Null Modem Question

Hi Doorman,

You may have been here, but it looks like it has a pretty extensive list of pinouts.

Hope it helps!

I did notice there were some null terminal pinouts there as well. Could it be that you got one of those by mistake?

Thanks Mikerho, but the link does not appear. Can you try again?

More info about what is going on. I am installing a device called a CIP module that enrolls credentials into the computer lock managing software. This item is only available in serial, and must be used with a proprietary serial/serial cable. When an end user updates to a new machine without a serial port, I need to use a CablesToGo 26886 USB to serial adapter. This adapter is not quite right, and the null modem crosswires for the correct combination (in theory).

Listed in order, I will have a USB/serial cable adapter, the proprietary cable, a null modem, and then the CIP module. I have tried not using a null modem at all, switching the order of the cable/null modem, and I have tried the cable and null modem both directions. The factory tells me the CablesToGo 26886 is correct, no doubt about it (I have two, and have tried them both). This is a lot of dinkin' around for a doorknob!

Has your combination of PC/laptop(?) and USB/RS232 adapter worked before?

I've had problems with some USB/serial dongle things that are A) OK on PCs but not laptops, B) OK with some devices but not others (specifically PLCs), or C) OK with some PCs but not others.

Usual disclaimers, but the only one I've found to work everywhere I've tried it is the Edgeport, from Digi.

No, this is a new laptop, for a customer/end user.

I have an understanding the USB/serial adapters do vary in their internals. This is why the factory advises the selection of the one captioned above for this setup.

The software has been in place and functioning for four years, and the end user (a business) upgraded his computer (old machine has a serial port, so no troubles). The new machine (USB only) has the software loaded and functioning. The software does make you select a computer port, and you always have to use that same one.

This has happened a number of times: time for a computer upgrade, and the end user has a belief there is no further need for the old equipment so it is traded in, given to their high school kids... whatever happened, it is gone now, and I need to get the new combination working. Previous times this incompatability has not been an issue.

The factory advises it - but have they tried it with a laptop? Just a thought.

Just another thought: if they've tested it and know it works, how come they can't specify the required cable connections?

Is there a PCMCIA slot? Have you thought about a PCMCIA-RS232 converter? They are available for under $30.

Maybe this will help

http://www.camiresearch.com/Data_Com_Basics/RS232_standard.html

What is the pin connections on your one?

I have some old books and printouts of connections and tricks to solve problems -

Way back we called RS232 a "Random Selection of 232 possibilities"

Thanks Hendrik. That is an informative site, lots of good stuff.

From the site: "During this 40-year-long, rapidly evolving period in electronics, manufacturers adopted simplified versions of this interface for applications that were impossible to envision in the 1960s. Today, virtually all contemporary serial interfaces are EIA232-like in their signal voltages, protocols, and connectors, whether or not a modem is involved. Because no single "simplified" standard was agreed upon, however, many slightly different protocols and cables were created that obligingly mate with any EIA232 connector, but are incompatible with each other."

Ugh! Is this a common problem?

That was an interesting read. Is there a way to flip the null-modem around through the use of gender benders or gender changing cables? May be worth a try if you haven't already done so.

re: From the link:

"IMPORTANT: Signal names that imply a direction, such as Transmit Data and Receive Data, are named from the point of view of the DTE device. If the EIA232 standard were strictly followed, these signals would have the same name for the same pin number on the DCE side as well. Unfortunately, this is not done in practice by most engineers, probably because no one can keep straight which side is DTE and which is DCE. As a result, direction-sensitive signal names are changed at the DCE side to reflect their drive direction at DCE."

What I have done in the past was to create a breakout connector where one could use jumpers to connect and swap connections to determine the best / right coupling .

Some times revers strapping signals inputs to the common temporary solved the problem.

small crocodile clip jumpers may also help.

The other method nay be to set the self enable register.

I have a 30Mb (RLL) drive here with all the software solutions but nowhere to plug it in.

The "handshaking" pins might be a problem. Some cables & null modems force the handshaking pins and others pass them. I don't have a 9 pin chart handy but in the 25 pin connector we often forced handshaking signals by shorting 4 & 5 together and also shorted 6 & 8 & 20 together.

