Looking for digital buffer(??) or something

V

vroemm

Guest
Hi..

The signal is a pulse of 337 Khz.
The voltage is 3 volt.
Comming from a source which can give not much current.
It starts to do strange things with loads of less then 50Kohm

The goal is to switch the pin 13 of a parallel port with this signal.
I have been experimenting, with no success.

Does anyone know of a cheap easy to buy component or circuit which can do
this ??
I think its output should be a current source ??
It does not need to give a perfect pulse out.
As long as the pulse can be read by the parallel port.

Thank you very much...
..
--
...............
 
On Wed, 07 Jul 2004 23:42:16 +0200, vroemm <vroemm@wanadoo.nl> wrote:

Hi..

The signal is a pulse of 337 Khz.
The voltage is 3 volt.
Comming from a source which can give not much current.
It starts to do strange things with loads of less then 50Kohm

The goal is to switch the pin 13 of a parallel port with this signal.
I have been experimenting, with no success.

Does anyone know of a cheap easy to buy component or circuit which can do
this ??
I think its output should be a current source ??
It does not need to give a perfect pulse out.
As long as the pulse can be read by the parallel port.

Thank you very much...
.
Click to expand...
I think I'd use a comparator. you can have as high an input impedance as you like.
Go to Farnell, or Digikey or RS or wherever and search for comparator. you can get
something very cheap in an 8 pin package which will run off a 5V supply. You could
even power it off the parallel port.

However, I am doubtful that you will be able to read the parallel port fast enough
to resolve a 337kHz pulse train. But I could be wrong, it has been known.
 
vroemm wrote:
Hi..

The signal is a pulse of 337 Khz.
The voltage is 3 volt.
Comming from a source which can give not much current.
It starts to do strange things with loads of less then 50Kohm

The goal is to switch the pin 13 of a parallel port with this signal.
I have been experimenting, with no success.

Does anyone know of a cheap easy to buy component or circuit which can do
this ??
I think its output should be a current source ??
It does not need to give a perfect pulse out.
As long as the pulse can be read by the parallel port.

Thank you very much...
..
Click to expand...
Making some assumptions that the unstated low value of your pulse is at
ground, use a fet. 2N7000 comes to mind. Most any fet with threshold
below 3V will do.
Now, for "reading" the pulse. Depending on it's width and your computer
speed and the operating system and the language you use to write the
program, what about the pulse you actually need to "read",
and...and...you may have difficulty with the software.
mike

--
Return address is VALID.
Bunch of stuff For Sale and Wanted at the link below.
Toshiba & Compaq LiIon Batteries, Test Equipment
Yaesu FTV901R Transverter, 30pS pulser
Tektronix Concept Books, spot welding head...
http://www.geocities.com/SiliconValley/Monitor/4710/
 
Hi.

Try a comparator, or an op-amp in follower mode. Both will have very
high impedance inputs, and can be used to drive the parallel pin
directly, or through a transistor or logic gate.

Yours, Mark.

vroemm wrote:
Hi..

The signal is a pulse of 337 Khz.
The voltage is 3 volt.
Comming from a source which can give not much current.
It starts to do strange things with loads of less then 50Kohm

The goal is to switch the pin 13 of a parallel port with this signal.
I have been experimenting, with no success.

Does anyone know of a cheap easy to buy component or circuit which can do
this ??
I think its output should be a current source ??
It does not need to give a perfect pulse out.
As long as the pulse can be read by the parallel port.

Thank you very much...
.
Click to expand...
 
mike wrote:

vroemm wrote:
Hi..

The signal is a pulse of 337 Khz.
The voltage is 3 volt.
Comming from a source which can give not much current.
It starts to do strange things with loads of less then 50Kohm

The goal is to switch the pin 13 of a parallel port with this signal.
I have been experimenting, with no success.

Does anyone know of a cheap easy to buy component or circuit which can do
this ??
I think its output should be a current source ??
It does not need to give a perfect pulse out.
As long as the pulse can be read by the parallel port.

Thank you very much...
..

Making some assumptions that the unstated low value of your pulse is at
ground, use a fet. 2N7000 comes to mind. Most any fet with threshold
below 3V will do.
Now, for "reading" the pulse. Depending on it's width and your computer
speed and the operating system and the language you use to write the
program,
Click to expand...
I thought it was fast enough..
Just tested it better.

The pulse is 0.71 microseconds.
My fasted computer, celeron, can only do 1.2 microseconds for 1 read.
To slow :-( :)
There goes the plan :)

I'll have to see if there is a not to difficult way to make 1 output puls
for every 5 to 10 input pulses, or something like that.
Else, i do not know..

--
...............
 
vroemm wrote:
mike wrote:


vroemm wrote:

Hi..

The signal is a pulse of 337 Khz.
The voltage is 3 volt.
Comming from a source which can give not much current.
It starts to do strange things with loads of less then 50Kohm

The goal is to switch the pin 13 of a parallel port with this signal.
I have been experimenting, with no success.

Does anyone know of a cheap easy to buy component or circuit which can do
this ??
I think its output should be a current source ??
It does not need to give a perfect pulse out.
As long as the pulse can be read by the parallel port.

Thank you very much...
..

Making some assumptions that the unstated low value of your pulse is at
ground, use a fet. 2N7000 comes to mind. Most any fet with threshold
below 3V will do.
Now, for "reading" the pulse. Depending on it's width and your computer
speed and the operating system and the language you use to write the
program,


I thought it was fast enough..
Just tested it better.

The pulse is 0.71 microseconds.
My fasted computer, celeron, can only do 1.2 microseconds for 1 read.
To slow :-( :)
There goes the plan :)

I'll have to see if there is a not to difficult way to make 1 output puls
for every 5 to 10 input pulses, or something like that.
Else, i do not know..
Click to expand...
Exactly what is it you're trying to measure? How accurate do you need
the answer to be?

--
Return address is VALID.
Bunch of stuff For Sale and Wanted at the link below.
Toshiba & Compaq LiIon Batteries, Test Equipment
Yaesu FTV901R Transverter, 30pS pulser
Tektronix Concept Books, spot welding head...
http://www.geocities.com/SiliconValley/Monitor/4710/
 
mike wrote:

Exactly what is it you're trying to measure?
Click to expand...
I found a PM2544 logic multimeter on a secondhand market.
The LCD screen is partly black, the number and signs are unreadable.

I have connected the signal to the LCD screen to the pin 13 of the parrallel
port.
This way the puls train which contains the data to be displayed at the LCD
can be read into the computer.
The pulses of this signal are MUCH longer.
This allready works.

But the puls train is unstable in timing.
It is on a msdos 486 computer, which i always thought where perfect real
time.
Here is a picture of the signals read with the parallel port:
http://home.wanadoo.nl/hihihi/pm2544/delay2000-ohm-vdc-vac-amp-2.png

Then i found the puls signal mentioned in my previous messages.
I hoped this signal would have the same difference in timing.
Perhaps then the X'th puls would always come at the same time as the Y'th
bit in the data puls train.
Like in a I2C signal.

But for now i have abandoned the puls signal mentioned in my previous
mails..
I found new ways to extract the data from a unstable puls train.
Like searching the train for a sequence of 7 bit which together make up a
valid number.
And searching for the longer puls which comes before most number.


--
...............
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top