JBTR
Aluminum
- Joined
- Mar 19, 2020
- Location
- USA, Northern Illinois
Wow, if pin 6 is ground, then I don't know how it's wired. Non-standard in that way could make things difficult. Do you have any documentation?
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Largest Manufacturing Technology Community on the Web
Stay Connected:
RS-232 Trouble! Ancient Dyna Mechtronics Mill.
- Thread starter AndrewZ
- Start date
- Replies 46
- Views 2,686
Unfortunately, no documentation besides the manual which seems incorrect. I confirmed again, and it does seem that pin 6 is ground.
I read on wikipedia that the control and data voltages should be opposite. Not sure if true, but perhaps one way to proceed. At least it seems to be possible to say that pin 7, 8 or 4 is pin 3 of the standard setup. Since pin 1 does not seem even connected it's not that many combinations to test.
Last edited:
JBTR
Aluminum
- Joined
- Mar 19, 2020
- Location
- USA, Northern Illinois
As far as I know, RS232 lines should have enough impedance and protection that nothing should be damaged by making a wrong connection in your testing.
Yes, I looked at the standard documentation which is proprietary and somewhat difficult to get hands on and it says so. Then again, the line does not seem standard.
Mostly the difficulty is in knowing how long to wait to confirm the pins are not correct. The setup is now PC on receive mode and the controller on send mode and I change the pinout hoping something would transmit.
Just a thought, with software handshake would it be enough to get the transmit on the controller and receive on the PC correct along with the ground for transmission?
Also, bridging 4-6, 7-8 on the PC end to make sure it is not stuck on a waiting state. This may also be an issue on the controller.
JBTR
Aluminum
- Joined
- Mar 19, 2020
- Location
- USA, Northern Illinois
The actual software handshake (aka flow control) needs data flowing in both directions. But if your bit rate is slow enough, it's possible that flow control would never be needed, that is the receiver can process the data as it comes in and its buffer never fills. Beyond that, whatever higher layer communication protocol is being used may require communication in both directions; that would be common. So if you had success, you would know it's working, but if nothing happens, you still wouldn't know the cause.
If both sides are set to software flow control, I don't think it would matter, but I don't think it will hurt either, as something may not be configured as expected.
angelw
Diamond
- Joined
- Sep 10, 2010
- Location
- Victoria Australia
Pin 5 of a DB9 connector is Signal Ground. Protective Ground is the Frame.
@AndrewZ
The Control is a PC Based platform; I would be surprised if the Cable Pinout didn't comply with conventional configuration.
For Xon Xoff Handshaking the following would apply.
Machine Side --------------------------------------- External Equipment Side
DB9 Female ---------------------------------------------- DB9 Female
DB9 Frame --------------- Shield Trace --------------- Not Connected
2 ------------------------------------------------------------------- 3
3 ------------------------------------------------------------------- 2
7
| Bridged
8
1
|
4 All Bridged
|
6
5 ------------------------------------------------------------------- 5
It won't hurt to complete the Bridging of 7&8 and 1&4&6 at the External Equipment Side, but normally its not required.
Voltage levels between +5 to +15 and -5 to -15 are the usual norm. The + voltage will yield logic "0" and - voltage logic "1"
Regards,
Bill
Ok, I checked every pin configuration with ground as pin 6. At least 5 second wait between each. Then changed the COM port to "2" on the controller configuration just in case the port is number 2 (I only see one port on the machine) and did so again.
Unfortunately, nothing, not even an error messages. I wonder if the handshake is required to be hardware or if the boud rate is required to be a certain number, or whether the port is functional at all. One problem could also be that the pins 4 and 6 (whatever they are in the configuration) are not bridged.
Last edited:
Yes, this pinout is on the manual as well for software handshake.
However, with volt meter pin 6 turns out to be the ground measured from a screw. Rest have at least kilo ohms of resistance. In this case it's also a male - male connection.
Last edited:
angelw
Diamond
- Joined
- Sep 10, 2010
- Location
- Victoria Australia
What do you measure between pin 5 and Protective Ground of the Control DB9 connector and what do you measure between pin 6 and Protective ground of your external PC?
Regards,
Bill
Last edited:
On the USB-RS232 between pin 5 and the USB end of the cable around 0.5 ohms. For pin 6 some mega ohms.
On pin 5 of the controller I believe it was some kilo ohms, pin 6 around 0.5 ohms. Many pins were on the some kilo ohms range, some on the mega ohms range.
Last edited:
I am thinking about opening the case and investigating if it is a regular PC motherboard.
I searched how to install RS-232 on a motherboard since that may be a next step, and below is how a connector port looks on one board.