The RS-232 lines should be "live" all the time the PC or other device are on. You should be able to identify transmitters with a multimeter. On a modern PC the voltage should be >3VDC or <3VDC on driven lines. The receivers (and open circuit pins) should be 0VDC. This test allows you to make sure you don't have transmitters fighting each other.

Two transmitters connected together will not work but I have never seen it cause any damage.

Some insight might be gained by connecting two computers together and using Hyperterm or another program to talk over the serial lines from one computer to the other. If this works you know for sure the serial ports are "good" and you know for sure that the cables and connectors are "good" (not necessarily correct for the other application but "good").

I would doubt the end device (the interface module) need a handshake; it is a dumb device (not sure that is correct term). It simply reads the value of a Dallas chip that is used as an access control token. It is unpowered so, there is no voltage to look for if it is not connected. About the size of an ice cube relay.

1) USB-to-RS-232 adapters always install on a virtual COM port. The COM port is visible in Device Manager.

Have you checked to see which COM port the adapter is installing on so that you can enter the appropriate COM port when needed?

2) Why is a Null modem adapter required? A "proprietary serial to serial" cable should connect both ends together with no other adaptation required. Should . . .

Did the old system use a "proprietary serial to serial" cable AND this particular null modem adpater?

3) The system worked before. With the new laptop, all you're doing is creating a serial port with a USB adapter and reconnecting the old stuff, right? Was the null modem adapter part of the old stuff?

Who says a null modem adapter is required? Do they define the null modem adapter?

4) Although there is no standard for RS-232 pinouts, PC's have used a common convention for pinouts on the DB-9 for almost 15 years now.

In the control panel go to System Properties. Select the Hardware tab and click on the Device Manager button. Go to Ports(Com & LPT), click the + and you may need to delete the Comx and reinstall or you may possibly be able to just reassign it the necessary number for x in Comx.

Before you even try that did you try, as Iris suggested, using it with out the Null adapter? If so are you sure it is a NULL adapter. A NULL adapter will have Pin 2 on one side connected to pin 3 on the other side. Use a small paper clip to access the contact in the hole on each side to connect your ohm meter.

Good stuff guys. Thanks for your responses.

"The factory advises it - but have they tried it with a laptop? Just a thought." Yes, it is supposed to work on laptops or desktops with no trouble. This is the first such trouble I have ahd after a dozen or so successful upgrades.

"Just another thought: if they've tested it and know it works, how come they can't specify the required cable connections?" The factory tech support has advised I am using the correct parts, and cannot figure out why it won't work.

1) USB-to-RS-232 adapters always install on a virtual COM port. The COM port is visible in Device Manager." "Have you checked to see which COM port the adapter is installing on so that you can enter the appropriate COM port when needed?" Yep; this is correctly identified in the software management. There is, in addition to this part, a cable that connects the computer to a handheld device that you use to tour the facility and program individual doors. The port for both functions must be the same, and the handheld device communicates correctly with the computer, so the port setup is correct.

2) Why is a Null modem adapter required? A "proprietary serial to serial" cable should connect both ends together with no other adaptation required. Should . . . Did the old system use a "proprietary serial to serial" cable AND this particular null modem adpater?" An adapter is required because the specified USB/RS232 adapter ends in a male connector. The serial/serial cable is male on both ends. With the machine previously used, there was a serial port available, and the cable connected there; so, the null modem and the USB/RS232 were not required.

"3)...all you're doing is creating a serial port with a USB adapter and reconnecting the old stuff, right?" Correct. Thus the trouble. It should be working. This should be easy. "Do they define the null modem adapter?" I have no information about the specific null modem, but one was provided by the manufacturer.

"...Serial port sniffer." If the other connection cable and device are working correctly, should I still consider this? I am running with the thought that if one works, then I know the port is identified correctly. Is this right?

"Is there a PCMCIA slot? Have you thought about a PCMCIA-RS232 converter? They are available for under $30." I have been proceeding with the factory recommended USB/RS232 adapter; but, that is something I had not considered and I may pursue it (if I can't get anywhere with the recommended components).

