panic mode

MrPLC Admin
  • Content count

  • Joined

  • Last visited

Community Reputation

96 Excellent

About panic mode

  • Rank
    Mitsubishi Moderator
  • Birthday 12/15/69

Contact Methods

  • Website URL
  • ICQ 0

Profile Information

  • Gender Male
  • Location Mississauga, ON
  • Country Canada
  • Interests many...

Recent Profile Visitors

10478 profile views
  1. Modbus Read/Write registers

    check this out  
  2. Modbus Read/Write registers

    if the monitor says it is writen correctly, it is writen correctly. in Modbus, inputs, coils, holding registers etc are diffreent memory blocks. your first example is reading input block. your second example is writing to a register but you are still monitoring the input block.  not sure how they are mapped in Omron but check D or H areas for example and check if they are already used, maybe the location is immediately overwriten by some running code  
  3. Warehouse ethernet cable

    since everything is connected to same network, ethernet switches are fine. router is only needed when connecting two networks. Cat5 and Cat6 simply tells how much speed you need. and unless your nodes are some data centers or something else needing crazy bandwith, Cat5 will do. after all this is not Netflix or Google server, just a warehouse which currently does not have any connectivity.  btw copper cables are good for up to 100m / 300ft. if you need longer connection you can add intermediate switch so that no cable segment exceeds maximum. ideally you could also upgrade to fiber optic backbone. then you get best of all worlds (speed, distance, interference immunity, insulation/lightning protection etc.)    
  4. is that how they call Mitsubishi in Taiwan? read manual, there is a function called ReadDeviceBlock... 
  5. Using PTO's to control multiple stepper motors?

    it does not matter what path you choose, result will be the same... if not you have a problem. btw. following manual is never a bad idea (unless there is a typo  ) this has nothing to do with polarity of I/O. this device would still need resistor if driven by NPN outputs since that does not change 24V. if the NPN outputs were open drain (with no internal pullups or RC suppression ) then it would be possible to power positive rail from diffreent voltage such as 5V. But this is getting ugly for sevefal reasons and one of them is reference potential that is not standard.    Btw. in spirit of "no exercercise is too small": most common LEDs (not power devices) can tolerate up to 20mA... but this is the max current that they can survive. for practical use, it is common to use half of that or 10mA. it would be fine to bring it down to 5mA too but most people are conservative and concerned with leakage current and sensitivty to interference, larger current reduces sensitivity. forward voltage drop of LED is related to emitted spectrum. red devices have lower Vf (about 1.6-2.2V) , blue have higher Vf (about 3-3.8V). since optocouplers are sealed devices, they are using infrared and Vf is on the order of about 1-1.2V. and UV LEDs have larger Vf too since on the other side of spectrum.  in industry standard control voltage is 24V. voltage drop across outputr is usually very low, specially at such low current (few mA). then you get for current something like (24V - 1.2V)/Rtotal if external resistore is 2k2, and internal is 270ohm, that would get LED current to about 9.23mA which is just where we want it. manual states 2k... this is not a problem to find today although such value is still less common than older series value 1.8k or 2k2. resistor values come in certain values depending on tolerance or "number of values per decade". for example 2.2 is standard value in every series since E3. on the other hand value. 2.0 value is only available in series E96 or E192. that is fine if such tight value is necessary but this is absolutely not the case here. btw if you don't know what i'm taling about:   and with 2k, Rtotal is 2.27k and If is 10mA.  using 1.8k one would get 11mA. all of these are very close and all three choices are fine (this is not critical).  but actual current will depend on exact vlaues of components and their tolerance but for sake of the argument, lets neglect those. lets' check power as well. power is I^2*R 1/4W is 0.25W and it is good to have some resereves (10-20% at least) using above examples we get Rextern (Ohm) Rcurrent(A) Rpower(W) ReservePower 1800 0.011 0.2178 12.88% 2000 0.01 0.2 20.00% 2200 0.00923 0.18742438 25.03%   yes, there are larger resistors like 1/3W, 1/2W etc. again 1/4W used to be very common. and 2k2 collors bands match so less confusion for person doing actual wiring. also it is easier to fit clear heatshrink tubing over the resistor. btw. tubing is great to reduce chance of accidental shorts when fidlding with circuit (troubleshooting etc.).. and the tubing is not just electrical insulation - it is also thermal insulation so the margins you have there may be gone.   just my $0.02   besides, how else do you whip juniors into shape, get them to read, do math and compete? they have to practice and reinforce what they learned. one of my favourite interview question is just such simple DC circuit but with low voltage incadescant light bulb instead of LED. it is pathetic that less than 5% of candidates are able to solve it for resistance. only maybe 1% actually also considers power and usually still get it wrong. (sigh) ... power is everything... just watch Apolo13 movie for a good example. it is embarassing to watch somone browse DigiKey or Mouser website to select 0.1W SMT resistor which is two orders of magnitude too little. but this is why such question is a great filter to weed out unwanted candidates early on, long before we get to more interesting topics.  likely doing some interview later this week so... hope he/she finds this thread. 
  6. Using PTO's to control multiple stepper motors?

