Current Mirror Question

steveb

Joined Jul 3, 2008
2,436
Okay, figured out that all my signal transistors are pretty fried and causing problems. Switched to an NPN array and everything seems to be making sense.

By the way, can alpha (Ie/Ic) be greater than 1?
You asked a loaded question.

Alpha is Ic/Ie (not Ie/Ic) and it is less than one in the active region of operation (forward biased base emitter and reverse biased base collector).

If I remember correctly, alpha=B/(B+1) which is clearly less than one.

However, Ie/Ic is greater than one when operating in the active region.
 

Thread Starter

blah2222

Joined May 3, 2010
582
Yeah <snip> I meant alpha = Ic/Ie, sorry. Well I used the NPN array and it's fully functional and this is what I'm getting for my readings:


I measured this circuit over and over again and each time Ie2 is less than Ic2. I didn't mix the emitter and collectors up either, as I initially thought. Ie2 is less Ic2...
 
Last edited by a moderator:

steveb

Joined Jul 3, 2008
2,436
Yeah <snip> I meant alpha = Ic/Ie, sorry. Well I used the NPN array and it's fully functional and this is what I'm getting for my readings:


I measured this circuit over and over again and each time Ie2 is less than Ic2. I didn't mix the emitter and collectors up either, as I initially thought. Ie2 is less Ic2...
You also have Ie2 less than half Ie1. These are all very strange and possibly due to measurement error or human error. Can you explain how you are measuring the emitter currents? What equipment, how is it set up and how are you doing the measurement.

Edit: I just noticed that the transistor is near or in saturation. Let's think about what that will do.

Edit Again: Still think it's human or measurement error. If I add all the currents at nodes, your values disobey Kirchoff's current law.
 
Last edited:

Thread Starter

blah2222

Joined May 3, 2010
582
haha why is it so hard for me to learn transistor basics ahhhh. Nothing so far has worked so far as I have expected, thus creating my aimless confusion. I apologize for the lack of knowledge in this but darn this is frustrating not getting simple circuits to work...

If you guys set up this circuit what kind of results should I be getting? I know the Vbe is at a good voltage 0.7ish and the collector current has a gain of 100 over the base current (half of Ish). But the darn emitter currents are weird...

Adding to this problem is that I am running my laptop off a USB stick for it's harddrive (ubuntu linux) and has no space for an simulation software to help me out. My uncle who helps me with this stuff is out of town, and I thought I knew the results that I should be getting but apparently not...
 

Wendy

Joined Mar 24, 2008
22,647
The circuit worked as predicted from here. The two collector currents match within experimental errors (and no two transistors are really identical).

Since you are dividing the base current two ways the formula is probably Ic = ß Ib * 2 . The currents going through the circuit without a second BE would still be Ic = ß Ib . Since you are controlling the total current they balance according to the equation Ic = ß Ib, it doesn't have a choice since the formula is a mostly accurate description. Unfortunately ß is a bit of a variable, it doesn't hold nearly as still as we'd like. ß at saturation is much lower than ß at the middle of the range for example.

The main point is, it doesn't matter what resistor you have on the collector of Q2, the current will be the same as Q1 current. If you shorted the resistor on Q2 collector it would still be the same current.

One side comment, current mirrors are interesting, but they aren't used that much because of the temperature problem. After you figure this out you will have the big 3 circuits to understand, common emitter, common collector, and common base.
 
Last edited:

steveb

Joined Jul 3, 2008
2,436
haha why is it so hard for me to learn transistor basics ahhhh. Nothing so far has worked so far as I have expected, thus creating my aimless confusion. I apologize for the lack of knowledge in this but darn this is frustrating not getting simple circuits to work...

If you guys set up this circuit what kind of results should I be getting? I know the Vbe is at a good voltage 0.7ish and the collector current has a gain of 100 over the base current (half of Ish). But the darn emitter currents are weird...

Adding to this problem is that I am running my laptop off a USB stick for it's harddrive (ubuntu linux) and has no space for an simulation software to help me out. My uncle who helps me with this stuff is out of town, and I thought I knew the results that I should be getting but apparently not...
Don't worry, this kind of thing happens all the time. Eventually, the issue is found.

