588,231 active members*
4,203 visitors online*
Register for free
Login
Results 1 to 13 of 13
  1. #1
    Join Date
    Apr 2006
    Posts
    100

    Rs232 Wiring Diagram?

    does anyone here have the rs232 wiring diagram for a Minimill. I will be recieving my 2002 minimill this Friday and would like to have the cable made by then. The cable will be 9 pin on pc side and 25 pin on minimill side I believe. any help would be greatly apprieciated.
    thanks
    Mike

  2. #2
    Join Date
    Sep 2005
    Posts
    767
    Here's a cable that works with ass Haas, Fanuc, Yasnac, Mitsubishi, etc.:

    CNC side (25-pin male) ----- PC side (9-pin female)
    pin 1 ---- cable shield
    pin 2 -------------------------- pin 2
    pin 3 -------------------------- pin 3
    pin 5 -------------------------- pin 7
    pin 7 -------------------------- pin 5
    pins 6 & 8 --------------------- pin 4


    If your cable only has 4 wires, you can also do this:
    CNC side (25-pin male) ----- PC side (9-pin female)
    pin 1 ---- cable shield
    pin 2 -------------------------- pin 2
    pin 3 -------------------------- pin 3
    pin 5 -------------------------- pin 7
    pin 7 -------------------------- pin 5
    pin 6 ---
    pin 8 --- <-- jumper these 3 together on CNC side only
    pin 20 --

  3. #3
    Join Date
    Apr 2006
    Posts
    100
    what type of wire works best. I have cat5 around here somewhere. also Pin1 to cable shield what is the cable shield? Sorry I have almost no experience with rs232.

  4. #4
    Join Date
    Mar 2005
    Posts
    1498
    060409-1110 EST USA

    mikeh78:

    Assuming that HAAS has not changed the wiring, then it is simple, and you need to decide if you want hardware or software handshake.

    At the HAAS 25 pin end you need a male connector. And at the PC 9 pin end a female. At the HAAS end you can ignore pins 6, 8, and 20 because internally HAAS only jumpers these together. If you wanted to use this same cable on Fanuc, then you would need to jumper 6, 8, and 20 together, but not to anything else.

    For software handshake:

    At HAAS jumper 4 and 5 together, but to nothing else. This should not be necessary, but just in case you have problems and talk to someone at HAAS you can say 4 and 5 are jumpered.

    At the PC 9 pin jumper pins 7 and 8. May not be necessary, but some software might require this.

    Between HAAS and the PC connect pin 2 to 2, and pin 3 to 3, and HAAS pin 7 to PC pin 5. These are the data lines with 7 to 5 being common (ground). Pin 2 at HAAS is TxD, and 3 is Rxd.

    Your cable capacitance which is determined by the cable and its length will determine your maximum baud rate up to the HAAS limit of 115.2 kbaud.


    For hardware hand shake:

    Same as software except remove jumper between between HAAS pin 4 and 5, and PC pin 7 and 8.
    Connect HAAS pin 4 to PC pin 8, this is CNC RTS to PC CTS.
    Connect PC pin 7 to HAAS pin 5, this is PC RTS to CNC CTS.


    See my web site www.beta-a2.com for a discussion on isolation, noise, and ground paths. Also a wiring diagram for our E232 to HAAS is shown on the E232 PHOTO page. But be warned this is not the wiring to a PC. For PC wiring you have to interchange pins 2 and 3, and pins 7 and 8 at the PC end, and understand that the PC end connector is female instead of male.

    .

  5. #5
    Join Date
    Apr 2006
    Posts
    100
    what is the difference between software & hardware hand shake?

  6. #6
    Join Date
    Mar 2005
    Posts
    1498
    060409-1151 EST USA

    mikeh78:

    Cat 5 is good low capacitance cable, means longer cable length for a given baud rate. Also the staggered twist pitch reduces cross magnetic coupling between transmit and receive.

