Tanks, impedance, the j operator and excel

May 26, 2008 14 Replies

Hi gang, This one is directed at you math whizzes..... I have a formula for tank circuit impedance that I am attempting to analyze/plot in MS Excel. The tank consists of a resistor in series with the coil, a cap in parallel with the coil resistor combination and a another resistor in parallel with the cap and coil/resistor. The formula consists of two quotients; One has a numerator that consists of several terms multiplied, added, squared, etc and the denominator is similar. The second quotient is preceded by the "j" operator. (That is the entire quotient appears to be "j"ed rather than any single term or factor). This quotient has a numerator and denominator similar to the first). I assume that the first quotient is the real part of the impedance, while the second one is the imaginary part, yielding the rectangular Z =R +/-j form. Does this sound correct? Assuming the math is done correctly, I believe I can use the IMABS function on the result to get the magnitude of the impedance. Yes? In Excel, I have entered the first quotient as ordinary numbers, i.e. "=A1*(C2+B3)..../D4*6.28*C3......" I'm getting numbers that seem to make sense. The second quotient has me scratching my head a bit. I have tried using the same process on this part, and then using "=IMABS(first quotient results,+second quotient results)" to get the magnitude of the impedance but I think this may be incorrect. Any thoughts? If someone wants it, I can post my Excel file to them.....



Thanks



Bob WB0POQ


nk

f
.

le

Yes, sounds right.

n

formatting link

It appears that, if you want to form a complex number from its real and imaginary components, you need to synthesize the complex number with the COMPLEX function before given that synthesized number to IMABS.

e

ut

You probably already know that it is not necessary to use IMABS or COMPLEX.

Just plug in omega for first quotient, omega for second quotient. You will get two numbers, one for each quotient, X, and Y, respectively Compute the total impedance Z from them using Pythagorean theorem.

Z =3D sqrt (X^2+Y^2).

Naturally, you ignore the j on second quotient.

-Le Chaud Lapin-

According to my Excel help, the syntax for the function is =IMABS("5+12i") that is, the quotes go around the argument of the function, and there is a plus sign, not a comma, linking the real and imaginary parts.

John

I don't think you are using the correct abbreviation or 'wizard' here. But then again, this might be who you want. ;-)

Paul Hovnanian paul@hovnanian.com ----------------------------------------------------------------------- Have gnu, will travel.

You should post the spreadsheet workbook up in alt.binaries.schematics.electronic

Then, we can all have a look at the real thing.

I am pretty good with excel, and this sounds pretty simple.

If you just post the entire formula for each element of your equation, it can be worked out as well.

A diagram of the tank circuit will work as well.

Sorry....my frustration is showing.... Any other thoughts of where I might upload the file?

In any event, thanks for the replys.

Ok....here is the formula:

Z = Ry(RxRy+Rx2+w2L2) wRy2(L-CRx2-w2L2C) --------------------------------------------------- +j -------------------------------------------------- (Rx+Ry-w2LCRy)2+(wL+wCRxRy)2 (Rx+Ry-w2LCRy)2+(wL+wCRxRy)2

Rx is the resistor in series with the coil (represents coil losses) L is the coil C is the capacitor in parallel with the coil/resistor network Ry is the resistor in parallel with the capacitor coil/resistor combination w is 2*PI*F

Sample values: Rx = 0.2836 Ohms L = 0.868uH C = 335.2pF Ry = 11 k Ohms Fr = 9.336MHz

Just great.....the formatting is all screwed up. I give up. I just have to keep learning over and over what a pain in the @## these %#*&$#& computers are.

Hi Bob,

You have to understand that this problem is so trivial to most people in this group, we are having trouble figuring out what it is that you do not understand, so it might help us to tell us what it is that you do not understand. :)

You have an "equation" clump. One clump on the left, one clump on the right. The clump on the left has bunch of R's, C's, etc. in it, and so does clump on the left, the only difference being that the clump on the left does not have a j attached to it like the clump on the right.

What you need to do:

  1. Take the clump on the left, the one without the j. Plug in numbers you have above into this clump on the left. When you are done, you will get a number. Call that number X.

  1. Take the clump on the right, the one with the j. Plug in the numbers you have above, again, into this clump on the right. When you are done, you will get another number. Call this second number Y.

Find the value of sqrt(X * X + Y * Y). This is your answer Z. Do you have trouble computing X*X + Y*Y?

You should ignore the fact that this problem is related to electronics or complex numbers. At this stage, it is really a "put the key in the keyhole and turn" situation.

I think the first thing you should do is determine whether you want to:

  1. Understand what you are doing.
  2. Get the answer.
  3. Both.

We will be more than happy to help you with 1, 2, or 3, but you should decide first. If the answer is 2, and 2 only, just plug in the values you have as I described above and wait for Z to pop out. :)

-Le Chaud Lapin-

What is your Usenet newsgroup server? Frontiernet's servers drop all posts over about 15k bytes (it seems "reasonable" since they don't support binary groups, but this is undocumented by Frontiernet, and there's at least one newsgroup FAQ that I never saw because it was larger than that! It took a while to figure out I was missing the FAQ posts, then a little more poking around to figure out WHY... I haven't tested ETC yet, I'm just surprised they have a Usenet server).

Anyway, sorry, this was supposed to be about you...

There are plenty of free websites such as the old geocities and angelfire, but they've gotten a reputation of having lots of pages full of viruses and such.

As another alternative, Excel can save a spreadsheet as a .csv ("comma separated variables") file, which is just plain ASCII, and can be copied and pasted into a text-only post right here in this thread. That won't show us the graphs, but you can tell us the steps you go through to get them.

Have you tried this in LTSpice or another simulator? Do you want the "solution" to this circuit, or are you doing this as an exercise to learn the math? Regardless, you may want to do it in a simulator - it'll make it more satisfying when you finally make the exact same graph in Excel.

When posting a binary, it has to be pre-processed into a text code, and each "message" then has to be less than 10,000 lines in size.

You cannot simply post the raw thing.

The encoder is called "uuencode" and uudecode". Another is called yENC, but many old fogies here won't like that post.

Most news clients will transform a binary file for you on the fly.

A notice that you are using Outhouse Express... Not to be jumping on you or anything, but you should really use a real news reader client.

Also, posting a schematic (ASCII) of the tank, and the formulas you use would allow one to work on it.

Even a few of your cell formulae would help us to help you.

Naaaahhh... all one has to do is copy, then paste the line back together in a text editor. No problem. It looks like there is only one line that did it.

The only other thing I can think of is... what cell formula(s) are you using to arrive at your conclusion?

Are you graphing a series where you bump the value of one element a bit several times, and graph the impedance through those element value choices?

You could graph the effect of each element.

I made a post up in abse, and it is a 27kB file. Your client should be able to grab it just fine.

Dont post the formula you used in excel, with that damned "j operator". Post the actual electronic formula you are using, parasitics and all.

Or at least give a link to a site that has the formula you mention. All the tank circuit formulae I have found on the web do not include parasitic elements.

Perhaps there is a good simulator model that has it.

Anyone?

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required