## Tuesday, 2 May 2017

### Current sink: one of my first experiences with Eagle

I’ve learnt a few things the hard way while messing around with electronic circuits, here is a basic summary:

1. When designing a circuit for a basic task, keep it SIMPLE! The less components the less risk of component failure.
2. Complicated stuff either looks complicated but it isn’t or it isn’t working as it should.
3. Real capacitors and inductors are nothing like I expected them.
4. Simulators are simulators (this point should not be under estimated).
5. Gained experience seems to be proportional to the volume of black smoke you produce at ambient pressure and temperature from components overheating (read: from burned transistors).

Keeping in mind these points that I collected step by step in my electronics adventures, the main goal of this project was to make myself comfortable using Eagle.

On the plus side, Eagle seems to be a really nice piece of software to work with, although lately it was a bit under internet fire due to the subscription service announced by the new owner, Autodesk.

Why did I decide to design a current sink? Well, if you have studied any electrical circuit theory, there will come a time when you will be asking yourself what about current sources? I would like to make 2mA flow across my load, let me put a current source in series with it. Yeah, sure, first you’d have to make one! But do current sources even exist?

Whether they exist or not I cannot really tell you, what I can tell you is that, in practice, you can build something that closely resembles it!

For instance, take this beautiful CAD drawing

Assuming your current source should be able to output a constant current between 2 and 8 mA the circuit above should work. If you know your opamp rules, the opamp will do whatever it can in order to set V+ equal to V- (assuming Vos is negligible) and if you can put a 2-8V voltage on R1 then by Ohm’s law a nice 2 to 8 mA of current will flow through it.

But will it though?

Well, assuming Vcc is 10V the circuit should operate fine within the specs, however R1 should be exactly 1k Ohm. At this point you can either look for a precision resistor, or put a decent multimeter in series with R1, look for a nice adjustable potentiometer and fine tune V+ until you get exactly the desired current through R1.

At this point, you may have realized that the darlington pair was useless and you could have got away with just a single 2N3904 NPN. Oh well, what have I just said above? Keep it simple.

The final final Eagle drawing looks nice though. Next I’m going to try and make a PCB out of it.

If you’d like to download my drawing for this project, then you can find it here.