    At the HAAS end pin 1 connects to the machine chassis, the common lead of the PC power supply in HAAS, and the common to the serial UART. Pin 7 connects to pin 1 thru a 100 ohm resistor.

    Pin 1 will be labeled shield ground, and pin 7 signal ground.

    At your PC end pin 5 is connected to the PC chassis.

    I will not argue for or against shielding on your CAT 5 cable. It is important that you use one twisted pair for pins 2 and 7 at HAAS and one of the other twisted pairs for pins 3 and 7 at HAAS. The two wires to pin 7 should be connected together at the PC to pin 5. I will suggest orange to HAAS pin 2 and thus orange/white to pin 7, and green to pin 3, and thus green/white to pin 7.

    If you do not use a shield then probably connect HAAS pin 1 to 7. However, to minimize ground loop noise problems or electrical faults our I232 System can be of great value.

    .


    .

  7. #7
    Join Date
    Mar 2004
    Posts
    761
    Cat 5e cable has more twists per foot than standard Cat 5. Cat 5e is a better cable for ethernet. However stranded wire cable would be a better choice if the cable is going to be moved around alot. Solid wire, like in ethernet cable, tends to break from being moved around.

    This is the wiring I used on the Haas Mini mills.


    The cable connector for Haas Mills

    PC DB9 .............Haas Mill DB25
    --------------------------------

    2---------------------2
    3---------------------3
    8---------------------4
    7---------------------5
    5----------|----------7
    ................|----------1

    1-4-6 ................6-8-20

    -------------------------------
    Wayne Hill

  8. #8
    Join Date
    Apr 2006
    Posts
    100
    GAR what is the cost of your system? for the mini mill.

  9. #9
    Join Date
    Mar 2004
    Posts
    761
    Hi Gordon,

    Betatronics®
    Ann Arbor, MIchigan

    Hi neighbor - I live in Jackson. near I-94 and 127.
    Wayne Hill

  10. #10
    Join Date
    Sep 2005
    Posts
    767
    I do not recommend that you use CAT5 or CAT5e cable. These are UTP, or "Unshielded Twisted Pair" cables, and RS232 requires a shielded cable. The reason for this is simple:

    RS232 signals are transmitted on one wire, and the voltage of that wire relative to "ground" is read by the receiving device. If there is noise on either the ground wire or the signal wire, you will get corrupted data.

    Ethernet or RS422 signals are different. They use "differential" signals, where there are TWO wires for each signal. The only valid signal happens when one wire goes "high" and the other wire goes "low". A noise spike on the cable will place an identical spike on both wires, which is ignored by the receiving device. These type of signals are best transmited by twisted pairs, and the shield is not really needed.

    The best wire we've found for RS232 is a low-capacitance double-shielded cable. You can buy 7 wire cable from Black Box Corporation at www.blackbox.com Look for the cable EMN-07A. You can get 500, 1000, or 2000 foot spools, or you can get custom lengths. The cable averages about $0.39 per foot.

    A good generic wire would be 22 or 24 gauge stranded wire, with an overall foil or braided shield. Low-cap wire is good up to 500 feet, and generic wire is good up to about 100 feet.

  11. #11
    Join Date
    Mar 2005
    Posts
    1498
    060409-1216 EST USA

    mikeh78:

    Handshaking is a means to modulate the data flow from a sender to receiver to prevent buffer overflow.

    It is possible to set handshake to none, but that is generally risky.

    Hardware handshake a physical wire (signal path) separate from the data lines to signal when it is ok or not ok for the sender to send data. Requires 5 wires. The send or stop state is determined by the voltage level on the handshake line. So this would be defined as level sensitive. The hardware handshake could have been done as an edge triggered signal, but is not.

    Software handshake uses the same serial communication lines as the data uses with two special control codes to signal stop (XOFF) and start (XON). These can best be described as edge triggering signals. This requires 3 wires.

