Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - evanh

Pages: 1 2 3 [4]
46
Technical support / Re:TL53.jar almost works native in Linux
« on: November 10, 2005, 08:28:47 PM »
Ok, fixed the Windoze problem by moving a dll into path and looking at the full dump for the Linux run we have a similar but not the same error:  It's wanting a library called "jspLux86" instead of "jspWin" so I figure I've got another problem as that name is not present in the install files.


Warning: Cannot convert string "-b&h-lucida-medium-r-normal-sans-*-140-*-*-p-*-iso8859-1" to type FontStruct
Installation Directory = /home/diatronics/.cxoffice/Unsupported/drive_c/TRiLOGI/TL5/
Serialio.SerialPortLocal: version 9.1: build 9104
Copyright (c) 1996-2002 Serialio.com, All Rights Reserved.
os.name="Linux"  os.arch="i386"
java.lang.UnsatisfiedLinkError: no jspLux86 in java.library.path: Check that native library jspLux86 is in proper directory
Serial ports reported by system:
/dev/ttyS0;fsdir = /home/diatronics/.cxoffice/Unsupported/drive_c/TRiLOGI/TL5/FileService/
Getting Server IP addresses from O/S
IP Address 1 = blackie/127.0.0.1
java.lang.UnsatisfiedLinkError: SerOpenPort
java.lang.UnsatisfiedLinkError: SerOpenPort
        at Serialio.SerialPortLocal.SerOpenPort(Native Method)
        at Serialio.SerialPortLocal.open(Unknown Source)
        at Serialio.SerialPortLocal.<init>(Unknown Source)
        at b.f.a(Unknown Source)
        at b.f.new(Unknown Source)
        at TLServer21.TLServer.<init>(Unknown Source)
        at TLServer21.TLServer.main(Unknown Source)

47
Technical support / Re:TLServer21.jar may not like J2SE 1.5.0
« on: November 10, 2005, 08:10:36 PM »
Ah-ha!  Just noticed the missing library error.  :>

I found it in the "TRiLOGI/TL5/java lib" directory.  No idea where the lib should be located so I just copied it into good old c:/windoze and away it went.  :)


Evan

48
Technical support / TLServer21.jar may not like J2SE 1.5.0
« on: November 10, 2005, 07:52:29 PM »
Below is a dump of the error message.  There is a real COM1 on this laptop and it works with all the miriad of other PLC packages I use.  The editor, TL53.jar is working fine.  A side note:  The reason why I have not reported this earlier is because this is the first time I personally tried to program our T100MD+.  The usual laptop is not here for me to compare with at the moment.


Installation Directory = C:\APPS\TRiLOGI\TL5/
Serialio.SerialPortLocal: version 9.1: build 9104
Copyright (c) 1996-2002 Serialio.com, All Rights Reserved.
os.name="Windows 98"  os.arch="x86"
java.lang.UnsatisfiedLinkError: no jspWin in java.library.path: Check that nativ
e library jspWin is in proper directory
Serial ports reported by system:
COM1;COM2;COM3;COM4;fsdir = C:\APPS\TRiLOGI\TL5/FileService/
Getting Server IP addresses from O/S
IP Address 1 = acer-tm220/192.168.156.159
java.lang.UnsatisfiedLinkError: SerOpenPort
java.lang.UnsatisfiedLinkError: SerOpenPort
        at Serialio.SerialPortLocal.SerOpenPort(Native Method)
        at Serialio.SerialPortLocal.open(Unknown Source)
        at Serialio.SerialPortLocal.<init>(Unknown Source)
        at b.f.a(Unknown Source)
        at b.f.new(Unknown Source)
        at TLServer21.TLServer.<init>(Unknown Source)
        at TLServer21.TLServer.main(Unknown Source)

49
Technical support / Re:TL53.jar almost works native in Linux
« on: November 10, 2005, 07:35:33 PM »
Actually, I've just tried the server under Win98SE w/J2SE 1.5.0 and it's reporting the exact same comport error.  I guess there is a Java problem ...

Evan

50
Technical support / Re:sizing a dc/dc converter for T100MD1616
« on: November 10, 2005, 01:30:41 PM »
Mark:  You misunderstand me, I didn't post the whole thing again because I no longer have it.  I typed it online and it got erased when I clicked "Post".

51
Technical support / Re:sizing a dc/dc converter for T100MD1616
« on: November 10, 2005, 04:35:31 AM »
I had a big reply but the website decided to erase it.  :(

Short answer is:  Don't undersize your power supply and, as per the other thread, the DC-DC converter also has to handle the large switching spikes.


