in the circuit. I don't see it on your diagram. Also, it looks to me that you need an oscillator in order for the circuit to operate at 880kz. I would like to give this circuit a try, but I would like to see it in better detail.
Transformers can be modeled, but I need to know the turns ratio of the primary and secondary windings. The IRF7456 should not be a problem. If there is no model for it, I can build one from its data sheet. I'm confident I can model this circuit, but it will take some time.
Richard M.

[Gary Lecomte]

Using a Typical Mosfet in EWB, I can't get it to Simulate Either.
And I don't have a IRF7456 to try to duplicate your circuit using real components.
The best I could do would be to just use an IRF640, but it has a higher ON Voltage. But Might work.
Lamazoid, Do you know the inductance's of your windings?
That would help.

Quote Rmeyn: IRF7456 is a synchronous rectifier.
** No it is NOT. That is Just a possible Application for this device.

Quote Rmeyn: Also, it looks to me that you need an oscillator in order for the circuit to operate at 880kz.
** Lamazoid's circuit forms an oscillator. It doesn't need another one.
** Also, Any Secondary winding would allow for a Test. The Turns ratio between the Primary and Secondary is Not important.

[lamazoid]

I have tested many mosfets and the best result was obtained with IRL2905. Any "IRL" should work, while typical IRF will not (i tried them too), until they are special low voltage types like 7456, 7457, 7319, 7401 etc. The main problem with these transistors is high input capacitance. I can't find appropriate IC to drive them, and those i have and tried worked very bad. The self oscillating circuit gives MUCH better waveforms compared to those ICs. Also i have noted that adding a large cap across the source increases perfomance. Efficency varies from 10% with no cap to about 30% with 1000mkf. But i can't use it in final device due to small dimensions required.

As to transformer, i don't know actual secondary turns. The inductance meter shows: 1 uH (micro) on each primary and 26mH (mili) on the secondary winding. Core assembled with no airgap.

Going back to the circuit, i don't understand how to adjust the duty cycle which is currently far from its optimum point of 50%, and is about 20%. That's the main reason to use simulations.

[Gary Lecomte]

I Tried a Quick built Circuit and transformer, using the mosfets I have, but no real luck. At one point it was oscillating, but very poorly.

Unfortunately where I live there are no suppliers within over 200 miles.

And just to order one of those low voltage types is far to expensive in shipping charges.

One other point is that IRF7456 has full turn On at 4.5 volts.
Your Supply is Only 3 Volts.
You Might try Increasing the number of Turns on the Feedback, to help increase that voltage.

I never looke at the Spec's on those other mosfets you mentioned.

And base on your circuit, the freqency of operation is probably a function of the Inductance of the Primary, But also the Capacitance of the Mosfet, creating an LC Circuit.

Sorry but can't help you much without those mosfets.

[lamazoid]

Gary, can you help with calculations of transformer for this circuit? I suspect transformer is the main problem of low efficency.