Hi,

My in my application there is one central PLC which will be communicating with other 4 PLC which r 700 m from the central one. And 300m far from each other. I m planning to do peer to peer communication using CIP Generic method ie. by msg instruction. But my problem is How can connect all 5 PLCs. is there any wireless method available which will communicate upto 700m and will be cost effective to the end user.

This is highly dependent on the radios themselves. If the radios act like 802.11 wireless radios and form some sort of mesh network (this is not typical but there are some systems which can do this), then you'd just assign static IP addresses and use MSG instructions to read or write data, simple as that.

If however they are serial, you've got your work cut out for you. Under DF-1 protocol, it is assumed that there is only one other device on the cable. When you send a MSG command, there is a reply message involved and there is an assumption that there's no handshaking because there are no other nodes. Of course this doesn't work very well for your situation.

If you buy radios that operate in a sort of "broadcast" mode (whenever one node sends, all others receive), you can "roll your own" sort of protocol using the serial port. I know that you can definitely make the radios from Aerocomm do this and they will trivially work over the distances you are asking for (at least the 900 MHz models will). By the way, distance is inversely proportional to frequency...the higher the frequency, the shorter the distance (unless you compensate by using more coding and consequently either more bandwidth or slower speeds). This will be all serial so forget your MSG instructions. You'll have to manipulate the serial port more or less manually to do this.

The Aerocomm radios in a nicely packaged box with a DB-9 serial port on one end (do NOT get the ones with the Ethernet jack...you can't use them for bridging) cost about $120 a piece and you can buy them through www.mouser.com.

my quick internet search found this http://www.data-linc.com/faq/faqsrm.htm for a 900mhz radio capable of doing 20 miles.

The Aerocomm's claim about the same. One reason is because the wavelength is almost twice as long as 802.11 (2.4 GHz or 5 GHz). This makes 1" of steel effectively invisible and the lower wavelength translates to lower losses, almost tripling the range by itself. The consequence is less bandwidth.

The issue though isn't so much the radio itself. The distances are easily achievable even with conventional 802.11 gear. The data-linc stuff is pricey. If you want to see just what's possible in terms of range and speed, check out dragonwaveinc.com, or trangobroadband.com, or tessco.com. You can get gigabit Ethernet speeds licensed or unlicensed, even up to 20 miles away, but the price just might make you make alternative plans ($80K per radio is a bit steep for anyone except telecoms).

The issue is a problem of dealing with the fact that DF-1 is not designed to be a "party line", peer-to-peer protocol. So unless the radio itself has logic to take care of that issue (such as a "mesh network" radio), you've got to deal with the "party line" nature of the system and write your own protocol.

Do it the old fashion way.
Run fiber.

The heading says ML1400. So what is all the discussion about working around serial commnuications? Should be able to communicate Ethernet I/P to the other ML's via a wireless ethernet device. Am I missing something?

We have had success with MOXA equipment. You might want to look into moving up to a small compact logix and use an ethernet i/p gateway with wireless I/O. Banner has some pretty good offerings that are fairly cost effective. I am a pretty recent convert to this as we had a bad experience about 5 years ago. I guess technology has moved forward.

Fiber can get very pricey when you are paying $200+ per transceiver. Plus, the original spec was to communicate to a central PLC 700 meters away from all the others. The spacing between was stated as 300 meters.

If you run single moode fiber (and pay for the expensive laser transceivers @ about $350 for the cheapest LH last I looked), that will get you the range on single mode fiber. With multimode, you can reach 300 meters MAX with a standard (FX or SX) transceiver. You can go beyond this to 550 meters max. on an LH transceiver (it's called 1000LX mode), but not 700 meters on multimode fiber without having a repeater.

So fiber doesn't really look that great because at least based on the description the practical option is to run all single mode fiber. Plus, an all fiber switch is OK in terms of price if you are using only 100FX but in this case because of the range concern, you'd be using 1000LH transceivers which are quite a bit more money, especially for an entire switch populated with them. This makes the wireless option, even with the more expensive radios, look pretty darned cheap.

There was a big cost concern and no concerns for speed so I assumed serial comms. And I think I originally mentioned both.I didn't see anything mentioning ML1400.

f Ethernet is available you can use wireless Ethernet modems but they can be pricey...figure about $1200-$1500 per radio for industrial grade units. You will of course have to figure out how to place them in terms of antennas, figure on lightining protection if it's outdoors (and enclosures), etc. Some popular brands include Hirschmann, Cisco (but I have had lousy luck with theirs), Esteem (pricey), and Prosoft Technology. One possibe option that I'll throw out there is to head on over to tessco.com and buy a low end Motorola Canopy base station and radios off it. This is carrier/ISP grade equipment instead of industrial/consumer. The main radio isn't badly priced and the nodes can be had for about $500-$600 at most (my friend in the telecom business claims he can get them for about $250 in bulk orders).

Many wireless modems are well designed for industrial/noisy environments but contain a few surprises. For instance, the frequency hopping radios from Prosoft can punch through just about anything (similar to the performance from a Canopy system compared to el cheapo 802.11 stuff) and can do hopping from node to node but only do about 100 kbps for speed.

That's another thing to be aware of. If you go for 802.11 gear, it's very cheap and readily available. And in general, it will often interoperate with equipment from other vendors (including your laptop). But, it's also cheaply made. Most of it doesn't handle 24/7 operation without lots of little glitches and screw-ups. I have tried lots of brands (and so have my telecom friends) and the overall conclusion is STAY AWAY. It is consistently lousy even if you get one with say an Atheros chipset for the radio from a reputable brand name. It seems to be pretty consistent that if the stuff wasn't such crap, it would be a great option.

I'm not saying that 802.11 stuff isn't OK for home use. But if you are planning on putting in a dependable 24/7 system for PLC's, then don't touch this stuff. It's kind of like using an office-grade PC for server or industrial PC work...it might work OK for a few weeks but it will fail prematurely.

The problem now arised is for using Radio Frequency modem( PROSOFT) customer needs to buy licensed Frequency Bandwisth which he is not ready to do so. WLAN can communicate only upto 100m. There is only one way now traditional way. But we are planning to take all 5 PLCs on Modbus in master slave method. When last time i used modbus for AllenBradly VFDs with PLC in daisy chain method. Whenever one slave device goes into fault all the next goes int fault. I didnt find the solution for that but hopefully the same thing will not happen here in case of 5 PLCs

I don't know which country you are in but there is usually always a frequency band available (and a radio available) that is unlicensed. Be careful to look at ALL the options Prosoft offers. The FHE's can be set up for several different bands.

As to range of "100m", simply not true, EVEN with 802.11. If you go to serial modems (lower speed, better distance gains), I've had no problems going over 1000 meters with NO poles or external antennas across a plant (not even outdoors) or miles outdoors, as long as it's line-of-sight or nearly so. With higher bandwidth, generally you get less distance. However, 10-20 km is not unusual at all. The lower the frequency by the way, the longer the range. Every time you cut the frequency in half, you double the same range, given the same radio. That's why shortwave radios (<30 MHz) can reach over 1000 km without hopping off the atmosphere, or around the world when atmospheric conditions are right with less than 10 watts of power.

You can reach practically up to about 25 km point-to-point or up to about 10 km in point-to-multipoint mode with "typical" WLAN radios. The key is to get rid of the cheap, crappy antennas that they come with and put real antennas on them. An "omnidirectional" (all directions) antenna can increase range by decreasing the amount of signal received from space or earth (pointed up or down). The gains are modest but 500 meters or more with all omnidirectional antennas is not unheard of. A directional antenna has to be aimed but for your scenario (one central station and a bunch of satellites), this would be ideal. If you put 24 dBi antennas at the satellites and stick with a 6 dBi omnidirectional at the center, you will have a gain of about 24+6=30 dB of gain. Those little "rubber ducky" type antennas on home equipment are only good for usually -3 to perhaps 0 dBi of gain. Every 6 dB corresponds to doubling, so you'd be at 2^5=32 times the original number. Since you said 100 meters, 100 x 32 = 3200 meters. This is easily more range than your requirements and leaves room to deal with obstacles and other noise issues.

Since this is your first attempt though I highly suggest you talk to someone at Prosoft or B&B Electronics for instance. It is very important that you get everything right so that you don't have a disappointing result.

So is it possible connecting all PLCs by only using WLAN? Whatever u r saying abt this gain theory is hard to understand for me. Can u explain in Little detail or provide me some material or link. My project is in a refinery plant. There are so many obstacles in routes.

Yes, it is always POSSIBLE. Consider the ELF radio system used by the United States to communicate (albeit it VERY slowly) to submarines at the bottom of the Atlantic ocean (to send nuclear launch codes). Of course it helps when the transmitting antenna stretches across 2 states!

But, refineries are very tough. You shouldn't even consider wireless unless you do a good quality survey first to verify whether or not it will work. Chances are that you are going to have to mount your antennas up in the air above the pipe racks at a minimum. Certain radios can also do packet hopping...they will pass packets from radio to radio to the destination. These are also sometimes called "mesh" radios. This is still relatively new technology.
http://en.wikipedia.org/wiki/Wireless_mesh_network

First thing to remember is that when a wave is travelling through a substance, ANY substance except vacuum, it is slowly losing strength. This is absorption. Second, any time that it hits another substance, the wave is split. Part of it is reflected, and part is transmitted through the new substance. How much goes each way depends on the properties of BOTH substances. This dual-nature is most obvious when you think of seeing your reflection in water, and still looking at objects beneath the surface at the same time. Even more confusing, the properties are wavelength dependent. Here is the absorption vs. wavelength for air:
http://en.wikipedia.org/wiki/File:Atmosphe..._or_opacity.jpg

Also, the size of the object is very important. If the size is less than about 10% of the wavelength, then it is effectively transparent. There is a little bit of scattering going on (which is the reason that you see light beams when there is dust in the air) but that's about it. Between 10% and one wavelength, the properties go from pure transparency to the normal situation of partly reflecting/partly transmitting.

