Who isn't still learning! Even our EMI experts all disagree with what
needs to be added or not.

regards,
Charles

--- In lpc2000@lpc2..., "Rob Jansen" <rob@m...> wrote:
>
> Charles,
> 
> thanks for your elaborating answer.
> I know, stuff like motor control can induce big problems in a
system. I've
> seen spontaneous system resets when driving even
'small' motors that are
> used in washing machines (200V DC motor).
> I have also worked with 100V 3 phase motors that controlled an A0 size
> plotter (a surplus device we got almost 30 years ago - no integrated
> circuits yet ...)  with a lot of these 'strange' solution I did
not
> understand back then.
> 
> Even after 30 years I'm still learning.
> 
> Rob
> 
> >   We design allot of motion control and there is plenty of noise that
> > is generated that we have to take into account. One saving grace is
> > the CE approved power supply. We ran experiments with one at an
> > approved lab and found that it eliminates all the conducted emmissions
> > from the system if one is there. That saved us from having to put
> > chokes and such on our boards since we could be pumping as much as 12
> > amps through the board at any one time.
> >
> > regards,
> > Charles
>

Charles,

thanks for your elaborating answer.
I know, stuff like motor control can induce big problems in a system. I've
seen spontaneous system resets when driving even 'small' motors that
are
used in washing machines (200V DC motor).
I have also worked with 100V 3 phase motors that controlled an A0 size
plotter (a surplus device we got almost 30 years ago - no integrated
circuits yet ...)  with a lot of these 'strange' solution I did not
understand back then.

Even after 30 years I'm still learning.

Rob

>   We design allot of motion control and there is
plenty of noise that
> is generated that we have to take into account. One saving grace is
> the CE approved power supply. We ran experiments with one at an
> approved lab and found that it eliminates all the conducted emmissions
> from the system if one is there. That saved us from having to put
> chokes and such on our boards since we could be pumping as much as 12
> amps through the board at any one time.
>
> regards,
> Charles

----- Original Message ----- 
From: "charlesgrenz" <charles.grenz@char...>
To: <lpc2000@lpc2...>
Sent: Saturday, January 28, 2006 1:28 PM
Subject: [lpc2000] EMI - was Re: lpc2138 5V tolerant is not really tolerant!
	> --- In lpc2000@lpc2..., Rob Jansen <rob@m...> wrote:
>>
>> I'm not an EMI expert,
>> > I do have a 51
>> > ohm series resistor on all I/O lines including the RXD and TXD of
>> > USART0 to make sure that no EMI problems occur.
>> >
>> > regards,
>> > Charles
>> >
>>
>> I did do some ESD research/design once for a web tablet. Although
> the IC
>> used (forget which one) was said to be ESD safe, firing at the touch
>> screen with a Human Body Model ESD tester resulted in error. I then
>> added R/C combinations to resolve this problem.
>>
>> But I'm definitely not an EMI expert and would even think that
adding
>> resistors may make the problem worse ...
>> Are there any special considerations to take into account with (apart
>> from PCB design and use of separate ground/power planes) ?
>>
>> Regards,
>>
>>     Rob
>>
>> P.s: let me state that I was _not_ the original designer of that web
>> tablet :o)
>>
> Hi Rob,
>
>  I am not either even though I have been designing our products to
> eliminate EMI for CE approvals for the past 4 years on our products.
>
>  I have also used what Leon suggested as well (the AVX transguards),
> but I am using some resistor packs from CTS that are 0.5mm pitch which
> allows me to put them about 3mm away from the processor pins.
>
>  We also designed a motion control stage for a large Canandian
> company and they wanted 47 ohm resistors on all I/O's from a large 12
> x 14 inch, multilayer board. They still had problems with the EMI, but
> then we added some ferrits to the motor stages and that helped them
> drop to 5db below class B.
>
>  We do have clients (one in particular) who are and suggested adding
> an resistor to each processor line to eliminate any EMI and to help
> with ESD. They suggested a 30 to 51 ohm resistor on all processor
> lines along with 100pf to 1nF capacitor on ports that extend off the
> board. We have followed this advice 3 times and each of our boards
> that went through CE, class B with no problems. There are still other
> considerations like board layout, capacitor and other components to
> help reduce the noise as well.
>
>  We design allot of motion control and there is plenty of noise that
> is generated that we have to take into account. One saving grace is
> the CE approved power supply. We ran experiments with one at an
> approved lab and found that it eliminates all the conducted emmissions
> from the system if one is there. That saved us from having to put
> chokes and such on our boards since we could be pumping as much as 12
> amps through the board at any one time.

