RS485 needs 3 conductors and a shield. Many people say its a two wire network but it is not.
Two conductors are used to carry the RS485 Differential voltage signal.
The Shield is connected to earth/ground at one end only and provides shielding against induced noise.
So why the 3rd conductor ?
The driver sends data by modulating the differential voltage. The receiver must sense and decode the differential. There are limits to the voltages the transmitters and receivers can work with. These limits are specified by the code. They are -7Volts to +12Volts. What happens if you have two devices and a ground potential exists between the two devices of 24 volts ? You can see that one of the devices will be operating outside the specified voltage range. While you might expect that all the electrical equipment in an installation is ultimately connected to the same ground in practice this is rare especially in cold climates where building architecture and frozen ground can conspire against you. That why you need the 3rd conductor – to connect the ground (of each RS485 driver) to the same reference. Now we don’t care about ground potentials.
Ever wonder why you blew a 485 device when you connected your laptop or computer ?
Its this problem – there is a ground potential. Thats why its good practice to connect your laptop’s 485 ground conductor before you connect the differential conductors.
Can you get away with 2 conductors ?
Yes. Thats why lab or factory tests suddenly stop working when installed at a site. In your lab or on your desk you can be sure the devices are all commonly grounded. Now if you measure the difference between the ground of a RS485 driver on one device and another device you will find zero volts.
Can you get away without the shield?
The value of the shield is controversial. If you are using a twisted pair and you don’t ruin the twists by unwinding them by more than an inch or two at each end then the shield probably is of little value. But, most cables come with a shield. If its not easy to terminate the shield or you cant be bothered then at least have your design drawings tell the installer to to coil and tape the shield drain wire so that you can use it if you want to.
Where can you purchase 3 wire 485 cable ?
I don’t know. Purchase two twisted pairs with an overall shield/drain. Use one pair for the differential and join the conductors of the other pair to make your ground wire.
What size conductors ?
The bigger the better. Most installations are done with 24awg but remember the higher the baud rate the greater the signal frequency and the greater all kinds of captives and inductive effects. If you can, get bigger conductors.
What kind of cable ?
Choose one with designed to present a almost constant nominal impedance because it will make the addition of terminating resistors easy – just read the cable jacket and get a resistor of the same impedance. Most cables listed for 485 use are designed to present an almost constant nominal impedance. Almost constant means a cable who impedance is reasonably independent of length.
What about biasing ?
If a RS485 driver is idle then what ? The line is disconnected from the driver in its idle state which means that the + and – are floating. What happens if for an instant there appears a differential voltage of more than 0.2 Volts? Well a differential greater than 0.2 is considered a signal and hence constitutes data noise. Solve this problem by pulling the lines to no voltages when they are idle. To do this connect them to ground or some other voltage using pull up/pull down resistors. Nice vendors include this. Nicer vendors provide a choice of biasing resistors selected by switches/jumpers. The reason you might not always use the same value is that the cumulative effect of many device’s biasing resistors may make it impossible to signal at all. How do you calculate the value of a biasing resistor ? How would you know where to connect it ? Good luck getting answerers to these questions. An easier question to answer is this: How do you know if you need to bias the idle state ? If you have a scope you can see what the idle state floating voltages are ? Don’t forget you can only measure this when the device is idle.
For more information on RS485
Please contact us, we can resolve your issues with RS485
© Chipkin Automation Systems 2007
This is just some small feedback with regard to the Ground connection for RS485 that we discussed two weeks ago on the ECC:
1. Yes, the Ground wire is required, and thus RS485 is not a 2-wire network. Very few companies indicate this in their documentation and diagrams. Even reputable semiconductor manufacturers will only mention this connection in passing, and then one specifically has to look for the issue.
2. But, the problem is that a ground loop can be created when the power to all RS485 terminals is not isolated from each other. This is a problem since in most installations for commercial purposes, unskilled people are used. They generally don’t fully understand the principle of EARTH and RETURN connections.
3. Thus, normal installations will receive their power from one central power supply or have separate SMPS, but if these are not specified properly, then the EARTH and NEUTRAL will be connected by either a short or some small resistance to each other on the low cost versions!!!