I have verified all components (except the null modem in hand) are working correctly. I have used each of his parts on another system, and they operate as expected. If I could find a pinmap of this connector, I could confirm or eliminate it as a bad device. As I said, I am not certain this info will help to solve my trouble, but this is the only component of the setup that I can single out as the troublemaker. I have ordered some more, but they are not expected to be available until next week and my end user is impressing upon me the urgency of his circumstance.

Thanks to you all, Doorman.

An adapter is required because the specified USB/RS232 adapter ends in a male connector. The serial/serial cable is male on both ends. With the machine previously used, there was a serial port available, and the cable connected there; so, the null modem and the USB/RS232 were not required.

There are devices that called gender changers that look just like a null modem adapter, but don't do null modem pin connections, the device just changes a male end to a female end:

I think you need a gender changer, not a null modem adapter.

I'd be inclined to agree. Too often, the terms are used incorrectly. In a null modem, the send and receive wires are swapped, and in a gender changer they are not.

Echoing what others have said, often a proprietary cable is much like a null modem in function, (or has some other simple wire switching) and sometimes (often) it seems that the proprietary cable is used just to generate revenue.

I might try going to an installation that works, and trace the wires in the cable, null modem, etc.

USB to RS232 adapters are quite variable in terms of whether they work (or not) for a particular usage.

There are, by the way, typically very few wires in an RS 232 cable that actually do anything: if I recall, ground, send, and receive are all that's required for many devices.

Also, Windows computers are far less standardized than many would like you to believe. Not too long ago, I read the bug notes ("known issues") for an HP printer, and there were about 100 entries for specific computers with which the printer would not work correctly. Plug and play... maybe.

I think Iris is on the right path, for the following reasons:

1. Other legacy equipment communicates properly with the USB adapter, so that rules out any configuration/operational issues with the PC out to the DB-9,

2. It's very easy to confuse a null-modem with a gender bender since they both look the same externally, but of course internally they are different.

It's easy to tell the difference though, a gender-bender is wired "straight-through" so that pin 1 on one side is connected to pin 1 on the other side, pin 2 to pin 2, etc., with the important proviso that there is never a connection to any other pin. A null-modem on the other hand is a "cross-over" device that connects DC (Data Communication) signals with DT (Data Terminal) signals so that the TX (Transmit) signal gets connected to the RX (Receiver) port, and vice-versa. There will also be some pins that are connected to each other on the same side of the connector, which is the distinguishing characteristic of a null-modem.

I have an update for us.

The CablesToGo 26886 USB to serial adapter, when first plugged into the laptop, went to the 'net and downloaded a driver (did not come with a disk). It downloaded something it found at Micro$soft, in lieu of it's parent site. Don't know why, some have told me MS love to direct plug-and-play drivers to their library.

I got rid of that, direct loaded the correct driver from the parent site, and all is well. The null modem provided is, after all, okay.

Now, sort of a peripheral question: Why do manufacturers of these things not even bother to tell me to go to their site to get the driver? No mention of this issue is printed on the package. The site at the link has the driver, but why does the device go get the wrong one? Plug-and-play? Is this a common issue? I dinked around with this for three weeks.

[edit] And another question: Why does it not caution the driver disk is not in the package? What if I don't use the machine on the internet? I am starting to see a few companies who dedicate a machine to this sort of task that is, due to security concerns, NEVER connected to the 'net.

I have encountered this frequently- something I thought might be due to the fact that I use Linux, and most manufacturers don't want to supply drivers for Linux for some reason. Interestingly, my usb to serial adapter requires a driver on Windows boxes, but just actually works without any song and dance when I plug it in to my Linux box. I also just purchased a new printer, which came with a CD. When I wanted to install it on a Windows box, it immediately wanted to fire up the Internet to grab the latest drivers (and I never, EVER put my Windows boxes on the Internet!). Funny- the printer started working immediately as soon as I plugged it in to my Linux box. No driver required...although I did have to go to Canon's site to get a driver for the scanner. I have added things like a new key board and a new style mouse to my Linux boxes without doing anything more than plugging them in. On a friend's Windows machine, he wanted to replace the keyboard- not only did he have to log on to the Internet for some unknown reason, before we could even get the machine to recognize the new keyboard, we had to go through some pretty serious security acrobatics...for a new usb keyboard???

Probably making sure the USB device is actually a keyboard and not a USB key spoofing as a keyboard, then scanning for an embedded key-logger...