Lpc2000

I have a problem where the MCU won't boot up. I discovered that if I have no power on my 
board but with a serial cable plugged in, the serial driver drives VCC to -0.5V (max), 
probable leak since RS-232 Rx input is negative voltage. When I power up my board, the 
chip won't boot up. But if I kill power to my board but reapply power again before VCC 
drops below 0V, the chip would boot fine (this is possible by the way because of some 
large caps on my board, so VCC doesn't immediately go down to 0V).

Now my question, should the MCU boot up when you power up from negative VCC as little 
as -0.5V? Should a reset chip (I'm using TPS3836 3.3) help? In my case, it doesn't. Thanks.

-John G-

--- In lpc2000@yahoogroups.com, "jpsg123" <gonzagajohn@h...> wrote:
> I have a problem where the MCU won't boot up. I discovered that if I
have no power on my 
> board but with a serial cable plugged in, the serial driver drives
VCC to -0.5V (max), 
> probable leak since RS-232 Rx input is negative voltage. When I
power up my board, the 
> chip won't boot up. But if I kill power to my board but reapply
power again before VCC 
> drops below 0V, the chip would boot fine (this is possible by the
way because of some 
> large caps on my board, so VCC doesn't immediately go down to 0V).
> 
> Now my question, should the MCU boot up when you power up from
negative VCC as little 
> as -0.5V? Should a reset chip (I'm using TPS3836 3.3) help? In my
case, it doesn't. Thanks.
> 
> -John G-

Hi John,

is there any option to connect ground through your serial cable
between your board and (I guess the PC).  This should help. If not,
unplugging the serial cable before booting should help too. Saw this
kind of problem with negative voltage before and it does screw up the
power on sequence.

Cheers, Bob

I don't suppose this has anything to do with invoking the bootloader
(P0.14 - UART1 Data Carrier Detect low).

--- In lpc2000@yahoogroups.com, "lpc2100_fan" <lpc2100_fan@y...> wrote:
> --- In lpc2000@yahoogroups.com, "jpsg123" <gonzagajohn@h...> wrote:
> > I have a problem where the MCU won't boot up. I discovered that if I
> have no power on my 
> > board but with a serial cable plugged in, the serial driver drives
> VCC to -0.5V (max), 
> > probable leak since RS-232 Rx input is negative voltage. When I
> power up my board, the 
> > chip won't boot up. But if I kill power to my board but reapply
> power again before VCC 
> > drops below 0V, the chip would boot fine (this is possible by the
> way because of some 
> > large caps on my board, so VCC doesn't immediately go down to 0V).
> > 
> > Now my question, should the MCU boot up when you power up from
> negative VCC as little 
> > as -0.5V? Should a reset chip (I'm using TPS3836 3.3) help? In my
> case, it doesn't. Thanks.
> > 
> > -John G-
> 
> Hi John,
> 
> is there any option to connect ground through your serial cable
> between your board and (I guess the PC).  This should help. If not,
> unplugging the serial cable before booting should help too. Saw this
> kind of problem with negative voltage before and it does screw up the
> power on sequence.
> 
> Cheers, Bob

Hi Bob,

I already have ground in the serial cable connected to my board gnd. About unplugging, 
you can never really tell people how to do stuff, some might connect all cables before 
powering on or the other way. Anyway, thanks for the info, I thought I was the only one 
seeing this problem.

-John G-

Hello John,

My 2 cents:

> Now my question, should the MCU boot up when you power up 
> from negative VCC as little 
> as -0.5V? 

Can this cause a different power-up sequence for the 3.3V and 1.8V
supplies (I guess the "-0.5V" is nominally 3.3V) ? Meaning:
- in the "normal" case, 3.3V rises before 1.8V,
- in the "abnormal" case, 3.3V lags a bit, and 1.8V appears first.
You will probably need to have acces to digital scope to check that.

Sorry if you have already checked, or if this is clearly part of the
data sheet, our project has not started yet, and I only have general
ideas  at this moment.


> Should a reset chip (I'm using TPS3836 3.3) help? 
> In my case, it doesn't. Thanks.

Just a thought here: I feel using a 3.3V reset device on a nominal 3.3V
supply is too tight. If on the board you use, "3.3V" is really 3.30V
nominal, of course. I would expect problems with glitches, aging &
temperature drifts. Unless your application warrants otherwise, I would
recommend a 3.0V reset device.

Hope this helps

Etienne

--- In lpc2000@yahoogroups.com, "Etienne Lubenec" <e.lubenec@g...> wrote:
> Hello John,
> 
> My 2 cents:
> 
> > Now my question, should the MCU boot up when you power up 
> > from negative VCC as little 
> > as -0.5V? 
> 
> Can this cause a different power-up sequence for the 3.3V and 1.8V
> supplies (I guess the "-0.5V" is nominally 3.3V) ? Meaning:
> - in the "normal" case, 3.3V rises before 1.8V,
> - in the "abnormal" case, 3.3V lags a bit, and 1.8V appears first.
> You will probably need to have acces to digital scope to check that.
> 
> Sorry if you have already checked, or if this is clearly part of the
> data sheet, our project has not started yet, and I only have general
> ideas  at this moment.
> 
> 
> > Should a reset chip (I'm using TPS3836 3.3) help? 
> > In my case, it doesn't. Thanks.
> 
> Just a thought here: I feel using a 3.3V reset device on a nominal 3.3V
> supply is too tight. If on the board you use, "3.3V" is really 3.30V
> nominal, of course. I would expect problems with glitches, aging &
> temperature drifts. Unless your application warrants otherwise, I would
> recommend a 3.0V reset device.
> 
> Hope this helps
> 
> Etienne

Thanks for the ideas. One thing I forgot to mention is, if I power up with the serial cable 
unplugged, the MCU boots up properly.

-John G-

jpsg123 wrote:

>I have a problem where the MCU won't boot up. I discovered that if I have no power on my 
>board but with a serial cable plugged in, the serial driver drives VCC to -0.5V (max), 
>probable leak since RS-232 Rx input is negative voltage. When I power up my board, the 
>chip won't boot up. But if I kill power to my board but reapply power again before VCC 
>drops below 0V, the chip would boot fine (this is possible by the way because of some 
>large caps on my board, so VCC doesn't immediately go down to 0V).
>
>Now my question, should the MCU boot up when you power up from negative VCC as little 
>as -0.5V? Should a reset chip (I'm using TPS3836 3.3) help? In my case, it doesn't. Thanks.
>
Just a tip but have you tried a bleed resistor across the supply? 
Normally any stray voltages that manage to sneak through the substrate 
can be damped this way.

This -0.5V can be doing all kinda wierd things with the substrate of the 
micro because you'll find that elements on the chip are only isolated 
from one another by normally reversed biased pn junctions. The -0.5V is 
probably an indication that these isolation junctions are now conducting 
and clamping, oowwhhh, that'll hurt on power-up!.

There is always a trade-off though because in low-power systems you 
don't want to waste power through a bleed resistor. But for most systems 
this helps enormously and hey, it's easy to put on the board or change.

my2cents

--
Peter Jakacki