17817Re: Communications error or invalid/corrupt data retrieve
- Nov 17, 2010RS232 can be very reliable as long as you have decent hardware and good software. I have worked with many industrial systems over the years and have had no issues whatsoever, even on systems that support many high speed channels simultaneousness without issues, some with automatic hardware handshaking. All and all it can be very reliable means of communication assuming that there is proper operating system support and properly written device drivers that are critical to their operation.
In the current PC and laptop market many vendors use a cheaper implementation of RS232 (RS232E I believe) that does not support the standard +/-12vdc, but instead use +/- 5vdc. While many devices can continue to work with mismatched voltages on the RX and TX, this is not always the case. Personally, I avoid the RS232E like the plague.
La Crosse did a really poor job with the serial ports on the WS-23XX and WS-36XX. The WS-23XX does not have a normal RS-232 signal ground but instead uses a funky "virtual" ground using the DTR and RTS signals, one set high, the other low. They also run one of these lines DTR or RTS through the consoles ground plane, out to the remote thermo, and all the way up to the wind sensor. This is why these consoles are so susceptible to lightning strikes. Also, one leg of the power pack is connected to the console ground plane. If you have a poorly manufactured power pack this just induces noise.
The WS-23XX cannot chew gum and walk at the same time when it comes to the WWVB reception and the PC serial port. They just flat out stop responding to the PC requests while they are busy.
The WS-36XX does not use normal serial character data but instead uses some funky scheme with the hardware handshake lines to send and receive data. The protocol itself was borrowed from the I^2 (I squared) protocol used for things like memory chips, very strange thinking there. Personally, I think they wanted to save 25 to 50 cents for the cost of a UART that would have done the job properly. This decision was really foolish with respect to what you said about most PCs not shipping with a real serial port. This attempt to do high speed handshaking over USB to serial converters is just a really bad idea.
Even the USB to serial converters are not without fault. MS has done an abysmal job of testing an certifying the adapters and there associated drivers. The fact that these adapters are crashing Windows at the Kernel level is really just inexcusable in my opinion. I cannot tell you how many emails and postings I have made to unsuspecting people who have crappy drivers and want to blame WUHU for causing the Blue Screen of Death (BSOD) on their machines. This happens over and over and over. I have probably spent man weeks of time explaining to users what the real deal is and recommended the IOGear adapter that seems to work better than any of the others on the market.
--- In firstname.lastname@example.org, "Ken" <yahoo@...> wrote:
> I don't really consider the problem reproducible. The example was 09 but I've seen 52 and whatever. I can run several days with no errors and then see a brief one here and there and then get the repeat hourly ones.
> I only responded because he has a similar problem. For myself, it is in the annoyance category until it becomes reproducible. I'd say 80% are happening overnight -- could be temp related.
> Serial ports may rightfully be the first boggy man to visit but it really is too intermittent to default to that to my eye. But trying to nail an intermittent can fool anyone.
> (I've hated serial ports all my life. They can glitch in so many ways. I had troubles with a Soroc terminal attached to an IMSAI and it continued when I ran a BBS. Every bloody one seemed to want something different. I wonder where my breakout box got to...) What's need is for LaCrosse to join the modern USB direct era. Until then...
- << Previous post in topic Next post in topic >>