Forums

problem with LPC2148

Started by Sutton Mehaffey November 27, 2006
I have a wierd problem that I am having trouble solving.

I have a serial display tied to TX1 of a LPC2148. It works great thru
UART1. RX is not used on this display. Only other connections are
+5V and GND.

I am also using P1.17-P1.24 as a virtual data bus on the 2148.
Whenever I put an address above 0xf on the data bus as an output
(switch port to output mode), I get an undesired adjusting of the
contrast on the display. I see that P1.21-23 are PIPELINE STATUS bits
as an output. What does that do? I'm guessing that when I switch to
data bus output mode that something is messing with the TX line on
UART1, thus doing strange things with my display. When I clear the
data bus, the contrast goes back to normal. I have PINSEL2 = 4, and
setting and clearing the data bus works fine. Any ideas on this problem.

Sutton

An Engineer's Guide to the LPC2100 Series

--- In l..., "Sutton Mehaffey" wrote:
>
> I have a wierd problem that I am having trouble solving.
>
> I have a serial display tied to TX1 of a LPC2148. It works great thru
> UART1. RX is not used on this display. Only other connections are
> +5V and GND.
>
> I am also using P1.17-P1.24 as a virtual data bus on the 2148.
> Whenever I put an address above 0xf on the data bus as an output
> (switch port to output mode), I get an undesired adjusting of the
> contrast on the display. I see that P1.21-23 are PIPELINE STATUS bits
> as an output. What does that do? I'm guessing that when I switch to
> data bus output mode that something is messing with the TX line on
> UART1, thus doing strange things with my display. When I clear the
> data bus, the contrast goes back to normal. I have PINSEL2 = 4, and
> setting and clearing the data bus works fine. Any ideas on this
problem.
>
> Sutton
>
The PIPESTAT pins are used when you enable the Enhanced Trace
Macrocell, so that is probably not part of your difficulty.

Others may have some comments on software possibilities, but there is
also a hardware point to consider. Whatever you are using on your
virtual data bus may be causing an increase in power supply current
consumption when the contrast change occurs. The contrast level of an
LCD is sensitive to voltage changes, so maybe your power supply is
varying?

-- Dave
I'll look at the voltage tomorrow. However, my backlight remains
steady when the problem occurs and that is the power hog. Just the
contrast dims. I would think if there was a voltage drop, then the
backlight would dim as well. By the way, the data and address buses
are tied to the same buses on a SCC2692 Dual Uart chip.

Sutton
--- In l..., "derbaier" wrote:
>
> --- In l..., "Sutton Mehaffey" wrote:
> >
> > I have a wierd problem that I am having trouble solving.
> >
> > I have a serial display tied to TX1 of a LPC2148. It works great thru
> > UART1. RX is not used on this display. Only other connections are
> > +5V and GND.
> >
> > I am also using P1.17-P1.24 as a virtual data bus on the 2148.
> > Whenever I put an address above 0xf on the data bus as an output
> > (switch port to output mode), I get an undesired adjusting of the
> > contrast on the display. I see that P1.21-23 are PIPELINE STATUS bits
> > as an output. What does that do? I'm guessing that when I switch to
> > data bus output mode that something is messing with the TX line on
> > UART1, thus doing strange things with my display. When I clear the
> > data bus, the contrast goes back to normal. I have PINSEL2 = 4, and
> > setting and clearing the data bus works fine. Any ideas on this
> problem.
> >
> > Sutton
> >
> The PIPESTAT pins are used when you enable the Enhanced Trace
> Macrocell, so that is probably not part of your difficulty.
>
> Others may have some comments on software possibilities, but there is
> also a hardware point to consider. Whatever you are using on your
> virtual data bus may be causing an increase in power supply current
> consumption when the contrast change occurs. The contrast level of an
> LCD is sensitive to voltage changes, so maybe your power supply is
> varying?
>
> -- Dave
>
Dave,

You were absolutely right. Apparently, if you manipulate the virtual
data bus (on my LPC2148) while the CE pin on the DUAL UART chip is
active low (my mistake - forgot to set it high on startup) causes a
serious current draw. Which caused the contrast on my display to dim.
Rail voltage dropped from 5V to 4.5V. The more data bits set, while
CE was active (low), the larger the voltage drop. Of course, if I did
it the correct way (disable CE, set the data bus, enabled CE), no drop
in voltage.

Moral of the story - when doing your own address decode, make sure ALL
signals are set correctly on startup.

Thanks,

Sutton

--- In l..., "derbaier" wrote:
>
> --- In l..., "Sutton Mehaffey" wrote:
> >
> > I have a wierd problem that I am having trouble solving.
> >
> > I have a serial display tied to TX1 of a LPC2148. It works great thru
> > UART1. RX is not used on this display. Only other connections are
> > +5V and GND.
> >
> > I am also using P1.17-P1.24 as a virtual data bus on the 2148.
> > Whenever I put an address above 0xf on the data bus as an output
> > (switch port to output mode), I get an undesired adjusting of the
> > contrast on the display. I see that P1.21-23 are PIPELINE STATUS bits
> > as an output. What does that do? I'm guessing that when I switch to
> > data bus output mode that something is messing with the TX line on
> > UART1, thus doing strange things with my display. When I clear the
> > data bus, the contrast goes back to normal. I have PINSEL2 = 4, and
> > setting and clearing the data bus works fine. Any ideas on this
> problem.
> >
> > Sutton
> >
> The PIPESTAT pins are used when you enable the Enhanced Trace
> Macrocell, so that is probably not part of your difficulty.
>
> Others may have some comments on software possibilities, but there is
> also a hardware point to consider. Whatever you are using on your
> virtual data bus may be causing an increase in power supply current
> consumption when the contrast change occurs. The contrast level of an
> LCD is sensitive to voltage changes, so maybe your power supply is
> varying?
>
> -- Dave
>