Now imagine connecting wrong way around, the ground becomes 6!
Pin 2 becomes 9 and pin 3 becomes 8. Also explains why pin 1 seems as if it is not connected, being on the empty spot.
I searched how to install RS-232 on a motherboard since that may be a next step, and below is how a connector port looks on one board.
Now imagine connecting wrong way around, the ground becomes 6!
Pin 2 becomes 9 and pin 3 becomes 8. Also explains why pin 1 seems as if it is not connected, being on the empty spot.
Last edited:
CarbideBob
Diamond
- Joined
- Jan 14, 2007
- Location
- Flushing/Flint, Michigan
Why do I think that the machine and the computer do not agree on what ground is?
This is a ground loop and then which side has more power?
The computer plugs in the wall in most cases. The machine attaches to other.
One should always "ring out" the cable as a first step.
This is a ground loop and then which side has more power?
The computer plugs in the wall in most cases. The machine attaches to other.
One should always "ring out" the cable as a first step.
To be clear, when measuring there was no connection at all between devices. Instead the ground pin was separately identified on each device.
Above your post it is explained that the likely explanation for the incorrect ground pin seems that the RS-232 connector is connected wrong way around on the mother board of the controller.
I will try make it function one more time with this in mind and then open the case and try to correct the orientation if all fails.
Last edited:
Go buy a rs-232 jump box on ebay or borrow one. I did and it saved me hours of struggling with which wire goes where.Voltage changes + or - .
Also some of these old computers only talked in EIA or ASCI don't know which one is pretty much standard these days I think ASCI is the standard one.
What year is this bone?
It's probably ASCII as that's what the RS232 standard is IIRC, but only 7 bits so less special symbols.
I believe the machine is from the 90s or maybe early 2000s.
Success!
The pin numbers can be obtained by assigning the pins of the controller the number 11 - standard pin number. In other words, pin 2 is 9, 3 is 8, 4 is 7 and so on, pin 1 is not available. Clearly the RS-232 connector is connected incorrectly to the motherboard.
The reason there was no function when going through the combinations is that the software handshake also does not work. Only the hardware handshake works and must be turned on.
I also bridged the pins 4-6 on both the controller (apply the conversion) and PC. It may not be necessary. COM1 is selected as the RS-232 port on the controller.
Thank you very much to all, especially JBTR for help!
Now if someone knows a way to do a bulk dump and drip feed please post! Bulk dump may not be possible, but drip feeding is according to the manual, it won't say how.
The pin numbers can be obtained by assigning the pins of the controller the number 11 - standard pin number. In other words, pin 2 is 9, 3 is 8, 4 is 7 and so on, pin 1 is not available. Clearly the RS-232 connector is connected incorrectly to the motherboard.
The reason there was no function when going through the combinations is that the software handshake also does not work. Only the hardware handshake works and must be turned on.
I also bridged the pins 4-6 on both the controller (apply the conversion) and PC. It may not be necessary. COM1 is selected as the RS-232 port on the controller.
Thank you very much to all, especially JBTR for help!
Now if someone knows a way to do a bulk dump and drip feed please post! Bulk dump may not be possible, but drip feeding is according to the manual, it won't say how.
angelw
Diamond
- Joined
- Sep 10, 2010
- Location
- Victoria Australia
Hello Andrew,
Your machine seems unusual. I had the chance, with one of my clients that has a machine with the same control as yours, to test the pin out of the DB9 connector and it follows convention. that is, pin five is signal ground, the frame is protective ground, etc.
When you say that you had to turn the Hardware Handshaking on, the cable you're describing, with pins 4 and 6 bridged at both ends (index corrected at the control end) doesn't even comply with a Partial Hardware Handshake cable. So what is the complete pin out of the cable you used?
Regards,
Bill
Last edited:
Clearly the cable was accidentally connected wrong to the Pentium board in the manufacturing process. It was of course supposed to be connected in correct orientation, and surely other machines of the line should have the correct orientation.
The solution should work for any controller or computer where the connection is made in the reverse orientation by accident. See post #31 if you missed it.
So the complete pinout is the standard one once the mistake is corrected for
2->3
3->2
7->8
8->7,
where 7, 8 are for handshakes. and indeed I bridged 4-6 on both.
DavidScott
Diamond
- Joined
- Jul 11, 2012
- Location
- Washington
I had a 4400 from 1988 and think you had to use their communication software for drip feeding. My control could only hold 999 lines of code so the drip feed would do a hiccup every time it crossed that threshold. You don't say what year your control is so I have no idea if this helps.
I would have to check the year. It's one of the newer Pentiums though.
Last edited:
