data logger project- where to start?

[fugu]

I want to make a simple data logger that will monitor 8 switch
closures/5V logic levels. When any of the switches changes state, I want
to record the time of day (to 0.01 sec. if possible) and the status of all
8 switches. I expect switches to change state no more than once every
100ms or so.

I am thinking this should be doable using the parallel port on a PC, and
the PCs built-in RTC. I'd like the data to either be saved as a text file
or have it go directly into a spreadsheet (Excel or Open Office equivalent).

I am not much of a PC guy in either hardware or programming, but I
can learn. Where would I start to get this sort of thing going? Can the
programming be done within Excel or do I need something like Visual Basic,
or is there some other code development system I'd need? I have a bunch
of info on the parallel port, but the programming examples for accessing
it all look like they are written in C, which I do not yet understand. I
have programmed uCs in assembly language, so I have no problem learning to
program things.

Any suggestions?

Thanks,
F

 

[martin griffith]

I want to make a simple data logger that will monitor 8 switch
closures/5V logic levels. When any of the switches changes state, I want
to record the time of day (to 0.01 sec. if possible) and the status of all
8 switches. I expect switches to change state no more than once every
100ms or so.

I am thinking this should be doable using the parallel port on a PC, and
the PCs built-in RTC. I'd like the data to either be saved as a text file
or have it go directly into a spreadsheet (Excel or Open Office equivalent).

I am not much of a PC guy in either hardware or programming, but I
can learn. Where would I start to get this sort of thing going? Can the
programming be done within Excel or do I need something like Visual Basic,
or is there some other code development system I'd need? I have a bunch
of info on the parallel port, but the programming examples for accessing
it all look like they are written in C, which I do not yet understand. I
have programmed uCs in assembly language, so I have no problem learning to
program things.

Any suggestions?

Thanks,
F

couple of random ideas,
have a look at the AVR butterfly demo board, might fit the bill
and have a look at python, a scripting thingummy for PC's, its seems
rather good, and it doesnt seem to need any EULAs from micro$oft
and also pyserial, guess what that does.....


martin

 

[linnix]

I want to make a simple data logger that will monitor 8 switch
closures/5V logic levels. When any of the switches changes state, I want
to record the time of day (to 0.01 sec. if possible) and the status of all
8 switches. I expect switches to change state no more than once every
100ms or so.

You would need to run it in DOS and disable interrupts.

I am thinking this should be doable using the parallel port on a PC, and
the PCs built-in RTC. I'd like the data to either be saved as a text file
or have it go directly into a spreadsheet (Excel or Open Office equivalent).

You would need to find a PC that can be set to 8 bits parallel input
modes. Older one can't be set to 8 bits input. Newer one sometimes
don't have parallel port.

I am not much of a PC guy in either hardware or programming, but I
can learn. Where would I start to get this sort of thing going? Can the
programming be done within Excel

I am not aware of any input/output function in Excel.

or do I need something like Visual Basic,

Could be done, but ...

or is there some other code development system I'd need?

Visual C++ OCX.

I have a bunch
of info on the parallel port, but the programming examples for accessing
it all look like they are written in C, which I do not yet understand. I
have programmed uCs in assembly language, so I have no problem learning to
program things.

Any suggestions?

Yes, XC9536 to serialize the switches. I can send you the
preprogrammed chip.

 

[Luhan]

I want to make a simple data logger that will monitor 8 switch
closures/5V logic levels. When any of the switches changes state, I want
to record the time of day (to 0.01 sec. if possible) and the status of all
8 switches. I expect switches to change state no more than once every
100ms or so.

I am thinking this should be doable using the parallel port on a PC, and
the PCs built-in RTC. I'd like the data to either be saved as a text file
or have it go directly into a spreadsheet (Excel or Open Office equivalent).

I am not much of a PC guy in either hardware or programming, but I
can learn. Where would I start to get this sort of thing going? Can the
programming be done within Excel or do I need something like Visual Basic,
or is there some other code development system I'd need? I have a bunch
of info on the parallel port, but the programming examples for accessing
it all look like they are written in C, which I do not yet understand. I
have programmed uCs in assembly language, so I have no problem learning to
program things.

Any suggestions?