    Our I232 System provides electrical isolation to approximately 2000 volts at each end of the system, maximum baud rate of 115.2 kbaud, and at 4000 ft of interconnect cable. The electrical isolation eliminates ground loop noise common to HAAS machines with brushless servos. In the simplest form the I232 must operate in a software handshake mode. Higher cost to do hardware handshake. Also the I232 can be provided without bandwidth limiting for higher baud rates.

    The I232 System is typically $200 plus interconnect cable, and shipping.

    .

  12. #12
    Join Date
    Apr 2006
    Posts
    100
    Gar I am interested in your product. Could you send me a quote on the I232 system? I need to go about 30 feet max from pc to machine. Shipping zip code is 55329.

  13. #13
    Join Date
    Mar 2005
    Posts
    1498
    060409-1243 EST USA

    mikeh78 and Dan and Wayne:

    There are basically three ways to induce noise into an RS232 interconnect cable. These are capacitive, inductive, and ground path resistance. In this case I am ignoring RF radiation.

    Capacitive coupling is minimized by electrostatic shielding, such as Beldfoil or braided shield. Beldfoil provides the more compete shield. Electrostatic shielding has virtually no effect on magnetic coupling at low frequencies (those we are concerned with), at microwave frequencies it is a different story. The effect of capacitive coupling can be reduced by adding shunt capacitance to your signal line at the destination, but this reduces your maximum baud rate. Shielding is better and adds less capacitance for a given effect.

    Inductive coupling is minimized by twisted pair wiring, and/or magnetic shielding. If I put a short across a 12 V car battery I will generate a large magnetic field, but negligible electric field. Twisted pair has negligible effect on capactive coupling.

    Noise from ground resistance (more correctly impedance) and a current thru it is not affected by any of twisted pairs, electrostatic shielding, or magnetic shielding.

    What you need to do for shielding, or isolation is greatly dependent upon your expected source of noise. For very short runs ground impedance may be your most likely source of noise. Also for any length wire run the ground impedance is important. For runs by welding operations both magnetic and capacitive coupling will be important. For runs by large motors and transformers it is mostly inductive coupling.

    Note CAT 5 cable can be obtained with electrostatic shielding. Cable such as Dan is suggesting has larger conductors and thus is more rugged. This is better than CAT 5, but if you already have CAT 5 and do not run close to capacitively coupled noise sources and the run is short you could start with it. If you are close, 5 or 10 ft, you could probably use untwisted telephone wire. Tie the two center conductors together for common. You could also use multiple wires of Belden 8523 (single conductor), and not bother to create individual twisted pairs for a short distance. I have done this many times, even as much as 20 to 30 ft.

    To minimize ground noise when you start out I suggest that you get your 120 V for the computer from the HAAS machine. This means from HAAS's 120 V secondary if your load is small, like a lap top, or from a separate control transformer supplied from the power lines to HAAS and then use the electrical safety ground of the HAAS.

    Beldin 8723 is a good shielded cable with two separately shielded twisted pairs, but has about 5 times the capacitance per foot as CAT 5. Five ft is a realistic maximum for 8723 at 115.2 kbaud. Maximum baud rate will be approximately inversely proportional to length between some short distance like 10 ft and a long distance like 1000 ft.

    Wayne:

    I have been to Jackson many times. In the 60's I consulted with Mechanical Products when Lyle Trolz was president and his father-in-law owned the company. In the mid 60's the company was sold to Howard Aiken lndustries.

    Howard Aiken was a very interesting person. He was one of the first if not the first, in modren times, to say that the way to build computers was digital. This was like 1935 or thereabouts and generally everyone said it could not be done. He was the originator of the Harvard architecture, and the Harvard Mark I computer. This was really a general purpose programmable computer, but of an electro mechanical design. Programming was from a punched paper tape.

    It should be noted that Aiken was aware of Babbage, but not details, because Babbage's notes were not found until 1937.

    It was from the approximate 1935 time of his concept until 1939 before he could get any support for his ideas. Then it took until Feb 1944 to get the machine operational, partly delayed because of WWII. But then was put into use before war's end by the Navy.

    Alro in Jackson is a major metal source for my son's business.

    .

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •