Micro 1100 high speed inputs

[kcor 1]

Have a machine running in Honduraus that has a 1100 micrologix. Question - They have two encoders that are rated at 1000 ppr. They want to mount them on a roller that is moving at around 100rpm. From reading the 1100 manual, it only has one high speed counter. There are 4 high speed inputs. Are the high speed inputs capable of detecting the pulses from the encoders, or should I recommend they order a micro 1500, which has two high speed counter. The 1000ppr * 100rpm would be 100,000 pulses per minute. If I divide that by 60, I get 1666 pulse per second. Can the 1100 handle that. If I am not doing something wrong, I think the 1100 has a scan time of 5ms. So if I take 1000ms for 1 second and divide that by .005, I get 200,000. Any advice would be appreciated. Thanks

[paulengr 44]

What's important is looking at the COUNTERS. You can't really handle it with just a scan (5 ms = 1000/5 = 200 scans per second). So instead of trying to actually detect each pulse, you use the counters to keep track of the number of pulses. That's what they are for. The only other alternative (which won't work either) would be to trigger the programmable interrupt (if it has one...can't remember this part of an 1100). The counter (directly from the documentation...RTFM) is capable of handling up to 40khz, or 40,000 pulses per second...much faster than you are talking about. Unfortunately though it only has one counter and you can't get another high speed counter as a 1762 I/O card. And the counter won't handle quadrature inputs (not sure if this is an issue or not), so it won't quite do everything you want. So you essentially have two choices. You can upgrade to a Micrologix 1500 or you can add a second Micrologix 1100 to use simply as I/O. Considering the cost difference, I'd probably use the second Micrologix 1100. You can connect them together either using a null modem cable through the serial ports, Ethernet crossover cable, or Ethernet switch. Then simply MSG blocks of data from the slave unit to the master. Usually it's easiest to use reads from the master since this both minimizes traffic and makes it easy to monitor what's going on with the communication link. A third possibility which you didn't mention is to use a different type of encoder. You can get one with absolute coded outputs and wire this up to the inputs (or an input card) or get one with an analog output. Either one will require at most that you add an input card to the Micrologix 1100. I believe you can also get external high speed counters with analog outputs if you want to stick with the encoders, which would serve the same function as the second Micrologix 1100...but since I don't normally shop for these things I'm guessing to a certain degree. The few encoders I have are either resolvers that have on-chassis I/O boards (Gemco or AMCI) or I have a few analog output encoders from Electro Sensors (call them...you wouldn't recognize it for what it is just by looking at the catalog or web page because their primary markets are mining and grain).

[Ken Roach 194]

What will the encoders be actually doing ? Are you counting a number of pulses then firing an output (what they do best) or determining a speed based on the rate of encoder pulses ? I strongly recommend getting a MicroLogix 1500 controller that has the two High-Speed Counter hardware circuits.

[kcor 1]

Thanks It appears that it will have to be at least a 1500. only problem with that is , it is almost $700 just to add the ethernet to the 1500 (which they would come out with a 1500 that has ethernet). Will need that option in order to connect to it from here in the states if we have to troubleshoot it. It may be best to go with the slc 505. The encoders are going to be used to show the rate that fabric is being fed. Just for testing, I used a 1200 on the bench, and based on some info I got from a earlier post on this site, set up STI and HSC on the 1200 on my work bench in the lab. I connected the encoder and then used a cordless drill to turn the encoder, and with the STI was able to accurately measure the rpm that the encoder was running. To double check, put a hand held tach on the drill, and the plc and the hand held tach showed the same rpm, so I felt pretty good about using the 1100. But now they want another encoder to be added and that takes the 1100 out of the equation. From pricing I just got from ab rep, the 1500 with two analog cards, one extra input and one extra output was almost twice as much as the 1100. Not got any prices yet on the slc 505. Thanks for the suggestions

[kcor 1]

Wow, was I off. Been sitting in airport since 4am waiting on a plane, got bored, went out to this site, ( which is in my opion, the best there is ). Big differnce between 200,000 and 200. What I should have done was just went and got a usa today and read it. 1000ms and it takes 5 ms, just as you say is 1000/5=200, Not 200,000, I should have read what I posted again. Thanks for keeping me straight.

