What is DeviceNet?

Поділитися
Вставка
  • Опубліковано 27 лис 2024

КОМЕНТАРІ • 97

  • @patrickmaloney5772
    @patrickmaloney5772 5 років тому +119

    Original DeviceNet dev team member here (spec coauthor, device profile concept creator). It's heartening to see how DNet has been adopted around the globe and how it's still being nurtured by real professionals like the RealPars team. Great video.

    • @realpars
      @realpars  5 років тому +12

      Hi Patrick, thank you for taking the time and watching this video. We are so delighted to get such positive feedback from an original DeviceNet dev team member.

    • @travisbuckle1071
      @travisbuckle1071 4 роки тому +3

      Thanks for the awesome work!

    • @jrdz009
      @jrdz009 3 роки тому +11

      How the hell you created something so hard to work with? It's a real headache!

    • @ffxv9m
      @ffxv9m 6 місяців тому

      @@jrdz009 Chill bro...those days ere very limited..nowdays no one will prefer device-net anymore

  • @MrWaalkman
    @MrWaalkman 5 років тому +20

    As a battle-scarred veteran of DeviceNet on the Rockwell platform, let me give everyone a free piece of advice. Avoid it. Avoid it at all costs.
    Here is a list of problems that you may (and probably will) experience at some point:
    1) Cabling issues. Usually fixed by unplugging the bad joint and plugging it back in. The downside however, other than the obvious interruption of the equipment and processes using it is: a) Which connector is the "bad" one? b) Does it happen to be that connection 30 feet in the air located over the oven? (or in some other inaccessible location?).
    Some cables are better than others (or more correctly, some cables are worse than others), the DeviceNet thick cables usually do okay, Control Boss on the other hand is terrible (bad connections up to and including one that melted down on us).
    So how do you find a bad connection? One approach is to disconnect and reconnect every last joint in the system. Might fix it, might not. My approach was to disconnect power and disconnect the resistor at one end and place my ohmmeter there. Now you should see 121 - 124 ohms of resistance. Anything more than that and you are going to have problems. To find the bad joint, have an assistant break the network in the middle and place a terminating resistor there. If you resistance came down to the 121 - 124 level, you know that the bad joint is further down the network, otherwise work your way up closer to your meter.
    2) Voltage drop for your power. After several hundred feet your 24 volts ain't gonna be 24 volts anymore. More along the lines of "too low to operate properly any longer". This can be fixed by adding an inline power supply though.
    3) Signal degradation. This one is a biggie, hook up a scope to a troublesome DeviceNet network and you will see what I am talking about. The signals are all over the place, how it ever works is a mystery to me. And don't expect to see the pretty textbook waveforms, if that was what was going on in your network you wouldn't have needed to hook up a scope in the first place. One of the weird things you will see is nodes that offset their signal up by a volt or so effectively making the CANH higher, and the CANL higher as well (which isn't a good thing).
    We mitigated the noise problem somewhat by building a cheap filter network that we stuffed into a field termination connector. It consists of two 60.4 ohm resistors (yes, you can buy a 60.4 ohm resistor: www.digikey.com/product-detail/en/yageo/MFR-25FRF52-60R4/60.4XTR-ND/14774) in series with a 22 nanofarad cap connected to where the two resistors combine, with the other lead of the cap tied to ground. You get about 3db in noise reduction. Further details are on the web, that's where we got it from.
    I suspect that Rockwell might of made a poor decision in selecting their DeviceNet transmitter chips. They seem to have their slew rate set to around the 1 Mbs range which is faster than necessary (and can lead to ringing in the waveform).
    Okay, so that's some of the problems you might experience with cabling, on to hardware.
    1) DeviceNet is going away. Get used to shopping for parts on eBay if you plan on using it.
    2) Firmware revisions, and why a drop-in replacement usually won't work. So your spiffy input block bit the dust and you grab a new one from stores and slap it in. Good to go, right? Wrong. If you have electronic keying set to its most restrictive level (or even the intermediate setting if the firmware rev level difference is large enough), your scanner isn't going to let it in on the network. So break out RSNetworx for DeviceNet and get to fixing it. Be sure to not mess up your scan list when you do it though. Eventually you will, and then you will see what I am talking about.
    3) Same as item #2, but now you need an updated EDS file. Go have a look at the EDS file repository on the Rockwell site and see if that doesn't crush your soul. It does mine. So many thousands to choose from and usually none of them work.
    4) Same as #3 except now you need an EDS file from the vendor for an outdated piece of hardware. Good luck with that.
    And I would be remiss if I didn't save a special place in my -rant- heart for RSNetworx for DeviceNet. I tried to get the guys from Rockwell to build it into RSLogix. That would have solved many of its problems if the user could add the block and define its bitmap in one place. But no...
    1) So RSLogix for DeviceNet has had some bugs over the years, and if you work for a company that locks down the programming workstations (say for example, General Motors) you may find that you are not keeping up with the times. Therefore you probably won't be able to program that spiffy new block (one version of Networx omitted the "Okay" button for accepting and loading your new configuration to the node. However, hitting the "Enter" key still worked). Is this a DeviceNet problem? Not strictly, but does that matter when you get home several hours late because of it? It tends to blur the lines somewhat.
    2) Goofy mapping. You can overlap your map on top of itself. We've done it. And remarkably, you can get replacement blocks that require you to remove part of your mapping to get them to work (I can't remember what block that this was on, but we had just about every engineer in the place out looking at it that day).
    What's humorous (or makes this topical) is that I was just out to lunch with an old partner of mine that just hired on with our company last Thursday. On the way back from lunch he gets a call from a tech at Nissan asking for help on a VIN etcher that was connected to a DeviceNet network. Specifically, it was the *only* device connected to that network (besides the scanner), and it still had to be reset constantly.