Evan

52
General Discussions / Re:Full Time Job Opportunity - PLC Controls Engineer
« on: November 10, 2005, 12:20:29 AM »
Wow!  That's one high-pressure looking sales job.  Anyone that fits that description is a walking corporate dream.


Evan

53
Technical support / Re:12VDC machine
« on: November 06, 2005, 02:40:42 AM »
Err, the existing design doesn't appear to have a main capacitor but instead seems to have many large caps for various supply rails which I feel is a little duplicative.  But then I don't actually have the schematics so I might be spouting garbage.


Evan

54
Technical support / TL53.jar almost works native in Linux
« on: November 05, 2005, 08:25:37 PM »
I've recently been experimenting with setting up the office with a Linux based PC and just installed SetupTL5.exe under the MEPIS distro using CrossOver.  After chasing missing icons I found that there were plain .jar files that could be run straight from Linux.  I did so, using the command from TL5.bat, and the editor popped up just fine.  :)

Sadly, apon trying to load an example PLC program it would fail saying it couldn't find the file.  :(  However, I can make a new program and save and reload it in the examples directory so there must be a file format problem (eg: CR/LF vs LF only).

Also, it's getting tons of font errors listed via stdout/stderr which, not surprisingly, causes much lose of info in the rendering of the editor.

I suspect there is not much stopping this.

I tried running the server using "java -DSERIAL_PORT_LIST=/dev/ttyS0  -jar TLServer21.jar" but it was less succesfull complaining that:
java.lang.UnsatisfiedLinkError: SerOpenPort
        at Serialio.SerialPortLocal.SerOpenPort(Native Method)
        at Serialio.SerialPortLocal.open(Unknown Source)
        at Serialio.SerialPortLocal.<init>(Unknown Source)
        at b.f.a(Unknown Source)
        at b.f.new(Unknown Source)

Anyone in the development team been experimenting?


Evan

55
Technical support / Re:12VDC machine
« on: November 04, 2005, 05:27:50 PM »
A little note on reliability:  An average LM78xx regulator can not handle large spikes in either direction.  It is strongly advised you add some real protection between the battery and the PLC.  And you might have to be careful with ground bounce too.

At the minimum, this would be something like a 33 volt zenor or  low voltage schotty diode forward biased in series with an inductor or resistor supplying a large storage cap and another 33 volt zenor clamping the cap.

Maybe these components could be in future PLC board designs.  The existing capacitor could be user replacable for a larger one that will hold up for longer brown-outs.


Evan

56
Technical support / Re:12VDC isolated voltage measurement
« on: November 04, 2005, 01:37:33 AM »
Looks like you are in luck, I found that there is cheaply priced chips that perform isolation via capacitors instead of the more expensive transformer method and whats more they do it completely internal to the chip so you only need deal to power supply and scaling.  :)

http://focus.ti.com/docs/prod/folders/print/iso124.html

You should be able to wire it up using the battery sense wires as the input supply.  This will be quite accurate due to your circuit drawing a constant current and there not being any possible earth loop.  The only concern there is whether it's happy to operate from a single supply rail.


Cheers,
Evan

57
Technical support / Re:12VDC isolated voltage measurement
« on: October 28, 2005, 10:51:36 PM »
To be clear, for non-isolated analogue measurements there needs to be three wires.  Two of them (The pair) are for meassuring the level while the third one (Usually connected between the zero volt power rails at both ends, this gives a good common-mode reference) is heavier and used to absorb any currents that might flow back and forth.

The key is to have no current flow on the sense lines.


Evan

58
Technical support / Re:12VDC isolated voltage measurement
« on: October 28, 2005, 10:42:21 PM »
Full isolation is indeed the most reliable way to go but is way more expensive and also has many more parts that can fail including another power supply.

Instead, you can use differential sensing and rely on full earthing to give a resonable common-mode reference.  With this method you still use twisted-pair for sense wires but need just one op-amp to convert to single-ended and it can do the scaling at the same time.  Best to use an "instrumentation" op-amp, it doesn't cost much more and combines three op-amps into a particular config for better common-mode rejection.  Also, sometimes these only require a single resistor for gain setting.  :)

The limitation of this method is the noise and particularly the signal levels must stay between the power rails.  If it floats away then you will get wrong readings.


Evan

59
Technical support / Re:Stepper jogging
« on: October 27, 2005, 06:29:46 PM »
Not having used it yet but doesn't SETSPEED configure acceleration for you?  By experimenting, you should be able to arrive at a reliable acceleration for your setup.


Evan

Pages: 1 2 3 [4]