[TimWilborne 108]

If it will also need an operator interface, look at the Redlion G3. It is capable of routing through a virtual serial port from its Ethernet port to it's serial port. So you can come in on the Redlion Ethernet port, out of its serial port, and into the port of your Micrologix or SLC without having an Ethernet port on it

[kcor 1]

Thanks TW Always look forward to your input. TW Does that mean that you could have the display connected to your pc through the ethernet port, and bring up the software for the display and work on the display, or use the rslogix500 and look at the plc? Sounds good. Have you actually used one with a micrologix? Thanks

[TimWilborne 108]

I've got a rough draft of a Reflion G3 First Impressions article that is similar to my Micrologix 1100 First Impression article. I need to finish it up, it would answer a lot of your questions. What I was commenting on was the G3 ability to route from the Ethernet to the serial port for a virtual serial connection to RsLinx. This will allow allow you to go online through RsLogix or use an OPC/DDE server to connect to a database. Also, as far as monitoring it has a built in Web Server which you can view the screens that you have made for the G3 or you can make completely custom web pages. BUT, if this is the same project you were working on earlier, you might want to tell us what you decided to do. It was very marginal for the Micrologix with the I/O you were going to need. It seemed more of a job for a SLC, or my preference, the Compactlogix which gives you much more bang for the buck.

[paulengr 44]

Now we are getting somewhere. I wouldn't say they are the cheapest (I'm not sure who is) but all you are trying to do is measure RPM. I've had very good luck with the tachometers from Electro Sensors. They directly output in 4-20mA outputs. They are not the only game in town in this regard either. Since you'll be feeding the information to your PLC in 4-20mA format, the hardware (tachometer) takes care of the actual raw conversions. This will be less expensive than adding additional PLC's and/or PLC hardware. You can also get stand alone high speed counters but the pricing is usually pretty bad. I believe that the Electro Sensors unit usually only uses about 10-100 PPR. The reason is because it uses the timing technique instead of pulse counting. It is also important to understand the speed and resolution requirements that you have. As speeds (and pulse rates) get higher, the pulse counting technique gives you a progressively higher resolution. The fastest way to increase resolution is to go with more PPR using pulse counting. The exact inverse is true if you are timing the pulse lengths. At lower speeds, pulse timing is much more accurate. With car engines, the PPR "disk" (if you want to call it that) until very recently was usually 1 PPR. They didn't use pulse counting at all. Instead, the engine computer measured the time interval from one pulse (from a cheap magnetic switch) to the next. Depending on your resolution requirements, you could potentially read up to 4 encoders simultaneously with hardware less than $100 PER ENCODER using pulse width (timing) instead of pulse frequency (counting).

[kcor 1]

Thanks to everyone for your suggestions. The encoders are already mounted on the machine and our manager wants us to reuse them. Have already tried to talk him into something else, but he said that we would keep the encoders. They are ab encoders with 1000 ppr. TW What this is - It is a piece of machinery, that to make a long story short- is feeding rolls of cloth through various chambers, etc. that apply finish, dye, etc to the fabric. After all this is done, we are going to build a prototype of a chamber that when the fabric goes through it, steam will be applied based on the humidity in the chamber. There is entrance motor and a exit motor as well as 4 other motors in the chamber that will be used to feed the material. There is a temporary chamber at the moment with 4 motors that I put together with older baldor drives and motors that I took off some old equipment that are being used to control the speed of each motor individual with a seperate pot for each drive. That is what our RD folks wanted to experiment with, and only had a couple of days to build it, so that is what I put together. Quick, and simple. Now the concept seems to work and they want us to build another quick prototype, with the 6 motors and use the speed of the entrance motor to automatically tell the other motors the speed to run. The entrance motor will be slaved to another motor that will tell it the speed to run, ,and then the entrance motor will be the master to other motors. Seemed like a good application for control logix, but this thing has to be done in a couple of weeks, also, we do not have the control logix software, or any control logix hardware, so our manager decided, that for another "quick" prototype with just the 6 motors, that we would use the two encoders that are already on the equipment, a micrologix 1100, ( but now a 1500), a powerflex drive for each motor, a panelview 300 micro display. His reason behind this, is the fact that the prototype will only be used for a couple of weeks then trashed, so he is not going to dish out the money for control logix. So, I will use the 1500 with the two hsc inputs, then use that to send a analog signal to each powerflex 70. If the concept works, and corporate approves the money, then we will go to control logix, because then overall, there will be 27 motors all slaved together. But for now, the 1500 is the only option that I have. Thanks again for the advice. Use to be 8 other guys who worked with me on all of our stuff, now only two of us, and without this site, not so sure what I would do. May have to do something about the other one that works with me though, he used my laptop at work a few days ago to go to another site that I use from time to time and posted some stuff that he should not have, I am still trying to figure out how to handle it, one thing i learned, is to alway log my pc off when I leave the office and go to the lab, but that is a different story. thanks alot

