Paul Lamar wrote:
< big snip >
Sounds good to me Finn. It would make a good article. Send me all
the stuff in crude form and I will work something up and submit it
to EAA under your name. I promise not to laugh at your code. I
don't care how good a programmer someone is. Code can ALWAYS be
improved.
The article would be a good start for someone.
BTW do you know of a single chip that will do this?
It needs to be designed to take 4 bit wide LCD data and store it
in an 8 bit wide flash memory chip. This, as shown, would take two
quad AND gates and an inverter.
I forgot. Is your LCD a four bit interface or an eight bit interface?
Paul Lamar
Forget this chip. I decided to use a PIC 16F877 instead and make the
data acquisition stand alone. It will interface to any 8 bit or a 4
bit standard LCD interface and dump data in email mode to an RS232
serial PC port using any email program such as NetScape or Internet
Exploader. It will look just like an external modem. Anybody have the
Internet communication protocol handy for email programs?
You will also be able to program it through the RS232 as to number of
analog channels you have and how often you wish to sample data using
the email program.
A $30 512K by 8 Flash memory chip would save 16 channels A/D three
digits each channel 21,845 times. At a sample rate of once a second
it would be 364 minutes or six hours. A complete cross country
flight's worth of data.
It would output the data in comma delimited ASCII form so it could
be easily imported into a spread sheet or data base program. The file
size would be several meg in comma delimited ASCII numbers and decimal
points.
Put a connector on your EFIS LCD interface so this data acquisition
system will merely plug in.
This will also work as super low cost black box data recorder in the
event of a crash to help the NTSB figure out what happened.
Paul Lamar
Paul & Group -
Most of this Single Board Computer (SBC) stuff sounds pretty easy - but
I admit I have never done it since college. (I'm an Electrical
Engineer, Programmer, Analyst) It would mostly be a matter of time,
energy, and stuff to test with (engine, sensors, eval boards, etc.)
As I've read the newsletter over the last couple months, I have been
intrigued by the out-of-box thinking that the group exhibits, even if
Paul DOES occasionally jump on someone. :) The points brought up above
are good ones. Taken one at a time ...
Forget this chip. I decided to use a PIC 16F877 instead and make the
data acquisition stand alone. It will interface to any 8 bit or a 4
bit standard LCD interface and dump data in email mode to an RS232
serial PC port using any email program such as NetScape or Internet
Exploader.
Hmm. Interesting. I like the idea of separating instruments from data
acquisition. A text data dump sounds good, but why use an Internet
protocol (like smtp or pop3, I assume)? I'd think a text dump to some
persistent memory device would be adequate. Flash ROM, cheapo memory
sticks/cards like in digital cameras, etc.
I am open to suggestions. That just looked like the easiest PC interface
thing
to do since email software is ubiquitous.
It will look just like an external modem. Anybody have the
Internet communication protocol handy for email programs?
Yeah, I've got it. If you want the smtp or pop3 protocol, let me know,
and I'll post it. Just let me know HOW DETAILED you want it. You
probably don't want the full RFC.
Send it to me private and I will filter out what I think will
get the job done in the simplest way.
You will also be able to program it through the RS232 as to number of
analog channels you have and how often you wish to sample data using
the email program.
Wouldn't something like HyperTerminal (in the Windows world) be good for
this, without involving Internet Protocol overhead/complexity?
Yes! I was wondering about that. In the good old DOS days dumb terminal
or communications programs as they were sometimes called were numerous.
Since I am not a Windows user I am not familiar with what
is out there. A com or dumb terminal program with text capture would
be ideal.
A $30 512K by 8 Flash memory chip would save 16 channels A/D three
digits each channel 21,845 times. At a sample rate of once a second
it would be 364 minutes or six hours. A complete cross country
flight's worth of data.
Then again, a memory stick/card from a digital camera would cost about
the same, last a long, long time, and you can get them anywhere. If you
design to a widespread "standard" memory device, the interface chips and
media would be dirt cheap. Granted, there's probably more up-front
design work.
I suspect interfacing it would be much harder than just the chip.
Plus you need that expensive connector while the chip can just
be soldered in. There are bigger chips available. Toshiba
has a 32 meg by 8 Flash for $82.
It would output the data in comma delimited ASCII form so it could
be easily imported into a spread sheet or data base program. The file
size would be several meg in comma delimited ASCII numbers and decimal
points.
This text data dump sounds great. Each data record should be
date/timestamped to accomodate variable sample rates. The data
aquisition device should be as simple as possible - just read-n-store.
Each flight could store a new "file", to be downloaded to a PC ...
whenever. Just like your digital pictures.
Put a connector on your EFIS LCD interface so this data acquisition
system will merely plug in.
This will also work as super low cost black box data recorder in the
event of a crash to help the NTSB figure out what happened.
Yup. If the data aquisition device (relatively tiny, light, and with
cheapo memory) is stored in a accessible, lightweight, tough,
fire-resistant box, there you have your NTSB black box. As a bonus, you
can record whatever parameters you want from your aircraft - engine,
instruments, GPS, etc., and keep them as a detailed electronic log of
your aircraft.
To the group - keep up the good work! Hopefully we can fill in the
gaps, and get the Mazda Rotary Engine to a point where builders can
really (REALLY!) purchase a reasonably priced, firewall-forward solution
from competing vendors. Capitalism is a Good Thing. I'd set target
price for a FWF solution to match that of, say, Van's, plus the PSRU.
--
Vince Orton
Chino Hills, CA, USA
Not building anything yet (RV-7 maybe?)
Lurking on ACRE Newsletter for months