Need Some Grounding on "GROUNDING"

[[email protected]]

I recently decided to undertake my first shot at a micro-controller
project, just for fun, to learn more about the little beasts.
Something simple, like a digital speedometer for the car. Okay, I can
hear the laughter, but I thought DSP was a little too ambitious for my
first project.

The design is very straight forward, a Microchip PIC uC, a 3 digit LED
display, the car's VSS signal, and a separate power supply (12V dc- 5V
dc converter, beefy enough to handle future auto projects).

The software is really straight forward, count some pulses, a little
time slicing, a quick binary to bcd conversion, and voila, digital
speed readout...

My concern is in the dc and signal ground paths. I'm not sure, but I
don't think the dc-dc converters negative terminal is really tied to
the vehicles ground system at all. So, what does the PIC see on the
VSS input line, since that signal is referenced to the vehicles
ground? I'm aware I could get around this with a different power
supply option, like a voltage regulator tied to a series pass
transistor from the cars +12V (13.whatever), and the ground would be
common to both the VSS and the power supply. But, I'd like to try and
make it work using the dc-dc converter. So, I would most happily
entertain any suggestion you folks may have regarding the proper way
to tie these pieces together.

As you can tell, I don't have an extensive background in electronics,
but I'm ready to expand my knowledge.....and who knows, for my next
project, maybe a TFT display driven from my little PIC (just kidding,
but I had to go there after reading some other postings)

Thanks!

John Clement

 

[scada]

I recently decided to undertake my first shot at a micro-controller
project, just for fun, to learn more about the little beasts.
Something simple, like a digital speedometer for the car. Okay, I can
hear the laughter, but I thought DSP was a little too ambitious for my
first project.

The design is very straight forward, a Microchip PIC uC, a 3 digit LED
display, the car's VSS signal, and a separate power supply (12V dc- 5V
dc converter, beefy enough to handle future auto projects).

The software is really straight forward, count some pulses, a little
time slicing, a quick binary to bcd conversion, and voila, digital
speed readout...

My concern is in the dc and signal ground paths. I'm not sure, but I
don't think the dc-dc converters negative terminal is really tied to
the vehicles ground system at all. So, what does the PIC see on the
VSS input line, since that signal is referenced to the vehicles
ground? I'm aware I could get around this with a different power
supply option, like a voltage regulator tied to a series pass
transistor from the cars +12V (13.whatever), and the ground would be
common to both the VSS and the power supply. But, I'd like to try and
make it work using the dc-dc converter. So, I would most happily
entertain any suggestion you folks may have regarding the proper way
to tie these pieces together.

As you can tell, I don't have an extensive background in electronics,
but I'm ready to expand my knowledge.....and who knows, for my next
project, maybe a TFT display driven from my little PIC (just kidding,
but I had to go there after reading some other postings)

Thanks!

John Clement

Using a DC to DC converter is a good idea! There can be lots of noise spikes
on the 12V battery. As far as the PIC input goes, I would use an
opticIsolator IC. The input goes to an LED inside the chip (powered by the
cars 12V, the output is a transistor (powered by the converter). 4N28 is one
of many numbers out there.

 

[Fritz Schlunder]

I recently decided to undertake my first shot at a micro-controller
project, just for fun, to learn more about the little beasts.
Something simple, like a digital speedometer for the car. Okay, I can
hear the laughter, but I thought DSP was a little too ambitious for my
first project.

The design is very straight forward, a Microchip PIC uC, a 3 digit LED
display, the car's VSS signal, and a separate power supply (12V dc- 5V
dc converter, beefy enough to handle future auto projects).

The software is really straight forward, count some pulses, a little
time slicing, a quick binary to bcd conversion, and voila, digital
speed readout...

My concern is in the dc and signal ground paths. I'm not sure, but I
don't think the dc-dc converters negative terminal is really tied to
the vehicles ground system at all. So, what does the PIC see on the
VSS input line, since that signal is referenced to the vehicles
ground? I'm aware I could get around this with a different power
supply option, like a voltage regulator tied to a series pass
transistor from the cars +12V (13.whatever), and the ground would be
common to both the VSS and the power supply. But, I'd like to try and
make it work using the dc-dc converter. So, I would most happily
entertain any suggestion you folks may have regarding the proper way
to tie these pieces together.

As you can tell, I don't have an extensive background in electronics,
but I'm ready to expand my knowledge.....and who knows, for my next
project, maybe a TFT display driven from my little PIC (just kidding,
but I had to go there after reading some other postings)

Thanks!

John Clement

What is wrong with your post? It doesn't seem to automatically include the
reply carats...

What kind of DC-DC converter is this? Is it a little pre-packaged module?
Does it have a part number and technical specifications you can point us
towards?