[paulengr 44]

If you are using a drive, why don't you just connect the encoders into the drives and do full vector control. Then you can use the analog output from the drive for the conversion and not even bother trying to do the RPM stuff yourself.

[kcor 1]

Thanks Thought of that, but not quite so sure at the moment how to do it. Have you ever done this, and if so, could you share the application? Also, liked the looks of the hmi that TW suggested, but boss says, no, plc and hmi will stay allen bradley. Thats just the way he is. Now, is there a way to add ethernet to the 1500, and still have a serial port to connect a display to. I know ab makes a device to allow you to to connect to the plc by using the serial port in and then be able to use the ethernet connection, but then that does not leave a extra serial port for a hmi such as the 300 micro. any suggestions

[TimWilborne 108]

If you really want the routing and Ethernet, look into the Redlion Data station. It will give you Ethernet, the routing and the web server without the display. There are other brands that can do this also, Redlion just gives you the most bang for the buck in my opinion. Then you can use whatever brand display you want. One of the 1500s has two serial ports though I can't remember what protocols the second port supports

[kcor 1]

thanks tw will call my ab rep tomorrow and find out

[Clay B. 1]

LRP is the CPU on the ML1500 that has dual com ports. Both ports have the same ablities just one is a Sub D 9 and the other is the "micro" port. You need to figure out which one you want to use with the ENI card so you get the right cable. One thing to note: Your using 2 encoders for rate measurement. ML1500 allows only one STI so you will have to sample both encoders at the same time. I can not access my program sample library to post a sample but I do remmeber I got sample code from the knowledge base.

[kcor 1]

thanks for info

[paulengr 44]

