Computer Display?

[Ramius]

Hi All!
I have a sensor that when I add about 1 meter of cable to the readings become wrong! I think the main problem is the pulses from the sensor are too fast so the PIC can only be a short distance. It would be great if there was a program I could download and display the results on the screen or maybe a different cable type than just shielded twisted wires?

[Normnet]

Hi All!
I have a sensor that when I add about 1 meter of cable to the readings become wrong! I think the main problem is the pulses from the sensor are too fast so the PIC can only be a short distance. It would be great if there was a program I could download and display the results on the screen or maybe a different cable type than just shielded twisted wires?

Hyperterminal is one program which will display RS232 protocol.
It is sometimes found in the Windows menu under Start>Programs>Windows Accessories>Communications.
If not it can be downloaded.
Another is TeraTerm.

You will need a PC with a serial port or for newer PC's use a USB virtual serial port adapter.
For longer distances a converter to and from RS485 with RS232 at both ends works well.

You didn't say the baud rate in use.

Others will have more info as well.


Norm

[Ramius]

Thanks Norm!
I do have a "newer computer with both USB and RS-232. If I look at the pulse width from the sensor with a Saleae Logic Analyser the pulse width from the sensor is very short an looks almost like it is at 2.4 MHZ although not closely spaced! The sensor clock is at 32,768. Since I am only displaying the data from the sensor I would think that a baud rate of 9,600 should be good? Don't really know for certain, never ran into this before. If I keep all the wires short (1' or less) everything seems to work fine. I just need to go a longer distance to read the sensor without having to build something complex and involved.

Best, Ed

[rsocor01]

If I keep all the wires short (1' or less) everything seems to work fine.

Hi. RS-232 should work for a distance much greater than 1'. I suspect something else is going on. What type of protocol are you using to communicate between the PIC and the sensor? Maybe noise in the line????

Robert

[HenrikOlsson]

Hi Ed,
A link to the datasheet for the sensor in question, or at least a part number would be a good idea to post. Somehow Norm knew or assumed its output was RS232, something you didn't say in your first post, so my question is: Is it?

If the sensor is indeed outputting async serial it may be doing so at TTL levels and your PC may not like that (even though it usually works for short distances) so you may need a MAX232 driver or something.

A diagram showing how you have it connected would be nice as well.

But again, without even knowing what type of sensor you're talking about it's just guesses.

/Henrik.

[Ramius]

Hi Henrik!
The sensor is the infamous MS5541-CM and I believe the protocol is SPI. Basically the sensor to a 3v to 5v converter to the PIC then from the PIC to a MAX232-cpe to the computer. At short distances from the 3-5v converter the sensor readings are "normal", that is you will accurately read room temperature for example. As the cable from the 3-5v converter becomes longer the room temperature reading becomes abnormal such as -1434C. First cable was a shielded 6-wire, 22 gauge cable. Then I tried a ribbon cable with signals on every other wire with ground at one end only for the wires inbetween the signal wires to avoid cross talk and any ground loops.

Best, Ed

[richard]

this doco from ti may help

http://www.ti.com/lit/an/slyt441/slyt441.pdf

[Ramius]

Hi Guys!
Thanks and I actully did place some .1 caps across the 3v and 5v side. Just added a 100mfd to both the 3v and 5v side of the TXB0104. How things are built is there is a small proto board about 4" from the pic which has the TXB0101. Then the sensor connects to the same board. I would love to be able to add a .1 on the back side of the sensor with the 47mfd cap. and just connecting what I have connected has been a real challenge. Damn stuffs too small! Lol Best, Ed