If this is a pre-packaged DC-DC converter module then it may have a fully
isolated output. If this is true then all you likely need to do is hook the
most negative pin of the DC-DC converter output to the vehicle ground
(normally the chassis), and voila, mission accomplished, the PIC will
receive the signal properly.

Without making this extra connection the PIC would simply see what appears
to be a more or less floating input.

If your DC-DC converter doesn't have a fully isolated output, but instead
you mean it is of some other topology (IE: inverting, such that it takes a
positive input voltage and produces a negative output voltage with respect
to the same ground, but then you use the ground as the PIC's positive power
rail and the negative voltage as the PIC's ground rail), then you may need
to use some technique such as optoisolators or somesuch as already been
suggested.

Rock on. A digital speedometer is a good little project.

 

[Tim Auton]

news:[email protected]... [snip]

What is wrong with your post? It doesn't seem to automatically include the
reply carats...

There's something wrong with your newsreader, as it's that that
includes (or doesn't) the quotation characters. What's wrong with it
is that it's Outlook Express, it does that every now and then. Use
something better.


Tim

 

[[email protected]]

What is wrong with your post? It doesn't seem to automatically include the
reply carats...

What kind of DC-DC converter is this? Is it a little pre-packaged module?
Does it have a part number and technical specifications you can point us
towards?

If this is a pre-packaged DC-DC converter module then it may have a fully
isolated output. If this is true then all you likely need to do is hookthe
most negative pin of the DC-DC converter output to the vehicle ground
(normally the chassis), and voila, mission accomplished, the PIC will
receive the signal properly.

Without making this extra connection the PIC would simply see what appears
to be a more or less floating input.

If your DC-DC converter doesn't have a fully isolated output, but instead
you mean it is of some other topology (IE: inverting, such that it takesa
positive input voltage and produces a negative output voltage with respect
to the same ground, but then you use the ground as the PIC's positive power
rail and the negative voltage as the PIC's ground rail), then you may need
to use some technique such as optoisolators or somesuch as already been
suggested.

Rock on. A digital speedometer is a good little project.

Thanks for the response Fritz...This is my first posting to a news
group and I am using the Agent news reader. I probably don't have it
configured properly just yet....It's a work in progress, like the rest
of my life ;-)

To answer your question, I am using a commercial dc-dc converter made
by Weidmuller, specifically the 9919371205 model. The tech specs can
be viewed at the following web site:
http://www.weidmuller.com/eprise/ma...ducts/whatsnewinus/datasheets/CP-DCDC_50W.pdf
Based on what I can glean from the specs, I suspect the output is
isolated, but I would enjoy your opinion if you have the time to look
at the specs.

I had originally bread boarded a couple of approaches to getting the
12 volt VSS input to the PIC via an op-amp buffer, but the results
were not too good. I also tried a commercial input interface, but it
seemed to draw more current than the VSS could withstand (around 20
ma). Not knowing exactly what is on the other end of the VSS wire
caused me some concern relating to current draw. I believe the PIC has
clamping diodes on the port (when used as input) which would take care
of the (relatively) high input voltage, but still present a high
impedance which would not load dow the VSS line.

Thanks again!
John

 

[[email protected]]

I recently decided to undertake my first shot at a micro-controller
project, just for fun, to learn more about the little beasts.
Something simple, like a digital speedometer for the car. Okay, I can
hear the laughter, but I thought DSP was a little too ambitious for my
first project.

The design is very straight forward, a Microchip PIC uC, a 3 digit LED
display, the car's VSS signal, and a separate power supply (12V dc- 5V
dc converter, beefy enough to handle future auto projects).

The software is really straight forward, count some pulses, a little
time slicing, a quick binary to bcd conversion, and voila, digital
speed readout...

My concern is in the dc and signal ground paths. I'm not sure, but I
don't think the dc-dc converters negative terminal is really tied to
the vehicles ground system at all. So, what does the PIC see on the
VSS input line, since that signal is referenced to the vehicles
ground? I'm aware I could get around this with a different power
supply option, like a voltage regulator tied to a series pass
transistor from the cars +12V (13.whatever), and the ground would be
common to both the VSS and the power supply. But, I'd like to try and
make it work using the dc-dc converter. So, I would most happily
entertain any suggestion you folks may have regarding the proper way
to tie these pieces together.

As you can tell, I don't have an extensive background in electronics,
but I'm ready to expand my knowledge.....and who knows, for my next
project, maybe a TFT display driven from my little PIC (just kidding,
but I had to go there after reading some other postings)

Thanks!

John Clement

Using a DC to DC converter is a good idea! There can be lots of noise spikes
on the 12V battery. As far as the PIC input goes, I would use an
opticIsolator IC. The input goes to an LED inside the chip (powered by the
cars 12V, the output is a transistor (powered by the converter). 4N28 isone
of many numbers out there.

