RC circuit differential equations

17 replies · 4K views · last reply
evol_w10lv
Messages
70
Reaction score
0

Homework Statement


Circuit schematic: a 2 A DC current source with a switch labelled tClose=0 feeds a node; R1 (2) shunts to ground, then C1 (1n) in series to a node with R2 (3) to ground, then R3 (4) to ground at the right.

Task is to write differential equation for this circuit.

Homework Equations




The Attempt at a Solution


I'll try to solve the task, but now I want to know, is it possible to use voltage source instead of current source. For example, I can calculate ekvivalent resistance R_1, R_2 an R_3 parallel. Then current source * ekvivalent (total) resistance = voltage source?
Like this:
Circuit schematic of a 4V source switched by tClose=0 feeding an RC network: resistors R1, R2 and R3 with capacitor C1 of 1n in series between the nodes, nodes numbered 1 to 4.

Don't look at numbers, they are random.. I just want to know about idea.
If it's possible, then I can use V(Thevenin) and R(Thevenin).

OR:
I can calculate R(Thevenin) = (R2*R3)/(R2+R3), I guess.
But I don't know, how to calculate I(Thevenin). Or I have to use Northon method, when we've got capacitor?

Can you suggest something?
 
In your original circuit the current source and first resistor form a Norton equivalent when the switch is closed. No reason why you couldn't turn that into a Thevenin equivalent.

Be sure to take into account the circuit properties affected by the switch if you need to determine the circuit operation if the switch is re-opened at some future time.
 
Do you mean like this:
Circuit schematic of a series RC network: 4V source V1, resistor R1 = 2, capacitor C1 = 1n, resistor R2 = 3 shunt across the node after C1, and resistor R3 = 4 to ground at the right end.
?

And then again I need to change circuit to Thevenin equivalent as I did before with RL circuits?
 
Last edited:
Yes and yes, in your figure you can absorb R2 and R3 into the Thevenin model.
 
Then:
R(Thevenin) = ((R2*R3)/(R2+R3)) + R1
E(Thevenin) = E = I*R1

(((R2*R3)/(R2+R3)) + R1)*C*dUc/dt + Uc = E

Is it correct?
 
Easier than I thought. But which initial condition I have to use? From picture in my first post?
But there is Ic(0), but here I need Uc(0). Seems to me, that I don't understand, how to find intial condition.
 
You can use the initial conditions for your "new" circuit. If the capacitor begins with no charge then its initial potential is zero. It will "look like" a short at time t = 0, so you can find the initial current, too.
 
Ic(0) = 0 and Uc(0)= E ?
Anyway, I'll try to finish all calculations tomorrow. It's to late today in my country.
 
evol_w10lv said:
Ic(0) = 0 and Uc(0)= E ?
Anyway, I'll try to finish all calculations tomorrow. It's to late today in my country.

With the Thevenin circuit, when the switch closes there will be a closed loop. The capacitor will look like a short, so just the Thevenin resistance is seen as the load at time t = 0.
 
It means that there is no voltage through the capacitor, so Uc(0)=0.
 
evol_w10lv said:
It means that there is no voltage through the capacitor, so Uc(0)=0.

"Across". Voltage is "across" a component, current goes "through". But yes, that's the right idea.
 
One difference between the two is that the current will be flowing in opposite directions and the potential across the capacitor will have the opposite polarity. To make them identical, change the polarity of either the current or voltage source.
 
Schematic of two circuits simulated in LTspice: each has an ideal DC current source IDC 0A in series with a 1k resistor, plus an inductor and switch; labels R1 to R6, L1 10uH, L2 10uH, U1 and U2 tClose=0.


How to redraw circuit here, when current source and resistor are in series? I guess, here we can't use Norton because resistor and current source not parallel.
 
evol_w10lv said:
How to redraw circuit here, when current source and resistor are in series? I guess, here we can't use Norton because resistor and current source not parallel.

Big tip: A resistor in series with an ideal current source has no effect on the circuit! The current source will produce its specified current no matter what is in its path. It won't even contribute to Norton/Thevenin resistances for the circuit.
 
  • Like
Likes   Reactions: 1 person
Tried to solve second one (from post #15) with two methods:
Circuit schematic with a current source labelled J, I1 = 0A, and IDC. A branch with R2 = 1k in series with L1 = 10uH carries current I_L, and a separate branch with R3 = 1k carries current I_2, both shown by large arrows pointing down.

Handwritten derivation of a circuit equation: J = I_L + I_2, I_2 = U_3/R_3 = (U_L + U_2)/R_3 = (U_L + I_L*R_2)/R_3, then J = L*(dI_L/dt)/R_3 + (1 + R2/R3) I_L, with I_L(0) = J.


And method 2:
Circuit schematic: current source I1 labelled IDC 0A in parallel with R3 1k, feeding series branch of R2 1k and L1 10uH on the right.

Then like before I used Norton/Thevenin equivalent:
Circuit schematic: voltage source V1 labelled 0V in series with resistor R3 (1k), resistor R2 (1k) and inductor L1 (10uH) in a single loop.

E = R3 * J
Schematic of a series RL circuit: voltage source V1 at 0V, resistor labelled 1k with R3+R2 beneath it, and inductor L1 labelled 10uH on the right side.

E(Thevenin) = E = R3*J
R(Thevenin) = R2 + R3

E = L*dIL + (R2+R3)*IL

Actually the first circuit (from post #15) is simular, but why there are switch without anything? Is there something tricky in the solution?
 
Last edited:
evol_w10lv said:
Actually the first circuit (from post #15) is simular, but why there are switch without anything? Is there something tricky in the solution?
The switch changes the circuit by eliminating R3 when it closes (it shorts it out). So there will be a different steady-state situation for each position of the switch.