Thanks, MrAl, but my original post was just seeking design knowledge, with no particular application in mind. A designer might wish to know, for example, if discharging a 1 Farad cap without using an external current-limiting resistor was feasible/safe, or how much time has to be allowed for a cap to discharge before a monostable can be re-triggered.
Hi there alec,
Oh ok, that's more like it
Now i have a much better idea what kind of information you are after. It appears you want to know the longevity of the 555 ic when subject to a large cap where it is questionable whether or not it's going to survive. That's a very valid question.
This really involves at least two things...
When we look at the data sheet we see that they do show some specs for the current, and they do show about 80ma at 1v for the discharge pin. What they dont show is the absolute current level limit. I dont know if they actually could spec the chip like this because it probably varies a bit from chip to chip and also with temperature due to the way bipolar transistors work with attention to the beta variations. This means one chip might draw 100ma and the other 150ma. It's very hard to put an exact number on this because of the possible variations. This in turn means it would be hard to determine how long a 1F cap would take to discharge down to say 0.1 volts when it starts out at say 15 volts.
What we can turn to is the power heating and see if that helps.
At 15 volts and 100ma that would be 1.5 watts, which would cook even the best package type i think if it ran for a long time that way. It only discharges for a limited time however, so we have to look at that.
With 1F and 100ma
dv=i*dt/C
dv=0.1*dt/1
dv=0.1*dt
15=0.1*dt
dt=150 seconds.
Thus it would take 150 seconds to discharge the 1F cap at 100ma.
So the question then becomes, would the package overheat in 150 seconds? It's hard to say without knowing the thermal time constant. I guess we could estimate that, but for now lets assume that it wont overheat in 150 seconds because the average power is about 0.75 watts over that time interval, and we wont retrigger until we force the average power down to 300mw. 300mw is chosen based on the thermal resistance of the 8 pin plastic package and the fact 600mw will cause a temperature rise of about 80 deg C, and we have to limit to a package rated at 70 deg C max.
So given the 1.5/2=0.75 watts and the target limit of 0.3 watts, we get the duty cycle of 0.3/0.75 which equals 0.4, or 40 percent. This means that since it takes 150 seconds to discharge we can not retrigger again until 225 seconds later, for a total cycle time of 375 seconds.
Now that was for 100ma, but if it draws 200ma that means we discharge in 75 seconds. This means at first it appears that we can retrigger 112.5 seconds later, for a total cycle time of 187.5 seconds.
However, the power increased to an average of 1.5 watts over the discharge time period, meaning now we have to limit to 0.3/1.5=0.2 or 20 percent duty cycle. This brings us back to a 375 second cycle time, even though we discharge faster now.
You might start to see the differences here and the similarities.
So in the end we end up with a little formula:
cycle time tc=C*Vcc^2/(2*0.3)
Of course this assumes that there is no significant dissipation caused by any of the other pins, especially the output pin. If there is, then that has to be subtracted from the 0.3 above:
cycle time tc=C*Vcc^2/(2*(0.3-Pd))
where Pd would be the output pin power, subject to a limit of less than 0.3 watts.
I think this now explains how the limitations of pin 7 come into play through the power dissipation of the chip.
Personally though i dont think i would use a cap as large as 1F.
Also, as i noted before, i had trouble with as little as 22uf but i didnt look into this too well to see if it was a chip defect or what.
The standard data sheet seems to indicate a max capacitance of 100uf. The more advanced versions go up to 10000uf.
Also interesting, the Texas Instruments version TLC555 (CMOS) states 150ma output OR discharge current limit as an absolute max for the chip. They also give some data for the discharge switch resistance vs supply voltage and temperature.