4. The whole grounding issues thus becomes the problem of the equipment supplier and in particular the installation / hardware design engineer.
5. Most ‘isolated’ RS485 Ground connections refer to the main Ground with a 100E resistor. This is fine if the common mode current (low frequency syndrome) is minute in the cabling. This phenomenae can be minimised by using a Shielded Twisted Pair where the shield is normally terminated only at ONE side of the equipment. In applications, such as ARC furnaces this will NOT work since 100′s to 1000′s amps are generated, and the ground wire will be conducting.
6. The track on the PCB thus becomes a fuse when the rating is exceeded e.g. 8 mil track maximum current carrying ability under optimal conditions is 500mA.
7. Proper isolated RS485 networks remove the low cost resistor implementation, and use a proper transformer to generate a isolated RS485 ground. This is also the best and most expensive circuit to implement. This will minimise common mode current pickup and voltages spikes on the line, but each installation has its own unique requirements (the arc furnace is just one of worst case examples that I know of…)
Regards,
Arno Netter.
You wrote this in one of the blogs about rs485:
2. But, the problem is that a ground loop can be created when the power to all RS485 terminals is not isolated from each other. This is a problem since in most installations for commercial purposes, unskilled people are used. They generally don’t fully understand the principle of EARTH and RETURN connections.
3. Thus, normal installations will receive their power from one central power supply or have separate SMPS, but if these are not specified properly, then the EARTH and NEUTRAL will be connected by either a short or some small resistance to each other on the low cost versions!!!
4. The whole grounding issues thus becomes the problem of the equipment supplier and in particular the installation / hardware design engineer. I’m currently in Kuwait working on a project which involves rs485 connections. I face problems with the rs485 connection grounding. Right now, I’m using a Robust DC 485-fiber convertor to convert my data from my substations to transmit across the FO to my local control room. At the local control room, I’m using same convertor to convert the signals back to 485 signals.
My problem is there is no common reference which ideally should have. Now I’m having line breaks in communications are time and data “hanging” during transmissions. Out of the 8 units I have, only 1 is working.
Can you please help me with it?
Dear Chong,
“I’m currently in Kuwait working on a project which involves rs485 connections. I face problems with the rs485 connection grounding. Right now, I’m using a Robust DC 485-fiber convertor to convert my data from my substations to transmit across the FO to my local control room. At the local control room, I’m using same convertor to convert the signals back to 485 signals.”
Are you using the recommended isolated power supply as per Robust DC specification or perhaps another type. I stress that this must be a ISOLATED power supply that supplies power to the transmitter and receiver of your RS485 convertor. When you measure the GND on the input side of the power supply with reference to the power output ground/negative, this must be open circuit or at least in the mega Ohm region. In effect, this is also called in industry an isolation transformer that is not always identical with a ‘normal’ transformer, although some
“My problem is there is no common reference which ideally should have.”
That should be the problem of the isolated power supply as these units are referenced on their input side to each other over distance. The communication should not suffer unless there are other problems associated with your network. RS485 is a differential pair, thus in your specific application with fibre their is no need for a common reference, but that depends on the driver/receiver pair in the actual convertor. How are these driver/receiver pairs (the actual semiconductors inside the box) referenced to the power of the actual unit. Are you introducing additional external resistors to make a common? If so, why because from a quick inspection of Robust DC datasheet this is not required?
“Now I’m having line breaks in communications are time and data “hanging†during transmissions. Out of the 8 units I have, only 1 is working.”
I suspect that your problem is more related to a ‘bad’ fibre optic cable. Do you use single mode or multi mode cable in your installation? Are you exchanging these units with each other to confirm a bad unit? Or are all the units on the fibre optic cable at the same time? Has the fibre been properly terminated? Is there is damage to the cladding? Are you within the length specification of the fibre? Are there any loops in the fibre cable that will cause higher optical dispersion in both single mode, but especially multi mode fibre? Are using exactly the same fibre optic cable for all the units?
“Can you please help me with it?”
For help, more information is generally required. Also never underestimated the physical environment in which you are installing and operating equipment. Kuwait for an analysis would imply a desert which implies relatively little humidity thus higher electro-statics, thus your equipment needs to be more robust to withstand continous static discharge. You probably have a shot driver/receivers on either the fibre interface or the RS485 interface especially when you say that data is intermittent.
