# Obtaining dynamic resistance of a zener diode in LTspice Please left-click to enlarge

Above is a circuit where: D1 is a 10V(breakdown) zener diode, V1 is a voltage source which varies from 11V to 5OV DC and R1 is a 1k resistor. My aim is to identify the dynamic/bulk resistance of the zener diode by using LTspice.

I simulated the circuit such that 0 - 40mA current passes through the zener. The resulting V-I plot can be seen above.

Lets call Vz is the zener voltage and Iz is the zener current. What I uderstand from what I read is that, the dynamic/bulk resistance of the zener diode refers to the derivative dVz/dIz when Vz is above the knee voltage.

My questions are:

1-) Can I use this plot to obtain Vz/Iz which correcponds to the dynamic resistance? If so, is there a standard at which current the derivative corresponds to the bulk resistance?

2-) How can I plot the derivative of this plot i.e dVz/dIz in LTspice?

Edit: DC Sweep: Please left click to enlarge

When I did a DC sweep, results at 5mA (since data sheet takes 5mA as reference):

dVz/dIz (total dynamic impedance I guess) is 25 Ohm from the above plot. But the data sheet says at 5mA max zener impedance is 20 Ohm.

And the voltage at 5mA Iz is 10.2V in the plot. But the data sheet says nominal voltage at 5mA is 10V.

Is there anything wrong with the way I compare?

• Vz/Iz is a static resistance. Dynamic resistance = dVz/dIz in LTspice rd = d(Vz)/d(Iz) = d(V(n002))/d(I(R1)) – G36 May 26 '16 at 12:29
• I see you're using a transient (time) analysis. Not that that will give wrong results (it could though if you would use a very short pulse, like a few nano seconds long). However, I would use a DC analysis. I do not use LTspice, I use Spectre (very expensive) and from a DC simulation results I can easily plot the derivative dVz/dIz. You want to analyse DC behavior, so use the DC simulation ! You will need to replace the pulse source by a DC source that is swept. – Bimpelrekkie May 26 '16 at 12:34
• Like .dc V1 10 20 .1 – Steve G May 26 '16 at 12:35
• Indeed V1/Iz will not tell you anything. Note also that you can only evaluate the total dynamic resistance of the zener. So the sum of the intrinsic zener's dynamic resistance plus that of the bulk, bondwires etc. And also: only if that is modeled ! – Bimpelrekkie May 26 '16 at 12:37
• 25 ohm vs 20 ohm: then I expect that the model is inaccurate and pessimistic. Same for the 10.2 V versus 10 V. All your numbers are reasonably close in my opinion, I would not worry too much about the difference. Remember that models are "just a simplification of reality". Only very good models have good accuracy (like within 1%). Good models are not easy to find. – Bimpelrekkie May 26 '16 at 14:06

C:\Program Files (x86)\LTC\LTspiceIV\Draft289.asc

D1 0 Vz BZX84C10L
Iz 0 Vz {iz} AC 1 0
.lib standard.dio
.ac list   1kHz
.step dec param iz  1m 40mA 20
.end
Plot V(vz)/1A


Unfortunately I was not able to upload a picture.

• What is this about? – user16307 May 27 '16 at 15:45
• Ah ok this is a simulation by code – user16307 May 27 '16 at 16:34