I should look this up. But I'm lazy, and I know who to trust on this group.
I'm writing a set of instructions for a board that has two physicaly identical connectors, one with RS-232 signals on it, and one with USB. I know that a user can plug USB into the RS-232 side with impunity -- what happens when a USB port gets -7V applied to it's D- input?
Donkey shorn.
Why does the RS-232 have a USB-like connector? ...Jim Thompson
The connector is neither RS-232-like nor USB-like. Rather, It's what the customer likes.
Den fredag den 10. januar 2014 21.11.00 UTC+1 skrev Tim Wescott:
depends what is driving it, but I doubt it would be happy
There is probably some clamping on the USB other than the ESD diodes on the 3.3V USB chip input, can you put a resistor in to limit the current?
-Lasse
"Massage" the customer ;-) ...Jim Thompson
The board is cast in stone. I can tell the customer not to plug things in wrong, though.
Mostly, I wanted to know how much stress to put on "don't do that".
We have two devices one that uses USB and the other serial on the same socket (USB-A), I haven't seen any equipment damaged by this sort of mishandling.
I think the rs232 driver's 1200ohm output impedance is high enough that the USB can withstand the current.
Have you tried it? Plug it in this way and that with various orders of things, power on and off. If you're worried about the -7V, then turn it up, and see where it fails.
George H.
...or modify the RS-232 signals to single polarity (0V, 5V) with commensurate relaxation of the 25V fault spec..
The problem with any clamping to a rail is power supply rails expect to source current, not sink it. So the external event just pulls the rail high until something fries. Perhaps a zener or transorb can be used to clamp the supply rail.
Now going negative I suppose is OK because ground is ground. If the current is limited, it might be OK. But of course this is a latch-up scenario.
There is no set current limit or impedance specification on a 232 driver. There are slew rate limits on the pin, but that doesn't translate to an impedance or current limit since slewing can be controlled in other ways.
"ground is ground"
That is a fairly dangerous assumption due to various "ground bounce" or "ground loop" issues :-).
Not really. The point is ground can source or sink. That is what you need if you expect a clamp to work. Everything else is just good layout practice.
Clamping to a power supply rail really depends on the capability of the power supply rail to absorb being hit. Capacitance will help a clamping situation where the disturbance is transient. But the situation here is the disturbance is itself a power supply.
In qualifying 232 chips, They are expected to be able to handle shorts to
232 levels. Nobody assumes the source is a 232 transmitter. Rather you put a hard supply on the pin and see if it will survive. In addition, the chip has to work properly once the disturbance is removed. Those 232 chips had charge pumps on them, and the pumps had to start even under the transmitter shorted condition.While I say this is for 232 chips, really any interface chip needs to be this rugged.
the
Oh lordy, one of those customers.
?-)
Have something to add? Share your thoughts — no account required.
Ask the community — no account required