    • @maintenanceguy2479
      @maintenanceguy2479 4 роки тому +1

      Thanks for the info i am so desperate at my devicenet network always having busoff deteced

    • @MrWaalkman
      @MrWaalkman 4 роки тому +4

      @@brianbeebee1731 In the early 2000's when GM first began to adopt DeviceNet I could see what a hot mess that we were getting ourselves into. I wrote an email to our section leader in the (un)Common Controls Group and gave him my opinion that DeviceNet wasn't ready for prime-time (IMHO it still isn't), and it would be irresponsible to use it in our factories.
      He got his nose out of joint when I told him that not only was DeviceNet ill-suited for a factory floor, but their "Common Controls", while working nicely for equipment in the Body Shop, didn't work so well anywhere else. I ended the email telling him to think "PLC" not "PLC++".
      At that, he demanded an apology, or he would have my job. He got neither. I retired in 2015, and have come back upon occasion for project work. On the other hand, there's not too much interest at GM these days in expanding DeviceNet networks. Nor did he last all that long as a section leader...
      In any case, my advice fell on deaf ears, and GM is still paying for it.
      So why all the vitriol over DeviceNet? Well, let's take some other I/O networks into consideration. Just the ones that I have had personal experience with:
      Data Highway+ Remote I/O: Pretty old, but nearly bulletproof. The last problem that I had with it (around 2015) was when the boss, who wasn't exactly solid in electronics, said: "We don't need terminating resistors". In this case he did.
      GE Genius I/O: Old as well, but even more reliable than DH+. The last time that we had a problem with it was when an electrician inadvertently left the terminating resistor off when he demoed out one of the presses. The other time was when the production dumped their excess plastic from the injection molder on top of the cable. Sorta melted it.
      As a side note on Genius I/O, it had a nasty habit of popping you with high voltage from time to time. Not often, not dangerous, but always annoying.
      Modbus: What can you say? When the heat death of the universe eventually occurs, all that there will be left will be cockroaches and Modbus. The bad thing about Modbus is that companies get pretty loose with the specs and manage to make what is probably the most open of all protocols out there somehow manage not to work.
      EtherNet: Once you get BootP to work on your PC (have fun with that), and when the Field I/O modules finally settle down and decides to accept (and retain) your IP address, it's just fine.
      For perspective, I could take all of the problems that I have had with these networks from the 30+ years that I've been an engineer and they would total up to about the same amount of problems that we had every six months or so with DeviceNet at GM.

    • @MrWaalkman
      @MrWaalkman 4 роки тому +1

      @@maintenanceguy2479 Just saw this. Did you get it fixed?