Hi,

Here is my version...

http://members.cox.net/berniekm/spydar.html

Luhan

 

[Rich Grise]

Here is my version...

http://members.cox.net/berniekm/spydar.html

It only has 6 inputs. )-;

Thanks,
Rich

 

[Donald]

It only has 6 inputs. )-;

Thanks,
Rich

Rich,

Did you look at the schematic at all ?

The 8 inputs are there, just connect the lines !!!


donald

 

[Jan Panteltje]

I want to make a simple data logger that will monitor 8 switch
closures/5V logic levels. When any of the switches changes state, I want
to record the time of day (to 0.01 sec. if possible)

The 10mS timing will be probelmatic, even with the PC connected to the
internet, and synchronized with a timeserver, you may see the PC real
time clock drift say >10 seconds a day.
Also you would need fast interrupt response from the PC.

So you would need an atomic clock, on a separate controller board,
sending the data at normal speed via for example RS232 in a format
that holds time and switch status everytime a switch is pressed.
If switches can be pressed within milliseconds of each other then
you need to store and chain the data localy on that board, so it needs memory.

Not difficult to make, simple PIC can do it, except for the atomic clock.

Maybe somebody else will have away around the clock problem.

 

[Jan Panteltje]

Maybe somebody else will have away around the clock problem.

Ah, add a radio clock perhaps.

 

[budgie]

The 10mS timing will be probelmatic, even with the PC connected to the
internet, and synchronized with a timeserver, you may see the PC real
time clock drift say >10 seconds a day.
Also you would need fast interrupt response from the PC.

So you would need an atomic clock, on a separate controller board,
sending the data at normal speed via for example RS232 in a format
that holds time and switch status everytime a switch is pressed.
If switches can be pressed within milliseconds of each other then
you need to store and chain the data localy on that board, so it needs memory.

Not difficult to make, simple PIC can do it, except for the atomic clock.

Maybe somebody else will have away around the clock problem.

GPS receiver.

 

[fugu]

The 10mS timing will be probelmatic, even with the PC connected to the
internet, and synchronized with a timeserver, you may see the PC real
time clock drift say >10 seconds a day.
Also you would need fast interrupt response from the PC.

So you would need an atomic clock, on a separate controller board,
sending the data at normal speed via for example RS232 in a format
that holds time and switch status everytime a switch is pressed.
If switches can be pressed within milliseconds of each other then
you need to store and chain the data localy on that board, so it needs memory.

Not difficult to make, simple PIC can do it, except for the atomic clock.

Maybe somebody else will have away around the clock problem.

I can work the clock issue with a PIC. I don't need accurate absolute
time, just relative intervals. This is very easy to do with a PIC and one
of its built-in timers. I can use a 16 bit counter value as the time
indicator. The events I want to time will occur in under a minute, so 16
bits is plenty for a 0.01 second intervals.

I've looked into RTC in a PC and it seems like the hardware clock is only
able to indicate whole seconds, and the software clock generates
interrupts in 58.xx ms intervals, also not very useful.

So now it looks like I need to send data from the PIC to a PC and store
the data into a text file. Maybe the PIC has to buffer the data if the
logic levels change state quickly.

Thanks,
F

 

[fugu]

Hi,

Here is my version...

http://members.cox.net/berniekm/spydar.html

Luhan

The is the closest thing I have seen to what I am trying to do. How is
the timestamping done? Are you using the PCs RTC or are you generating an
elapsed time value in the PIC?

Thanks,
F

 

[Donald]

The is the closest thing I have seen to what I am trying to do. How is
the timestamping done? Are you using the PCs RTC or are you generating an

First you don't look at the schematic, now you don't bother to look at
the code given on the web site.

donald

 

[fugu]

First you don't look at the schematic, now you don't bother to look at
the code given on the web site.

donald

plonk!
F

 

[Donald]

plonk!
F

Hmmm, that was easy.

donald

 

[Rich Grise]

Rich,

Did you look at the schematic at all ?

No, I was going by:
"The unit is based on a PIC16F84. Six digital data lines are brought in
via a 74hc165 shift regiser."

The 8 inputs are there, just connect the lines !!!

OK, I see it now.

Thanks,
Rich