My suggestion is to increase your resistor values. Right now your transistor is very near saturation. While this should not make Kirchoff's current law go out the window, it may be creating an issue that is confusing the situation. Try 20K resistors instead and see what happens. Report the data to us, and I'll bet we can figure it out.
 

Thread Starter

blah2222

Joined May 3, 2010
582
Edit Again: Still think it's human or measurement error. If I add all the currents at nodes, your values disobey Kirchoff's current law.
I am still puzzled because this was the first thing that I noticed after I measured this. I am going to try it again right now and report back.

Edit. Thank you for the support. I appreciate the help from you guys. So a 20K for both? I guess Ic2 wouldn't really matter if it's a current sink.
 

steveb

Joined Jul 3, 2008
2,436
I am still puzzled because this was the first thing that I noticed after I measured this. I am going to try it again right now and report back.

Edit. Thank you for the support. I appreciate the help from you guys. So a 20K for both? I guess Ic2 wouldn't really matter if it's a current sink.
Yes use 20K for both. This will take the transistor away from saturation.

Edit: Crap, I take that back. Use 20K on the diode bias only. If you already used two 20K resistors, report that data too. It could help.
 
Last edited:

Thread Starter

blah2222

Joined May 3, 2010
582
After switching both 10Ks with 20Ks here's what I get from the same circuit source still at 6V, current mirror:

Vbe1 = 0.693 V
Vbe2 = 0.693 V

Iref = 0.270 mA
Ic2 = 0.265 mA
Ish = 0.005 mA

Ie1 = 0.258 mA
Ie2 = 0.162 mA What the heck...?
 

steveb

Joined Jul 3, 2008
2,436
After switching both 10Ks with 20Ks here's what I get from the same circuit source still at 6V, current mirror:

Vbe1 = 0.693 V
Vbe2 = 0.693 V

Iref = 0.270 mA
Ic2 = 0.265 mA
Ish = 0.005 mA

Ie1 = 0.258 mA
Ie2 = 0.162 mA What the heck...?
Sorry, I realized too late that it's better to have 20K on the diode bias and 10K on the transistor load. Using two 20K resistors will still leave the transistor near saturation. See, now you don't have to feel so bad. :)
 

Thread Starter

blah2222

Joined May 3, 2010
582
Rlimit = 20K and Rload = 10K, Source = 6 V; Current Mirror:

Vbe1 = 0.694 V
Vbe2 = 0.694 V

Iref = 0.271 mA
Ic2 = 0.270 mA
Ish = 0.005 mA

Ie1 = 0.263 mA
Ie2 = 0.166 mA Hrm...
 

Thread Starter

blah2222

Joined May 3, 2010
582
Here's my circuit:

http://img819.imageshack.us/f/dsc03881r.jpg/

The 6 Volt power supply leads are out but here is the main circuit.

Blue resistor is the Rlimit (20K) and the black resistor is the Rload (10K).
Grey wires are the emitters to ground and brown wires are the Q1 collector to base and Q1 base to Q2 base.

Edit: click on the image to zoom in for better view.

Any ideas?
 
Last edited:

steveb

Joined Jul 3, 2008
2,436
Here's my circuit:

http://img819.imageshack.us/f/dsc03881r.jpg/

The 6 Volt power supply leads are out but here is the main circuit.

Blue resistor is the Rlimit (20K) and the black resistor is the Rload (10K).
Grey wires are the emitters to ground and brown wires are the Q1 collector to base and Q1 base to Q2 base.

Edit: click on the image to zoom in for better view.

Any ideas?
One first thought is that you don't have any filter caps on the power supply at the board. If you make measurements without an Oscope, then it's possible to have oscillations and/or noise give false readings on a standard meter.

I'll keep looking, but using power supply filter caps is good practice even if it turns out not to be the issue.
 

Thread Starter

blah2222

Joined May 3, 2010
582
One first thought is that you don't have any filter caps on the power supply at the board. If you make measurements without an Oscope, then it's possible to have oscillations and/or noise give false readings on a standard meter.

I'll keep looking, but using power supply filter caps is good practice even if it turns out not to be the issue.
I checked it out on an O-scope (tektronix) that my uncle lent me and it's a constant 6.00 Volts coming out.

It's getting late and I think I'm going to head off to bed, but I will be back at this tomorrow. I really appreciate the help I'm getting with this!

Thanks a lot guys, especially steveb!
 
Top