As to your "lost serial port problem", you can get a Digi One IAP. This has 2 serial ports and an Ethernet port. You can software configure them any way you want. One of the most obvious things is to configure the second serial port as "passthrough" while still routing CIP or AB Ethernet packets to and from the Ethernet port. This way you essentially get all the advantages of an AB NET ENI but you still have a free serial port. Source: www.bb-elec.com. The IAP's aren't cheap ($600-$700), about the same price as a NET ENI. The first time you configure one, use the tech notes on the web. How you configure it is not very obvious at first. By the way, the biggest reason for having that free serial port is so you can connect to the processor with a laptop. If you've got an Ethernet port and you add a small (<$200) dumb switch in your cabinet (highly recommended), you've just networked your processor and you can plug your laptop AND your HMI in at the same time and be programming/troubleshooting all of them at once. Yes, have used drives many times before in different configurations. Anything that provides "sensorless flux vector" is going to give you speed output WITHOUT an encoder. Anything that supports an encoder will work, too. The thing is that these options are so inexpensive that it is actually getting to be like soft starts...it is actually getting harder to find the less expensive drives that don't do vector control. There are many other drive manufacturers but since your boss is strictly an AB guy and this is an AB forum, I'll use the AB terminology. A Powerflex 4/40/400 series for instance won't work. These are true "volts/Hertz" (old style) drives. They are good enough for controlling speeds on pumps and fans but not even recommended by AB or anybody else for your application. As the load (and %slip) changes, the motor speed will change. Since a volts/Hertz system doesn't compensate for this, you lose speed control. Vector drives keep track of the shaft angle in some way and give you compensation for changing torques. The entire Poweflex 70/700 series can do this, because these do sensorless flux vector or encoder-based control. They are essentially 90% of what you get when you buy a servo drive, but work with regular AC motors (within their limitations) instead. Setting up either vector mode is so simple on those drives that using Volts/Hertz is almost like driving an automatic transmission in "1" or "2" instead of "D"...it's there but unless you are doing something very strange and unusual, why would you ever want to use it? An automatic transmission is...well...automatic. A vector drive is vector, not Volts/Hertz. As to applications...hmm. Let's see. I have an oven that has a conveyor chain running through it. Someone bought a Reliance vector drive and put it in Volts/Hertz mode. I changed it to vector for better torque control and because I got a torque monitor "for free" (from the drive). Vector drives provide faster/better starting/stopping so if I put one in, I can usually buy a regenerative drive and I don't have to pay for brake resistors or mechanical brakes. There are several places where I have variable torques such as cement augers where the loading changes. Vector drives make the speed control constant. I have several "cars" moving lances (the product is 20 feet long) for painting, lining, grinding, etc. All ideal vector drive applications. Grinding stones by the way are a torque-based application. So rather than using speed control, I can use torque control which reduces wear and tear on the grinding stone...the motor "automatically" speeds up and slows down depending on how "hard" the grinding is. More notes: 1. If you use Sensorless Flux Vector control, you get RPM readings for free. Just configure an output channel as RPM. In this mode, the drive monitors the current (actually the phase difference between the generated voltage and the current draw) and directly measures the motor slip. From there it is a very simple matter to calculate the actual motor RPM (output frequency X (100% - slip) = actual motor RPM). If you play with a drive in this mode the first thing you'll notice is that it is much more smooth and very responsive compared to Volts/Hertz since the drive is no longer "guessing". 2. Check out for instance Publication #20D-AP004A-EN-E which shows you how to use an encoder to do absolute positioning control. Essentially this is for doing servo-control with regular AC motors when a true servo motor is overkill. This is much more advanced than you need but gives you some idea of what is possible. 3. Wiring up an encoder for speed input is very easy. See page 1-16 of the user manual for a Powerflex 70 for instance. You don't really "do" anything except to wire up the encoder and configure the drive to use the encoder. User manual is document #20A-UM001L-EN-P. This is for a quadrature encoder (sounds like what you have). If you have an absolute encoder, then there's a lot more involved and you need to follow the above document because these encoders are more sophisticated than a simple quadrature encoder. 4. Whether you use Volts/Hertz, sensorless flux vector, or an encoder is very simple because the drive ASKS you that question when you set it up. If you use the drive wizard ("SMART"), it will ask you to which mode you want and then configure it for you as you go along. It might even be possible to configure the drive as say Sensorless Flux Vector and still read the encoder output via some sort of arcane drive parameter number but when the drive parameters number into the hundreds (never mind the "wiring" type modes), I'd suggest you call an AB rep to find out whether you can do this or not...it might take a while to find the right configuration stuff on your own.

[paulengr 44]

Additional followup...I just had to spec out a drive for a job I'm doing today that's closer to your project dimensions. A Powerflex 40 will run about $700-$800 for a 480V 3-phase 5 HP motor, depending on the discount rate you get for AB stuff. It does sensorless flux vector. It will output the RPM directly to an analog output. So it can do all the speed information "for free". If you go with the 1500, you can also "directly" interface to it via Modbus RTU. You just need a voltage shifter (AIC) to go from RS-232 to RS-485. The Powerflex 70 will interface directly to the encoder you are talking about. This gives you even tighter RPM control. This gives you full vector control. In terms of response, Volts/Hertz gives you 10% speed error control. Sensorless flux vector gives you 1-2% speed error. Full encoder vector gives you 0.1% error. I checked around the competitors. Unless you are going super cheap (automationdirect.com or somebody like that), at least in the "small drives" market, AB is very competitive as far as I can tell in terms of pricing.

[kcor 1]

thanks alot lots of good info