Where I used to work we had to put ferrite beads and capacitors on the 
serial port lines to get our units through the emissions tests. Connectors 
are available with ferrite built-in, which could be useful.

Leon

--- In lpc2000@lpc2..., "Gus" <gus_is_working@y...> wrote:

> On my board I have LPC2138 connected to MAX232
running at 5V. I had 
> problems with UART until I added a 4.7K resistor in series. and the 
> voltage high on it now is 2.3V
> 
> Are the pins real 5V tolerant? The data sheet says UART RX is 5V 
> tolerant!!!

Admittedly, I've never used the above chips, but are you sure you
don't have the MAX232 drivers hooked up backwards? There is the 0VDC
to +5VDC TTL logic level side, then the +/-5VDC RS-232 side. 

If you have the +/-5VDC side connected to the LPC2138, it would
explain why the LPC2138 doesn't seem 5V tolerant for you. Having -5VDC
applied to an I/O line is probably not what Philips meant by saying
"5V tolerant I/O." 

Also, the +/-5VDC side of the MAX232 is driven from a low current
charge pump power source, which might explain why adding a
4.7K series resistor would drop the voltage down to 2.3V. Otherwise,
maybe the LPC2138's input pin is not configured properly in your
firmware so that the LPC2138 is actually trying to drive the line as
an output low.

--- In lpc2000@lpc2..., Rob Jansen <rob@m...> wrote:
>
> I'm not an EMI expert,
> > I do have a 51
> > ohm series resistor on all I/O lines including the RXD and TXD of
> > USART0 to make sure that no EMI problems occur.
> >
> > regards,
> > Charles
> >   
> 
> I did do some ESD research/design once for a web tablet. Although
the IC 
> used (forget which one) was said to be ESD safe,
firing at the touch 
> screen with a Human Body Model ESD tester resulted in error. I then 
> added R/C combinations to resolve this problem.
> 
> But I'm definitely not an EMI expert and would even think that adding 
> resistors may make the problem worse ...
> Are there any special considerations to take into account with (apart 
> from PCB design and use of separate ground/power planes) ?
> 
> Regards,
> 
>     Rob
> 
> P.s: let me state that I was _not_ the original designer of that web 
> tablet :o)
>
Hi Rob,

  I am not either even though I have been designing our products to
eliminate EMI for CE approvals for the past 4 years on our products. 

  I have also used what Leon suggested as well (the AVX transguards),
but I am using some resistor packs from CTS that are 0.5mm pitch which
allows me to put them about 3mm away from the processor pins.

  We also designed a motion control stage for a large Canandian
company and they wanted 47 ohm resistors on all I/O's from a large 12
x 14 inch, multilayer board. They still had problems with the EMI, but
then we added some ferrits to the motor stages and that helped them
drop to 5db below class B.

  We do have clients (one in particular) who are and suggested adding
an resistor to each processor line to eliminate any EMI and to help
with ESD. They suggested a 30 to 51 ohm resistor on all processor
lines along with 100pf to 1nF capacitor on ports that extend off the
board. We have followed this advice 3 times and each of our boards
that went through CE, class B with no problems. There are still other
considerations like board layout, capacitor and other components to
help reduce the noise as well.