    • @MrWaalkman
      @MrWaalkman 4 роки тому

      @@brianbeebee1731 The worst thing about Genius I/O was that horrific cable that GE specked for it. Remember that junk? Early on, there was a problem with Genius blocks going to sleep (or dying) on you. Pretty rare though. And for very old Series Six systems your Genius Bus Controller could fail on you (most likely only needing a re-cap).
      The thing with ControlNet, was AB using an Ethernet connector rather than designing their own. This led to more than one device getting blown up.
      GM had a strange policy concerning Ethernet I/O. There was *one* guy up north who banned it, saying: "As long as I work for here, there will be no Ethernet I/O at GM". That's pretty much a direct quote BTW. So GM got on the Ethernet I/O bandwagon rather late compared to other companies.
      The funny thing was that we had Ethernet I/O in our paint mix room long before he was ever in the position to say such a thing. Never had a problem with it. :)

    • @maintenanceguy2479
      @maintenanceguy2479 4 роки тому

      @@MrWaalkman yes sir thanks for the info you shared it really help me a lot to overcome our devicenet problem thank you so much sir.

  • @serhiikozak1704
    @serhiikozak1704 3 роки тому +5

    Terminating resistors are to suppress signal reflection. Proper shielding is supposed to reduce electrical noise specifically in voltage signal (potential)

  • @Ryarios
    @Ryarios 6 років тому +8

    A couple of things I want to mention here. The maximum number of node address may be 64 but you will rarely if ever be able to achieve that number. Cable length restrictions and data throughput will generally limit you considerably. I generally shoot for about 24 node addresses per highway. However, if you're required to have a highway that is intrinsically safe, the number of nodes and main trunk maximum length is greatly reduced to 10 nodes and 100 ft.
    Also the terminating resistors are designed to make the data highway "look" like it's infinite in length as far as the electrical signal is concerned. That is to say an electrical signal propagating down the highway never returns. Unterminated highways can allow electrical signals to reflect back onto the highway creating ghosts of messages to appear when the electrical signal hits the end of the wire. This can cause a lot of problems and make it hard or impossible to get valid messages through. This can also happen to a much smaller degree if your trunk and drop lines have orphan strands where they terminate at terminals or connectors. We generally buy pre-made cables to help alleviate this problem. You can get custom length cables pre-made for you if you have the time and know the length.

    • @realpars
      @realpars  5 років тому +1

      Hey Ryarios, thank you for the explanations.
      Yes, it is absolutely correct that you may not be capable of using the maximum cable length and nodes in DeviceNet and depending on the circumstances it may differ from one network to another. Using the terminating resistors is inevitable as it has revealed in this video.

  • @jaikishank
    @jaikishank 3 роки тому

    Deeply appreciate the clear conceptualisation.

  • @Zeqqqq
    @Zeqqqq 5 років тому +1

    What a fantastic video! All needed areas covered and well treated. Thanks!

    • @realpars
      @realpars  5 років тому

      Thanks for your support!

  • @avidLii
    @avidLii 5 років тому +1

    As usual, very insightful. Thanks Real pars!

    • @realpars
      @realpars  5 років тому

      Thanks for your support David! Happy to hear that you are benefiting from our courses. Have you also followed our free PLC Hardware course? realpars.vhx.tv/browse
      Happy learning!

  • @oktaytekin4560
    @oktaytekin4560 6 років тому +3

    Thank you very much.This lesson is very perfectly as the other lessons.(ın İstanbul)

    • @realpars
      @realpars  6 років тому

      Thanks for your positive feedback! Happy to hear that.

  • @juanlorenzo5501
    @juanlorenzo5501 3 роки тому +2

    Excellent my friend

    • @realpars
      @realpars  3 роки тому +1

      Thank you, Juan!

  • @jhonez801
    @jhonez801 4 роки тому

    I like how you explain it so easily

    • @realpars
      @realpars  4 роки тому

      Happy to hear that!

    • @leocdms
      @leocdms 3 роки тому

      Expertise is not just know how but make it SEAM easy 😏.
      Great job as always

  • @edsoncastro2988
    @edsoncastro2988 2 роки тому

    Muchas Gracias, por el video.

  • @hemantjoshi9771
    @hemantjoshi9771 4 роки тому

    the explanation is too good

  • @drummer1315
    @drummer1315 5 років тому

    Excellent video, I will share the link on my social networks.

  • @matsietsimohulatsi775
    @matsietsimohulatsi775 4 роки тому +1

    THANK YOU,very informative

    • @realpars
      @realpars  4 роки тому

      You are very welcome, happy learning!

