I have a chip that is giving me data over a UART, and I have
encountered a very strange issue. Under some conditions (such as
another sensor being connected to the chip), the MSP stops receiving
data from the UART. To debug, I took out the UART Tx line and a ground
line and connected them to an external line driver (to view the data
in hyperterminal). To my puzzlement, when only the ground of my
circuit is connected (to the ground of the external line driver
circuit), the UART data is received. But when that ground is removed,
the UART data is no longer received (but I am sure it is still being
generated). The two circuits (i.e. mine and the external line driver)
are powered by two separate power supplies (both connected to mains
power). Hmmm..... any ideas? It seems that providing an additional
ground connection is making some sort of difference but I don't
understand why. Your inputs are welcome! Thanks.
Strange ground influance on UART receiption
Started by ●March 11, 2008
Reply by ●March 11, 20082008-03-11
--- In m..., "Leon" wrote:
> Are you depending on the two circuits being grounded via the power
supplies?
> They might not have a connection to the mains ground. You should have a
> separate ground connection between the two systems.
>
> Leon
>
No, not at all. The other circuit was just to debug and tap in to my
project. I wanted to spy on the UART so I connected Tx to Rx and of
course the ground. There is no other reason for the connection. So
what is strange is, that while just connecting my ground to another
power supply ground, that affects my circuit. I don't see why.
> Are you depending on the two circuits being grounded via the power
supplies?
> They might not have a connection to the mains ground. You should have a
> separate ground connection between the two systems.
>
> Leon
>
No, not at all. The other circuit was just to debug and tap in to my
project. I wanted to spy on the UART so I connected Tx to Rx and of
course the ground. There is no other reason for the connection. So
what is strange is, that while just connecting my ground to another
power supply ground, that affects my circuit. I don't see why.
Reply by ●March 11, 20082008-03-11
The minimum connection for RS-232 or TTL Serial is Tx, Rx, Gnd - 3
wires.
A 0 or 1 is determined relative to ground. The ground is also used to
return the voltage. If you do not have the ground there is no reference.
If you have a ground, then check to see if it is connected. Try to check
it with an ohm meter with the power off, AND a volt meter with the power
on.
Kip
On Tue, 2008-03-11 at 12:26 +0000, merapcb wrote:
> --- In m..., "Leon" wrote:
>
> > Are you depending on the two circuits being grounded via the power
> supplies?
> > They might not have a connection to the mains ground. You should have a
> > separate ground connection between the two systems.
> >
> > Leon
> > No, not at all. The other circuit was just to debug and tap in to my
> project. I wanted to spy on the UART so I connected Tx to Rx and of
> course the ground. There is no other reason for the connection. So
> what is strange is, that while just connecting my ground to another
> power supply ground, that affects my circuit. I don't see why.
>
>
wires.
A 0 or 1 is determined relative to ground. The ground is also used to
return the voltage. If you do not have the ground there is no reference.
If you have a ground, then check to see if it is connected. Try to check
it with an ohm meter with the power off, AND a volt meter with the power
on.
Kip
On Tue, 2008-03-11 at 12:26 +0000, merapcb wrote:
> --- In m..., "Leon" wrote:
>
> > Are you depending on the two circuits being grounded via the power
> supplies?
> > They might not have a connection to the mains ground. You should have a
> > separate ground connection between the two systems.
> >
> > Leon
> > No, not at all. The other circuit was just to debug and tap in to my
> project. I wanted to spy on the UART so I connected Tx to Rx and of
> course the ground. There is no other reason for the connection. So
> what is strange is, that while just connecting my ground to another
> power supply ground, that affects my circuit. I don't see why.
>
>