  We design allot of motion control and there is plenty of noise that
is generated that we have to take into account. One saving grace is
the CE approved power supply. We ran experiments with one at an
approved lab and found that it eliminates all the conducted emmissions
from the system if one is there. That saved us from having to put
chokes and such on our boards since we could be pumping as much as 12
amps through the board at any one time.

regards,
Charles

----- Original Message ----- 
From: "Rob Jansen" <rob@rob@...>
To: <lpc2000@lpc2...>
Sent: Saturday, January 28, 2006 6:11 AM
Subject: [lpc2000] EMI - was Re: lpc2138 5V tolerant is not really tolerant!
	> I'm not an EMI expert,
>> I do have a 51
>> ohm series resistor on all I/O lines including the RXD and TXD of
>> USART0 to make sure that no EMI problems occur.
>>
>> regards,
>> Charles
>>
>
> I did do some ESD research/design once for a web tablet. Although the IC
> used (forget which one) was said to be ESD safe, firing at the touch
> screen with a Human Body Model ESD tester resulted in error. I then
> added R/C combinations to resolve this problem.
>
> But I'm definitely not an EMI expert and would even think that adding
> resistors may make the problem worse ...
> Are there any special considerations to take into account with (apart
> from PCB design and use of separate ground/power planes) ?

Devices like AVX TransGuards can be very useful.

Leon 

---
[This E-mail has been scanned for viruses but it is your responsibility 
to maintain up to date anti virus software on the device that you are
currently using to read this email. ]

Just to chip in as well, I have never had a problem with 5V 232 chips at 
any baud-rate etc. The fact that you measured 2.3V when you used a 
series resistor indicates that you should perhaps check your 3.3V 
supply. Perhaps you have damaged these inputs because the 5V was present 
when your 3.3V supply was still ramping. The 5V tolerant inputs depend 
upon the 3.3V supply being present to operate effectively. The chips do 
work as they are spec'd, you need to look at your implementation.

Is your VDD supply at a nominal 3.3V and is it clean? Is your 5V supply 
the same? What do they look like (ramping) when power is first applied? 
I think if you scoped the inputs on both sides you can get a clear 
picture of what is really happening anyway.

*Peter*

Gus wrote:
> On my board I have LPC2138 connected to MAX232 running at 5V. I had 
> problems with UART until I added a 4.7K resistor in series. and the 
> voltage high on it now is 2.3V
>
> Are the pins real 5V tolerant? The data sheet says UART RX is 5V 
> tolerant!!!
>

I'm not an EMI expert,
> I do have a 51
> ohm series resistor on all I/O lines including the RXD and TXD of
> USART0 to make sure that no EMI problems occur.
>
> regards,
> Charles
>   

I did do some ESD research/design once for a web tablet. Although the IC 
used (forget which one) was said to be ESD safe, firing at the touch 
screen with a Human Body Model ESD tester resulted in error. I then 
added R/C combinations to resolve this problem.

But I'm definitely not an EMI expert and would even think that adding 
resistors may make the problem worse ...
Are there any special considerations to take into account with (apart 
from PCB design and use of separate ground/power planes) ?

Regards,

    Rob

P.s: let me state that I was _not_ the original designer of that web 
tablet :o)

Hello, Gus.

Saturday, January 28, 2006, 0:28:55, you wrote:

G>    On my board I have LPC2138 connected to MAX232 running at 5V. I had
G>  problems with UART until I added a 4.7K resistor in series. and the
G>  voltage high on it now is 2.3V

G>  Are the pins real 5V tolerant? The data sheet says UART RX is 5V 
G>  tolerant!!!

G>  Thanks in advance

G>  Gus

I use ADM232LJN (5V chip) with LPC2129 without any problems, but not
yet tried to hook any 5V peripherials to 2138.

By the way, which revision of chip you currently have? (just quote last
line of chip marking if you don't know where to look for it).
	-- 
Best regards,
 Yuri                          mailto:yuri@yuri...

An Engineer's Guide to the LPC2100 Series