  • @faizalraazy91
    @faizalraazy91 3 роки тому

    Great presentation

    • @realpars
      @realpars  3 роки тому

      Thank you, Faizal!

  • @erickontiveroslara209
    @erickontiveroslara209 5 років тому

    Thank you RP so much for the good quallity material. The far right device in 7:30 looks pretty much like an IO Link master. Now that I think about it DeviceNet, IO Link and ASi-Bus work pretty much the same way: that is in the field level, are open sandards, they use an independent or special power supply (at least IO Link and ASi), theres a flat cable abailable for installation of perforating conexions (DeviceNet, ASi), and other very similar accesories in common can be used. Not surprisingly there must be also many differences between them

    • @realpars
      @realpars  5 років тому

      Thanks for sharing your knowledge with us, Erick! We appreciate that.

    • @Letsgo_BrandonR
      @Letsgo_BrandonR 2 роки тому

      I o link is a scam, the aoi blows up your memory. People should just use devicenet

  • @baqarengr4823
    @baqarengr4823 5 років тому

    Sir Thanks for your Efforts 👍

    • @realpars
      @realpars  5 років тому

      You are very welcome!

  • @racuwave
    @racuwave 5 років тому +2

    the courses in the web are only for siemens?? do you teach allen bradley ???

    • @realpars
      @realpars  5 років тому

      Hi Rucawave,
      Thanks for your comment!
      We mostly focus our course videos on Siemens, we do have a couple of course videos on Allen Bradley.
      We are planning to add more courses to our Allen Bradley section in our course library, but we do not have any of those courses scheduled yet.
      Feel free to have a browse around through our course library to see which topics we cover at the moment. bit.ly/30ZrxWq
      Let me know if you have any other questions.
      Happy learning!

  • @qzorn4440
    @qzorn4440 4 роки тому

    yes installation and setup, if done wrong have fun chasing the faults. :/

    • @shearzy3152
      @shearzy3152 3 роки тому

      Yep, can confirm. DeviceNet is a nightmare vs EIP these days

    • @Metallizombie
      @Metallizombie 3 роки тому

      Yeahhh I’ve had my share of troubleshooting issues on our miles of conveyor system 😓

  • @pablomoreno6909
    @pablomoreno6909 5 років тому

    Your videos are perfect!

    • @realpars
      @realpars  5 років тому

      Thanks a lot! :)

  • @WeconTechnology
    @WeconTechnology 2 роки тому

    devicenet could be used in Servo?

    • @realpars
      @realpars  2 роки тому

      Yes, DeviceNet could be used, but the servo controller hardware would need to have a DeviceNet interface available.

  • @nguyenngocsang7748
    @nguyenngocsang7748 Рік тому

    thanks you so much

  • @paraspatidar7925
    @paraspatidar7925 3 роки тому

    Should have more like to this video

  • @renegadeflower575
    @renegadeflower575 6 років тому

    As always, you get top notch controls knowledge from realpars more than anywhere else. Sign up to www.realpars.com and I promise you won't regret it.

    • @realpars
      @realpars  6 років тому

      Thanks a lot for your support! Happy learning.

  • @shoortey1338
    @shoortey1338 2 роки тому

    Got a complicated problem with one of our machines using DeviceNet. Im kinda new to this but what happens is that we have several nodes in the machine including a staübli 6axis robot. For some reason when the (deadman's switch) or servo motors recieve power it knocks all of the nodes off like they are loosing connection. If we exclude the robot in the equation the nodes doesnt fault. Even with an external power supply to the robots modules this happens. Does anyone have any idea what to look for here? Have tested all the other nodes as well but works aslong as the robot doesnt recieve power.

  • @daffa_gamink
    @daffa_gamink 4 роки тому

    thank u

  • @JBCFAUTOMATION
    @JBCFAUTOMATION 6 років тому +1

    Can you create video about ethercat?

    • @realpars
      @realpars  6 років тому

      Hi there, thanks for the suggestion. I will pass your topic request on to our creator team.

  • @sandychoudhari7106
    @sandychoudhari7106 5 років тому

    Sir thank you for this valuable information
    Please suggest best software for P & ID designing

    • @realpars
      @realpars  5 років тому +1

      For P&ID's, the top two major competitors are AutoDesk (AutoCAD) and Bentley (Microstation). There are also add-on software packages that make creation and management of P&ID drawings easier, such as the AutoCAD P&ID package for AutoCAD. Before buying any CAD package, request a demonstration from your local distributor so you can see what is required in order to use their software (hardware, training, etc.) and so that you fully understand the costs of the software licenses, including any annual maintenance fees. Good luck in your search!

