I am so stupid about this, and so intent on starting the holiday weekend in a foul mood, that I found yet another terminal program to try. It is called “Terminal”. While at first blush this sounds very clear and simple, on reflection it has the potential to confuse, doesn’t it? Anyway, it is by somebody named Bray. And it works, with the three adapters I have sitting here now, or at least seems to with the few things I have tried so far. Advantageously, it has a huge screen bristling with buttons. A list of 7 COM ports grays out the ones that don’t appear to exist, and only COM7 is available. Fourteen standard Baud rates and one custom option are all just pushbuttons, so you can hunt for the correct one quickly. It also has lots of simultaneous display modes (ascii, decimal, hex, bin) and various other nice looking bells and whistles. Beats me what this guy did right that nobody else did, or if it’s just some quirky combination of my PC and the software choices, or what - but I end the year victorious!
By the way, Kevbo, I can be sympathetic. Not only do I believe Microsoft to be an evil shadow government, and think Gates became a philanthropist in a hopeless attempt to wash away his guilt complex, moreover, the way A-B implemented the IEC 1131-3 languages makes me dash frothing into the hallway looking for pencils with which to stab my eyes out.
There is a big difference between many of the adapters but them main difference is the chipset and drivers. Most low-cost USB to RS232 adapter cables are made with a chipseet called Prolific (Taiwan), I have experienced a lot of problems with this. If you invest a bit more $ then you can get a USB to serial adapter designed with a FTDI chipset and drivers. These are compatible with most devices and most versions of Windows, Mac and Linux. Personally I can highly recommed a FTDI based adapter.
HansNielsen, I am not certain I got representatives of the FTDI chipset, but I am certain I got more than just Prolific chips, and I did include expensive adapters. Note that I did also include a PCMCIA adapter that (I think) isn’t a USB serial adapter at all.
The laptop is a Dell, and Dell makes a legacy docking station that our IT fellow gave me to try recently. It includes a serial port. Since this device is attaching to the docking station connector, it might be a “real serial port” and not an adapter at all. The serial port on this device DOES work with Tera Term, Terminal by Bray, and PuTTY, though it does not work with RealTerm. On the downside, this docking station lacks a DVI connector, forcing a digital to analog VGA to digital conversion step for the display, but that might have been OK. Unfortunately, none of its USB ports worked - and when I went back to the regular docking station, now none of those USB ports worked either!
The big problem, though, was that with the legacy docking station installed, if I put a signal to the RS232 port, the CPU usage goes up to 100% on all four processors and stays there. Now it can no longer keep up with my typing, in fact not even close. It records about a character per second as I type, and drops a few. The mouse is practically unusable, and if I connect a serial device at 38400 baud the mouse dances merrily around as if the PC mistook the data acquisition box for a serial mouse.
The latest advance came early this week. I had bought an ethernet-based Lantronics UDS serial server, which provides 2 serial ports, has a configuration page, and sits on the network. It’s a nice, solid, pretty little thing, which supports (among other things) telnet. Though it took a little time to find out about some quirks, I have it working just fine. I am telnetting to it using Tera Term (for some reason RealTerm only partly works, transmitting but not receiving). Note that in this scenario, NOBODY is pretending that there is a COM1 or other serial port anywhere in the system.
I still loosely think that something about this PC will not support emulating a serial port.
The official API for port access for traditional Windows programs is through the file API, with CreateFile( “COM7:” ) and SetCommState() and SetCommTimeouts() to establish the channel, then ReadFile() and WriteFile() to do the actual communication. Presumably, .NET programs have an equivalent set of APIs. There are no facilities to monitor individual lines, AFAIK, so I assume that a terminal program that tries to show this information will end up with some sort of emulated picture of a UART.
At my office we’ve been communicating for years with various devices (cameras, microscopes), using some of the cheapest USB-to-RS232 adapters (usually Prolific) and most adapters have worked well/correctly/predictably.
Until… last year. We have these two Core i5 Dell computers with Windows XP, where the RS232 channel just stops responding after several hours of continuous I/O. We think it has to do with USB reenumeration with this particular USB chipset (“Intel 5 series / 3400 series”). We’ve had to program around it by detecting the suspicious silence, closing the file and reopening it.
Back in the day, the company I used to work for built devices that communicated via Macintosh style serial ports. We spent a lot of time screwing around with various ISA card, PCMCIA, and USB serial ports.
What I found pretty useful was having a digital storage oscilloscope. We’d stick a probe on the serial cable and could instantly see if the data was coming out of the port. It’s useful to have a dedicated connector with exposed lines to make probing simple.
Do you have access to a DSO or a logic analyzer?
Well, yeah if you’re talking DH-485, because XP SP3 broke the PIC driver for the serial port. That wasn’t AB’s fault. Talking DF1 works fine with any FTDI based USB-RS232 adapter.
I recommend a FTDI-based adapter. Their chips seem to actually work (unlike some of their competition’s), and they offer frequent driver updates for a variety of operating systems on their web site. None of this “track down the no-name company that made the whole adapter and ask them for drivers” nonsense.
The ones I’m using now are from USB Gear. While the enclosure isn’t as nice as (for example) the IOGear GUC232A, at least the ones from USB Gear seem to work reliably. [Note - IOGear finally released a driver update in February, after several years of no updates. Maybe their adapter works better now, but I’d already switched.]
I’ve dealt with a lot of different USB to Serial adapters over the years and the best compatibility I found was with Quatech adapters, but they are fairly expensive.
The basic problem is with software that tries to configure the serial port. It just didn’t work with USB to serial adapters and you had to config them manually in the OS. It’s possible that someone else has fixed that problem.
If you laptop has an express card slot, that would be an option also.
Here is a list of the adapters I tried, though incomplete because our ordering system stripped away parts of the description:
“micro innovations” (no other marks)
EasySync ES-U-1001-R10 US232R-10
TrippLite Keyspan USA-19HS
CP TECHNOLOGIES CP-US-03
Cables Unlimited USB-2920
Manhattan
Aten Technologies UC232A
I thought I had a separate record someplace but can’t find it.
Joel UpChurch
>If you laptop has an express card slot, that would be an option also.
What’s that? Is it the same as a PCMCIA or PC Card slot? I did try those, if it is.
>Do you have access to a DSO or a logic analyzer?
For one thing, I can put an AC voltmeter on Tx and ground and see a bit over 4 V when I send a long text file out the port. I also have some of those little adapters with LEDs in them. And an ordinary scope. No DSO or logic analyzers, tho.
Can anybody tell me where to find a list of adapters that use FTDI chipsets?
Express Card was the sucessor to Pc Card, but from I understand Express Card has been replaced by USB 3.0.