Thanks for your suggestion, I actually tried a similar approach, but
had some concerns about the current draw on the VSS line...See my
reply to Fritz for more details.

Thanks!
John

 

[Howard Henry Schlunder]

There's something wrong with your newsreader, as it's that that
includes (or doesn't) the quotation characters. What's wrong with it
is that it's Outlook Express, it does that every now and then. Use
something better.

Rather than every now and then, it is more accurate to say that Outlook Express does not add quotation markers whenever replying to a message posted with the "Content-Transfer-Encoding" header set to "quoted-printable". Does your Forte Agent software properly quote if you try to reply to my message? For the longest time, Forte Agent (or at least the free version) didn't handle quote-printable messages at all, so I'm really happy to discover that they finally added support for section 6.7 of RFC2045 [ http://www.faqs.org/rfcs/rfc2045.html ]. That would be absolutely fabulous if Forte's implementation exceeds Outlook Express's (which has supported it for a long time, albiet with the quoting limitation). I'm tired of darn unintended line wrapping!

 

[John Clement]

There's something wrong with your newsreader, as it's that that
includes (or doesn't) the quotation characters. What's wrong with it
is that it's Outlook Express, it does that every now and then. Use
something better.

Rather than every now and then, it is more accurate to say that Outlook Express does not add quotation markers whenever replying to a message posted with the "Content-Transfer-Encoding" header set to "quoted-printable". Does your Forte Agent software properly quote if you try to reply to my message? For the longest time, Forte Agent (or at least the free version) didn't handle quote-printable messages at all, so I'm really happy to discover that they finally added support for section 6.7 of RFC2045 [ http://www.faqs.org/rfcs/rfc2045.html ]. That would be absolutely fabulous if Forte's implementation exceeds Outlook Express's (which has supported it for a long time, albiet with the quoting limitation). I'm tired of darn unintended line wrapping!

So, is there a penalty for a first-time poster who obviously did not
research the importance of the quotation markers to the general
populous?

 

[Howard Henry Schlunder]

So, is there a penalty for a first-time poster who obviously did not
research the importance of the quotation markers to the general
populous?

Penalty? No. He deserves a big bravo for posting using the (relatively) modern format that he did.

Ordinarily, news posting software causes forced line wraps to occur (usually around 76-80 characters per line) before messages get sent to the Internet. This allows newsgroup messages to be human displayable without doing any calculations or reformating on the newreader's computer. By using Quoted-Printable, original text format and line feed positions defined by the author are preserved and no forced line wraps are added. As such, it becomes the newreading client's job to determine how the message should be displayed.

Computers and the Internet have advanced so much in the last couple of decades, and there is absolutely no reason why we should still be composing messages for the lowest common denominator of newsreaders. Back when text mode terminals with 80 column by 25 line displays and no processing power of their own were common, it made sense posting messages with line wrapping already embedded to make it display properly on those consoles. Now days, Palm computers and even calculators have phenominal amount of processing power and other resources available for them to reformat messages on the device for optimal displaying. The overhead required to reformat a quoted-printable message probably would have been quite severe back in the old days, but doing so today is trivial.

My personal wish is that eveyone would switch to using Quoted-Printable or even better an HTML varient with only a few tags defined, similar to what operators of many web based forums allow. People need to wake up and collectively agree that the following quotation is ugly, unacceptable, and fixable simply by adopting newer standards.

 

[Howard Henry Schlunder]

...composing messages for the lowest common denominator of newsreaders.

Oops, that should be: the lowest *possible* denominator of newsreaders.

 

[John Clement]

in message Oops, that should be: the lowest *possible* denominator of newsreaders.

I agree with your assesment....Sign me "The Accidental Visionary"

 

[Tim Auton]

in message
Rather than every now and then, it is more accurate to say that Outlook Express
does not add quotation markers whenever replying to a message posted with the
"Content-Transfer-Encoding" header set to "quoted-printable".

Ahh, I'd never worked out quite why it was.

Does your
Forte Agent software properly quote if you try to reply to my message?

Not quite It didn't word-wrap it, but did properly quote the one
enormous line. Word wrapping is usually done by the posting
newsreader, isn't it? It always was when I were a lad Perhaps
quoted-printable is supposed to deal the wrapping and Agent isn't
doing it right.


Tim

 

[Richard Mason]

On Wed, 7 Jan 2004, Howard Henry Schlunder wrote in

My personal wish is that eveyone would switch to using Quoted-Printable
or even better an HTML varient with only a few tags defined, similar to
what operators of many web based forums allow. People need to wake up
and collectively agree that the following quotation is ugly,
unacceptable, and fixable simply by adopting newer standards.