    Fig 3 is correct for PNP. i would use common value resistor 2k2 1/4w resistors to ensure about 10mA through optocoupler LEDs in DM556 inputs.  
  7. Using PTO's to control multiple stepper motors?

    sure, instructions do not care what is attached to the I/O. imprtant part is that wiring is correct (polarity, level) and that signal type and I/O are compatible.
  8. grounding and fluctuation

    you think it is fustrating to see some decimal places fluctiate in the HMI? imagine how these guys felt when interferrence is significant and tied to some output (robot)? in this very educational video robot works perfectly before welder is activated. but once the welder is turned on, robot feedback is compromised and robot goes wild (runaway axes). solution - always consider manufacturers guidelines for the application. make sure that grounding and shielding is fine.  
  9. grounding and fluctuation

    the best is to use small dedicated PSU with low ripple. also check local regulations as this will likely need to be double insulated etc. if not grounded. connecting one PSU terminal to GND with shared (high frequency) loads can have negative effect on noise immunity, specially if analog inputs are high impedance. inverters and drives (welders, VFDs and Servos) usually operate at 8-16kHz and are capable of producing plenty of noise with harmonics going to very high frequencies. when such devices are not installed properly this can be nightmare to troubleshoot. at those clocks, it does not take much to create incuctive or capacitive coupling. but if you do follow manufactures guidelines (separation, shielding, grounding etc.) of those devices, analog are not fussy. i had voltage analog inputs withe very long cables (20-30m) and wired to same cabinet with bunch of drives. don't remove much of shield, bring it close to terminations. use cable that is not just shielded but conductors are twisted pairs.    
  10. it cannot  just arbitrarily connect things to each other. this device does not care about polarity but it still matters what is on the other side and for it to work you need to complete the circuit so there is a current path.  
  11. this supports connections of either polarity so - PLC digital I/O could be any 24V type, either PNP/NPN. PLC analog output will need to be 0-10V of course. it will connect to VR and GND, +10V terminal will not be connected. 
  12. Communication

    speed vs. distance depends on used communication chips. not all are created equal, some have lower output impedance than others and can drive higher current. many fieldbusses are based of 485, and they have different specs regarding cable and terminating resistors. terminations must match the cable. some networks use 82 Ohm, 120 Ohm, etc... or series if 1kOhm and 1nF capacitor. picking some random cable (and Belden makes MANY kinds) and random termination is not a good way to start a project.  i am sure your product manuals have details on that. in general lower baud rate is less sensitive since slower operating time allows signal artefacts (ringing, edge rounding etc). to settle down. this is why some network support wide range of baud rates.  if this is a suspected problem, simply pick the slowest rate supported (1200?) and work your way up. but also check if you have duplicate nodes, ground loops, routing near interference sources etc. there could be number of other reasons this is not working as expected.  
  13. Using PTO's to control multiple stepper motors?

    PTOs are like normal outputs but can operate faster. hardware manual of your product shows detailed connections for both PNP/NPN type I/O.    
  14. Going Obsolete

    well, time to move on or dig deep into pockets.
  15. PLC digital input resistors

    you assume wrong.... circuit you show model of an input card for NPN devices. This is common in low power electronics where risk of injury is non existent (phones, fax machines, etc.) but those are not same class of devices discussed here and don't get same scrutiny. In industrial automation NPN is used as standard in Japan. Most of the world uses PNP. If you are looking for product with NPN configuration in Europe or NorhtAmerica, you will find them but they will be very few since not common. So for most of the world (including AB and Siemens) your circuit would need to be upside down and resistor would be pull-down, not pull up.  And since you show 24V and posted this in industrial forum, it is important to add that your circuit would still be unacceptable even from NPN perspective because you grounded negative side instead of positive. Grounded rail is common or reference potential.  All switching is on the "live" side (24V or phase if AC) due to regulations. When working with NPN type I/O in industrial automation, positive terminal of PSU is grounded and therefore 0V. negative terminal is "-24V" and it is live. and switch, fuse, etc would be on the live side. Some companies are from Japan and insist on NPN whereever they are. For example Honda and Toyota are among them. Honda in North America uses NPN for I/O, but for safety circuits, they use separate power supply and western style safety devices because those are more common. (NPN type OSSD exists but it is just too rare here). More over, do you need just one output or multiple I/O? If you are developing devices with I/Os for industrial automation, why not look at products meant to be used for this?  Yes they cost few $ more but they already contain TONS of features expected from today's products. And i am not talking about hardware specs alone, this also goes to approvals. And you need approvals if you want to marked industrial product, or every failure will be lawsuit with your name on it. so for North America you want at least UL/CSA and it starts not just with components that already have such approvals but also design practices such as PCB layout, spacing etc.