So at 5 GHz, the wavelength is just 6 cm. Any object thicker than about 0.6 cm starts to interact significantly with the radio wave. The wavelength at 900 MHz is now up to 33 cm, so a 2-3 cm thick slab of metal is still effectively transparent. Concrete is a little more complicated of course because it is so porous.

The best, least complicated situation of course is "line of sight", clear of obstructions even within the first couple Fresnel zones (http://en.wikipedia.org/wiki/Fresnel_zone).

Without that, what happens is that you typically see a "rich scattering" environment. Even if your antennas are not sufficiently high off the ground, you end up with two waves, a "space wave" (direct) and a "ground wave" (reflected). They have different path lengths and they cause an interference pattern which leds to "hot" and "cold" spots. It is still possible to communicate, and there are now multiantenna receivers that are designed to take advantage of scattering. However, it is important to be aware of this effect. Of course you might be able to simply put the antennas up on tall enough poles to eliminate the problem in many cases.

Radio calculations often involve very large numbers, so communications engineers always work in decibels. To convert to decibels, take the base 10 log of a number and multiply by 10. To convert back to "normal" numbers, divide by 10 and calculate 10^(number). More info here:
http://en.wikipedia.org/wiki/Decibel

Once you are in decibels, multiplying and dividing becomes addition and subtraction. One interesting calculation to consider is that 2 is equal to 3 dB. So every 3 dB is the same thing as doubling (or halving). 6 dB is quadrupling, and 10 dB is equal to increasing signal strength by 10 times. 20 dB is 100 times.

Here is the link budget which is used to calculate gains or losses in a line-of-sight scenario:
http://en.wikipedia.org/wiki/Link_budget

Note that there is a gain for an antenna. Why would this happen? Well, except for the theoretical "isotropic radiator" (perfectly spherical antenna), all antennas are directional to one degree or another...they do not receive equally well in all directions. As you narrow the range of directions that an antenna can receive in, it increases the signal strength in the direction that it does receive in.

Let's for instance imagine a big 3 meter satellite dish. They are highly directional. Any signal not directly in front of the dish even 100+ km out into space is invisible. They often have gains around 36 dB (signal strength is increased by nearly 4000 times!)

Now a satellite dish is quite directional, and it is pretty normal to see directional antennas with gains of between 10-30 dB. An omnidirectional antenna can also have a gain. It does this if you think three-dimensionally. It simply concentrates the signal into a plane or toroidal shaped region pointing out from the antenna, while signals received straight up or down are tuned out. The gains are more modest but 3-9 dB is pretty common.

So if you used a 6 dB omnidirectional antenna and a 20 dB directional antenna pointed at it, the signal strength would be increased by 6+20 = 26 dB. Converting this to decimal is 10^(26/10) = 400 times increased signal strength. Doubling the distance cuts the signal strength to 1/4 (waves spread out like a sphere...so the area on the sphere is what matters). So increasing the distance from 100 meters to 700 meters decreases signal strength by 17 dB, or 50 times. So the net result is an increase in signal strength of 8 times (9 dB).

In reality, there are just too many variables to consider. So most folks do a radio "survey". It's not too hard to do. If you have a wireless access point and your laptop will show you signal strengths (I recommend using netstumbler which is free software to do this), then you can do a very primitive survey. Set the access point up temporarily and start walking around with your laptop. It's not perfect and it's limited to the frequency that your access point & laptop support, but it should quickly give you some idea of what is possible/practical and what isn't. You can even get a better quality antenna to attach to your access point if you want to experiment further. Most industrial grade radios have some sort of signal strength (survey mode) built into them as well.

Paul:

Do you have some Navy experience? Not many people know about the ELF system. I retired with 21+ years of submarine service in 1989.

Bill

No. I used to live within 50 miles of the transmitting antenna when I went to school. In my EE degree, you had to specialize in 2 of about a half dozen areas. I chose analog electronics and communications. So when you're that close, ELF is a topic of conversation from time to time. At the time, the biggest water cooler topic was whether or not ELF would cause long term health or environmental effects. Answer: it does. Mostly, it seems to enhance vegetation growth near the antenna. There are other noticeable effects (bacteria growth in patterns aligned to the electromagnetic field lines) but nothing that could be classified as harmful...of course the environmentalists try to conjure up all kinds of unknown, cataclysmic effcts.

Not to disagree Bill, but anyone who read or paid attention to "Hunt for Red October" saw a Hollywood demonstration of the ELF on a sub. Probably not totally accurate, we all know how Hollywood is.

Ah, yes, Bob. The operative words "paying attention" and on top of that, having a clue what it all meant. I'd wager only the techies picked-up on any of that in the first place. It is not like ELF was our only way of staying in communications while submerged. The 1MW VLF transmitter in Cutler, Maine was our mainstay (in the Atlantic pond) for many years. In the winter, they would put a continuous carrier on the air and wind the transmitter up to 2MW long enough to melt the ice off the antenna array.

Enough of the good ol days. I don't want to hijack a thread.

Bill