Current Source/Sink Notation
Current Source/Sink Notation
(OP)
I have a relatively simple question:
Today when looking over some datasheets for digital buffers (74als652) I ran across specs for I(OH) and I(OL). For this particular device, I(OH) is -15mA and I(OL) is 24mA. I take this as meaning I can "source" a load (connected between my output and GND) with 15mA, and can "sink" a load (connected between VCC and my output) with 24mA.
I would have thought that the "-" notation would be used for the sink condition rather than source condition. Does anyone have any background information about this?
Thanks!
Today when looking over some datasheets for digital buffers (74als652) I ran across specs for I(OH) and I(OL). For this particular device, I(OH) is -15mA and I(OL) is 24mA. I take this as meaning I can "source" a load (connected between my output and GND) with 15mA, and can "sink" a load (connected between VCC and my output) with 24mA.
I would have thought that the "-" notation would be used for the sink condition rather than source condition. Does anyone have any background information about this?
Thanks!





RE: Current Source/Sink Notation
- when you are sourcing current, the current is extracted/subtracted from the output, therefore negative
when you are sinking current, the current is added into the output, therefore positive.
But I think it is just a definition somebody made up in the 60's, and it became a standard. It is the point of view from the logic gate side, when you think of the gate as a closed system.
- you can extract/sink more current from the output, if you don't care about the voltage levels. But look at the absolute maximum ratings, and it's not a good design practise.
The datasheet just says that at those current levels the output voltage will still meet the high/low logic voltage levels; it does not mean the output will explode at -16mA or +25mA.
- if you want your circuit as universal as possible, arrange your loads so that your logic gate sinks current, e.g. LEDs coming from Vcc (with R, of course). Most newer CMOS TTL don't care, but older TTL technologies (LS,ALS,etc.) can significantly sink more current than they can source. The same for (even newer) 8051 microcontrollers.
Volker
RE: Current Source/Sink Notation
You may be interested to note that the sink/source sign notation first gained acceptance in the early days of TTL in the changeover from DTL during the mid 1960's (that's Diode-Transistor Logic for those that don't know). DTL gates were generally what we would now call open-collector devices, consequently could not source any current at all unless a pull-up resistor was provided. Before the ubiquitous 54/74 series TTL became firmly established, at one point the term CCSL - Compatible Current-Sinking Logic - was coined for any DTL and TTL device family that could be interconnected.
RE: Current Source/Sink Notation
That certainly clears things up. I was looking at it from the perspective that it was a digital output and so current flowing out would be marked "+", kind of like nodal and mesh equations.
BrianG, Thanks also for your input. It's certainly interesting to see how stuff like this comes about!
RE: Current Source/Sink Notation
(- to + ). Someone correct me if I'm wrong.
-99
RE: Current Source/Sink Notation
My professors always pointed out that it's strictly a matter of convention, and that this is the convention currently used. I believe we were told that Benjamin Franklin used this convention in describing his experiments with electicity and it has stuck. Of course that was before the discovery of the electron and recognition that movement of electrons is the physical form of many currents.
There are places where direction of current flow is based on the flow of electrons; for example in U.S. Navy electronics schools (if they still do). But one doesn't mix and match the two approaches. So VolkeR's post is the good one here.