| Electronics > Beginners |
| LM393 comparator or op-amp question |
| (1/2) > >> |
| lordvader88:
I mapped out a fan controller circuit that uses an LM393 comparator, and I'm trying to calculate what it's doing. If I used basic ideal rules, a certain transistor would not be on at all due to the infinite input impedance of the LM393. (it's emitter resistor terminates at the inv. input, until the LM393 output goes low creating a 2nd path) When I simulate it with circuit wizard, they say about 4uA @4V, 6uA @6V into the inputs. For op-amps in general, where in the datasheet can find or figure out the current into them, with any confidence ? I see nanoAmps Whats a rule of thumb ? |
| Audioguru:
The LM393 has PNP input transistors. Their positive bias current (you must provide an input path to 0V for it) is shown as 400nA max. There is no rule of thumb, instead there is a datasheet with the specs of every part number. |
| JS:
Bias current or input vias current or input current (look also at current offset, drift and noise while dealing with any kind of precission) JS |
| Zero999:
Dave did a video about this awhile ago. https://www.eevblog.com/2013/06/07/eevblog-479-opamp-input-bias-current/ The LM393 has a PNP input stage so will source a tiny current out of its inputs, to whatever impedance is between them and 0V. The voltage generated by this current, will simply be the current, multiplied by the resistance. The datasheet lists a worst case of 400nA, for the LM393. http://www.ti.com/lit/ds/symlink/lm393-n.pdf The bias current will have a negative temperature coefficient, because it's inversely proportional to the Hfe, which increases with temperature. As Dave mentioned in the video, it's possible to minimise the effects of bias currents by keeping the input impedances the same, but this can be difficult in a comparator application. The simulation software you're using probably uses a generic model for the LM393. No doubt there are more accurate simulator packages than Circuit Wizard, but they most likely have a steeper learning curve. I use LTSpice, but it isn't interactive, like the simulators aimed at the educational market. |
| lordvader88:
Ok thank's so it's not something I could easily calculate or would be expected to know. I'll try some emitter currents and work back to see if the voltages I'm wondering about make sense. I'm just using the basic transistor model. |
| Navigation |
| Message Index |
| Next page |