4
\$\begingroup\$

I was trying to simulate the circuit mentioned in https://wiki.analog.com/university/courses/electronics/text/chapter-11 under paragraph 11.4.1. I need to make a current amplifier with a current gain of 1.5. Thus I made the following circuit: enter image description here As you can see the current mirror does not work as supposed to work. The current gain is not 1.5 times the input current. Is there any reasoning behind this or it is just a simulation error. Thank you.

\$\endgroup\$
2
  • \$\begingroup\$ Hint: have a look at section 11.5 of the document you link. My bet is on early effect, the input transistors Vce is ~650 mV, while the output transistors Vce is 24 V. Try adding some emitter resistance, or reducing Vcc. \$\endgroup\$ – Vladimir Cravero Aug 22 '19 at 18:57
  • \$\begingroup\$ Do you need to make this circuit, in reality, with real parts and perform a design that will get you reasonably close to expectations? Or is this all about getting a simulated result using parts which have zero variation to them and operate at a fixed (and identical across all parts) temperature? If the latter, are you allowed to select the BJTs used for simulation? Are you allowed to vary their model parameters to anything you please? \$\endgroup\$ – jonk Aug 22 '19 at 19:23
1
\$\begingroup\$

I simulated your circuit in LT Spice and got a ratio of 1.79. You only need 2 transistors for this circuit, so I revised the circuit using only 2 transistors but put resistors between emitter and ground. (I kept the 1k resistor in Q2.) So referring to your circuit, eliminate Q1, Q4, and Q5. Place a 30 ohm resistor between emitter of Q2 and ground and a 20 ohm between emitter of Q3 and ground. I now simulate a ratio of 1.5. As I decreased the 30 and 20 ohm resistors it diverged from the current ratio of 1.5. If you change the collector resistor from 1k you will have to make adjustments in the emitter resistor. When building this circuit with real transistors you may need to do some trimming.

\$\endgroup\$
2
  • \$\begingroup\$ I am curious why you put a current source and voltage source in series with Q1 and Q2. Remove the current source and change the 1k to 5825 ohms. (24-0.7)/0.004 = 5825. You will get a current ratio of 1.8, therefore you will need to add emitter resistors to balance it out. No, the transistors are not getting too hot. The power is not 24V * 4mA, the 1k is absorbing the power. Q1 and Q2 are seeing 4mA *652mV = 2.6mW. A 2N222 is good up to 800mA. \$\endgroup\$ – Rob B. Aug 23 '19 at 12:35
  • \$\begingroup\$ Also PSpice does not model thermal runaway unless you have purchased good models, the free component models do not have thermal components in them. I still don't understand how you got 25mA in Q3 -Q5. \$\endgroup\$ – Rob B. Aug 23 '19 at 12:35
0
\$\begingroup\$

The transistors get HOT (25v * 4mA), and the Vbe needed to support 4mA is reduced, so you get even more current.

In other words, you are seeing nascent thermal runaway.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.