| ??? 06/15/08 06:08 Read: times |
#155917 - What negative effect would that be? Responding to: ???'s previous message |
What negative effect would that be, Erik? This looks like a good topic for discussion, perhaps better in a separate thread.
Erik Malund said:
Richard Erlacher said:
I'd suggest a low (33-75 ohm) resistance in series with Vcc, and a 1 nf plus a 10 uF on either side of the resistor. Likewise on the GND side, but with the total capacitance in the range I mentioned. This is fuzzy. I read that you want to insert resistors in series with the ground pins Oh, but I do! I might feel differently with less-noisy circuits, but with the old MAX232, and splitting the resistance betwee Vcc rail and GND rail, say, 2:1, respectively, it's just to help motivate the cap's to do their work, though a sizeable ferrite bead might be better. Moreover, you can "tune" the filtering in the rail to the nominal frequency of the MAX232, since its effect on system noise will be about half to 80% of the total system noise, unless you're firing SCR's and relays and the like on the same supply as your logic. (which I'm sure you don't) and where does the other side of the 'likewise' caps go?
They go to Vcc on the MAX232 side of the resistors in the GND rail. Whenever the GND rail is pulled down, the cap mitigates that dip, and when it goes high, the cap mitigates the bump, just as it does on Vcc. The jury is out on this one, some use a small resistor in series with Vcc, I never do. I can see the purpose, but do not see the need. I think the negative effect of the resistor outweighs the positive.
The MAX232, unlike the MAX232A and MAX3232 are VERY noisy, perhaps more so than the entire reset of the circuit. There's a small gain, and, really, that doesn't matter much, so increasing the noise on the MAX232 side of the resistors in the supply rail doesn't matter. Erik |