  • @optisliprg8571
    @optisliprg8571 5 років тому

    Great video. When a video about ASi protocol for?.Thanks

    • @realpars
      @realpars  5 років тому

      Thank you! I will forwarded this topic suggestion to our creator team. Thanks for the topic request! Happy learning!

  • @mariomoralestorres1285
    @mariomoralestorres1285 5 років тому

    Excellent videos.
    Could you make a video about wireless HART technology?
    Best regards from Mexico.

    • @realpars
      @realpars  5 років тому +2

      Thank you, Mario! Thanks for the topic suggestion as well, I will pass it on to our creator team :). Happy learning!

    • @realpars
      @realpars  5 років тому

      Hi Mario!
      Please take a look at this video and blog post about Hart Protocol which may help you; realpars.com/hart-protocol

  • @pridechigamba1664
    @pridechigamba1664 5 років тому

    Good job

  • @jermanskingsly7773
    @jermanskingsly7773 6 років тому

    super sir thank youu

    • @realpars
      @realpars  6 років тому

      You are very welcome!

  • @user-of3yt1bs5o
    @user-of3yt1bs5o Місяць тому

    Typo at 7:07. Should be Cable Length, not Lenght.

    • @realpars
      @realpars  Місяць тому

      Hello, @user-of3yt1bs5o. Thank you for your question. Great catch! That video was made five years ago before we started using better editing software. We are watching these mistakes closer today, and it shows you were paying close attention to the video! Thanks again!

  • @muhammadaashir7582
    @muhammadaashir7582 6 років тому +2

    do make one video on AS-i

    • @realpars
      @realpars  6 років тому

      Hi there, I will pass your topic request on to our creator team.

    • @Ryarios
      @Ryarios 6 років тому

      It would be interesting to see a comparison of all the major data highways used in control. It would probably have to be a much longer episode though.

  • @elijahpuga3501
    @elijahpuga3501 5 років тому

    Can you create a video about Fieldbus?

    • @realpars
      @realpars  5 років тому +2

      Hi there! We actually have created a new video course on Fieldbus, it will be live this Monday!

    • @elijahpuga3501
      @elijahpuga3501 5 років тому

      @@realpars Thanks! Luckily can watch it before my examination starts! Thanks for the great content guys!

  • @firecasts
    @firecasts 5 років тому

    Is it fitable to bycicles communication?

    • @realpars
      @realpars  5 років тому

      No, I don't think so!

  • @pusatberk4193
    @pusatberk4193 3 роки тому

    there are master slave relationship in this protocol, how does devices communicate with scanner ?

    • @realpars
      @realpars  3 роки тому +1

      The scanner is actually a master. We call it a scanner because it polls each node address sequentially by node number and waits for a response from the node. Each packet sent by the scanner is "addressed" to a specific node with data specific to the request.
      Each node can be considered a slave node because it only responds to a request by the scanner. DeviceNet is similar to RS-485 but by using the CIP protocol over a CAN network layer, it is much more robust.

  • @chhabindrakumarbag
    @chhabindrakumarbag 5 років тому

    👍👍

  • @nguyenthanhquang8375
    @nguyenthanhquang8375 5 років тому

    nice

    • @realpars
      @realpars  5 років тому

      Thank you, Nguyen!

  • @cyrox1859
    @cyrox1859 6 років тому +2

    70% of broblems that i had with this system were the terminal résistance.

  • @adilsaiyad3684
    @adilsaiyad3684 3 роки тому

    if you can upload a video explaining different protocols Briefly , and put that video just before this one.

    • @realpars
      @realpars  3 роки тому +1

      Hi Adil!
      Thanks for your comment and your suggestion. I will pass this on to our course developers!
      Thanks for sharing and happy learning!

    • @adilsaiyad3684
      @adilsaiyad3684 3 роки тому

      @@realpars thank you so much

  • @konradb2010
    @konradb2010 5 років тому

    I'd rather use ASi.

  • @MrAdamk1981
    @MrAdamk1981 6 років тому

    Now do ethernet and DLR

    • @realpars
      @realpars  6 років тому

      Thanks for the suggestion, Adam! Will pass your topic request on to our creator team.