Please, no HTML.

I don't use OE so I wouldn't know what 'Quoted-Printable' is but I do
know that people use Quotefix http://jump.to/outlook-quotefix to fix
their OE quoting problems.

 

[Howard Henry Schlunder]

On Wed, 7 Jan 2004, Howard Henry Schlunder wrote in


Please, no HTML.

I don't use OE so I wouldn't know what 'Quoted-Printable' is but I do
know that people use Quotefix http://jump.to/outlook-quotefix to fix
their OE quoting problems.

Quoted-Printable is an encoding standard of the Internet defined by RFC2045. Although OE implements it, it is not in any way associated with Microsoft.

Quotefix looks like an awesome program for those using Outlook, but it will not work for users of Outlook Express. The lack of support for OE is surely because Outlook Express doesn't expose any COM objects which can be manipulated automatically by other programs. As a result, one would have to manually copy all text out of the composition window into the fix-it program, and then manually copy it back. An alternative fix-it program would be implemented as a proxy server, fixing messages before they leave the computer, but would have the highly undesirable effect of not being WYSIWYG editable. Nonetheless, I'm really tempted to write such a fix-it program since I use OE so much and do find it somewhat limited.

Howard Henry Schlunder

 

[Howard Henry Schlunder]

To answer your question, I am using a commercial dc-dc converter made
by Weidmuller, specifically the 9919371205 model. The tech specs can
be viewed at the following web site:
http://www.weidmuller.com/eprise/ma...ducts/whatsnewinus/datasheets/CP-DCDC_50W.pdf
Based on what I can glean from the specs, I suspect the output is
isolated, but I would enjoy your opinion if you have the time to look
at the specs.

I don't see anyone who replied, so I'd like to say that according to the specs, you have some 500Vac of galvanic isolation between the input and output. This means that, as you suspected, your output is isolated and your solution should therefore be as simple as connecting the two grounds togethor.

 

[Richard Mason]

On Thu, 8 Jan 2004, Howard Henry Schlunder wrote in

Quotefix looks like an awesome program for those using Outlook, but it
will not work for users of Outlook Express.

Sorry, wrong link. OE-Quotefix
http://home.in.tum.de/~jain/software/oe-quotefix/downloads.php

 

[John Woodgate]

I read in sci.electronics.design that Richard Mason

On Wed, 7 Jan 2004, Howard Henry Schlunder wrote in


Please, no HTML.

I don't use OE so I wouldn't know what 'Quoted-Printable' is but I do
know that people use Quotefix http://jump.to/outlook-quotefix to fix
their OE quoting problems.

There simply isn't any point in using quoted-printable and even less for
using HTML for either mail or news.

Unless you are/were one of those artistic types who decorate their
handwritten letters with borders, colours and/or sketches.

 

[Howard Henry Schlunder]

On Thu, 8 Jan 2004, Howard Henry Schlunder wrote in


Sorry, wrong link. OE-Quotefix
http://home.in.tum.de/~jain/software/oe-quotefix/downloads.php

Oh how neat. Thanks for the link.

 

[Howard Henry Schlunder]

There simply isn't any point in using quoted-printable and even less
for using HTML for either mail or news.

Unless you are/were one of those artistic types who decorate their
handwritten letters with borders, colours and/or sketches.

I agree that there is no reason to support borders, and colors should be overridable by news reading clients, but that isn't why I support a newer standard or quoted-printable. With quoted-printable, there is no such thing as a forced and unintended line wrap. You can post source code with long lines in it, and you can immediately copy it out of your favorite news client and paste it into your IDE and it will work. Links can be as long as they come and not get destroyed by a carriage return/line feed. Ascii art and schematics can be as wide as needed. When quoting quoted messages, words don't move down to the next line and become effectively unreadable. And, also important, it allows the newsreaders to fully expand text to fill whatever window pane size the user is using. At 76 to 80 characters per line, you waste a bunch of screen real estate on higher resolution displays.

Actually though, since the onset of this discussion, I'm getting the feeling that it may just be that Outlook Express sucks and other newsreaders may not have as many problems. If that is true, quoted-printable is still desireable since it helps remove the guesswork the newsreader has to do before formatting the message for the client computer, though.

Howard Henry Schlunder

 

[Richard Mason]

On Thu, 8 Jan 2004, Howard Henry Schlunder wrote in

Actually though, since the onset of this discussion, I'm getting the
feeling that it may just be that Outlook Express sucks

You're right there.

and other newsreaders may not have as many problems. If that is true,
quoted-printable is still desireable since it helps remove the
guesswork the newsreader has to do before formatting the message for
the client computer, though.

Mine doesn't.