Hello Guest it is April 23, 2024, 12:58:23 PM

Author Topic: input/output and pulse analysis probe.  (Read 3701 times)

0 Members and 1 Guest are viewing this topic.

Offline Perfo

*
  •  143 143
    • View Profile
input/output and pulse analysis probe.
« on: August 12, 2008, 06:09:43 AM »
Hello all.
With the help of Mr Leed3 I'm hoping to develop a small DIY type probe that can be connected to a spare input and will enable the user to place the tip of the probe on any other input or output and the PC screen will display if the input is high, low or if there is a pulse train on it (maybe even a graph showing the pulse train so pulse width etc can be viewed). First off, in practice how fast can data be read from the parallel port (I know this is dependent on machine spec but on average)? Is there already a plug-in or mod that does this? I suppose the last question is would anyone else find this a handy addition? I like to measure things when setting up or looking for faults and thus this is why I thought of it but maybe it's only me  :). All comments will be appreciated.

Offline Jeff_Birt

*
  •  1,107 1,107
    • View Profile
    • Soigeneris
Re: input/output and pulse analysis probe.
« Reply #1 on: August 13, 2008, 12:42:07 PM »
The purpose of such a probe would be to assist in troubleshooting, no? When one has a problem with an I/O pin the first step is to determine if the signal is getting to/from the I/O device. This helps determine if the fault lies in the I/O device (LPT/Mach in our case) or the external device (machine). So you have a 50/50 shot at the problem being on either side of the I/O device/external device connection.

What your proposing is using a pin on the very I/O device that you need to test to test the I/O device itself (the LPT). Sorry to say, you can't do that. Your testing apparatus has to be in a know good working state before you can rely on it to test anything. And, a testing apparatus cannot test itself!

If your probe was to find that a certain I/O pin was not acting properly how do you determine if the fault is the configuration of the port itself or some external device?
Happy machining , Jeff Birt
 

Offline Perfo

*
  •  143 143
    • View Profile
Re: input/output and pulse analysis probe.
« Reply #2 on: August 14, 2008, 07:17:24 AM »
Good points.  You can prove if the probe is working or not by touching it on a 5V supply somewhere it could even include a couple of LED's to perform a self test type thing is one of the probe cables will be connected to gnd one to a supply and one to an input so touching the tip on a supply should light one LED and touching it on GND should light the other and tell you the probe is working. If you then touch a known high pin on the port that isn't showing high on the PC then the ports is malfunctioning. I also get your point about how do you know if it is the device port or BOB that's not working. I propose once the self check has been carried out then you can check the port and the BOB with this device and through process of elimination it will ether be the device connected tot he port or the software/set up depending on if the bob etc pass the test. This also works the other way you should be able to touch the pin relative to the input tot he bob that corresponds to the limits and see it change when the limits operate then knowing there is something wrong with the set up in Mach if it still says you have limit /stop problem. I would hope you could also see if noise is causing your limit problem by displaying the wave form of the incoming signal but I've got to test this function yet. similarly I was hoping to do some fast Fourier analysis on it to give a spectrum analysis type graph telling you at a glance you have a DC component (expected and a say 3KHz compent (unexpected). My electronics and software knowledge is a little dated as I haven't done much in the last 18 years since I left My job where I did a lot of design work. However if the port idea doesn't work then the same probe with Little adjustment can be used on a mic or any audio input to the PC and data collected that way and thus eliminating any combined errors due to using the same port to test and control. The only real problem with that is when I used it before to make a PC based oscilloscope back in the 80's the highest sample rate was around 10KHz with a top of the range sound card which ment you could only reliably show wave forms up to 5KHz. Now I believe the standard sound card will go to 44KHz (the CD standard sample rate) and thus we could reliably show at least 22KHz this would need more investigation on my behalf. Thanks for your comments. The electronics for this comes to around £1 and is really simple to construct so quite do able for anyone fiddling with Bob's and Mach 3 etc.