But this is all maybe’s since I do not know more about the installation, the environment, etc.
thank you.
RS485 is a 2 wire network if done right. Profibus is a typical example of a true 2 wire RS485 system. The secret to trouble free RS485 is to use isolated power supplies to run the RS485 transeiver. Low cost, low current, high speed non optical isolators are now available from Analog Devices (the ADUM1300 series) and simple to implement power supply isolators are also available from Maxim (MAX253 & MAX845). Combining these parts with any standard RS485 transeiver builds you a fully isolated interface that truely needs only 2 wires. You must provide biasing resistors with this setup to center the RS485 signals within the isolated interface common mode range. A sample of an isolated interface can be found in the booklet “The rapid way to Profibus” or I can send you as sample circuit if you are interested.
Hello Douglas,
I’m sorry but basically you are wrong.
Voltages are not defined by themselves, they are only defined as the difference in the force of the electrical field. To measure a voltage you always need two points to put your probes on.
In RS485 one point is the signal wire A or B, and the second is the local ground of the receiver circuit. When you have an isolated device, the voltage of the isolated ground compared to the local ground of your device is undefined. This is especially true for the voltage difference between the signals and the isolated ground, because the signals are generated in respect to the local (isolated?) ground of the *transmitter* circuit. When these two grounds are completely isolated, there is no guarantee the receiver understands the data. It often happens the difference is too big. One proof of that is the real-world experience, transmissions become successful when you just add the ground line.
Biasing resistors do not really help, because for the purpose you describe you have to attach them to *every* device connected to the bus. This requires a recalculation of their values every time you add or remove a device from the bus.
(fact: calculation of resistors is not done by average customers.)
And in fact in this way you use the data lines to clear all voltage differences between the isolated grounds of the devices. At least you connect them via the resistors and the Data- line, which is pull-downed.
Even without bias resistors there may be some kind of parasitic connection between the data lines and the receivers ground, which also clears the differences between isolated components. That may also be the reason why in many applications the communication is fine with only two wires.
I agree, when using bias resistors as you describe it, there is only a reasonable current when a new isolated device is attached to the network. Later the isolated ground planes share the same electric potential, and there are only very small currents, if any.
By the way, even maxim-ic says the ground wire is required.
Hello Douglas,
Your point about Profibus is a good one but I think it re-inforces the the fact that rs485 requires 3 conductors.
The reason I say this is that Profibus is a standard that encapuslates RS485 and imposes other electrical requirments on it (like the isolation). Thus you get a bus with devices with a uniform electrical interface.
This is not the case with other RS485 applications. For example BACnet MSTP merely references the RS485 standard and does not impose other physical layer requirements.
I have to agree with Wilfried, viasing resistor calcs should not be required for field installations.
Hello Wilfried.
I think that the receiver side isolator output is measured by its ground. therefore receiver device can read data correctly.but the third wire is required as shield to reduce the noise.
Hi All,
I agree with Douglas, a third wire is not always necessary.
As long as the difference in ground potential between all the devices, including noise, is within the common mode range for TIA485 specification then it should work without a ground.
In this situation is side A more positive than B or vice versa, this is what is meant by differential signals.
This common mode range of TIA485 was mentioned early in the article, -7 volts to 12 volts.
If this is exceeded in the system than one or more isolation devices will be needed.
Though personally I would always use a ground.
Biasing resistors, not to be confused with terminating resistors, only need to be small enough so that noise on the signal line does not get through the receiver, at least 200mV.
This has become confusing over time as the original specification required the receiver chips to have a defined input resistance which means that there is always going to be current flowing into the input pins. This limited a bus to 16 devices owing to the drive available from the driver chips.
Since then there has been quarter power, eighth power and even sixteenth power devices, so resistors for biasing with early devices are now too low in value to work properly in large systems, the driver will become overloaded. This means that later systems will have higher value resistors, and can lead to providing insufficient bias for older chips allowing noise through the receiver and affecting the integrity of the communications.
Trevor
I have been struggling with the 2- versus 3-wire, grounded versus ungrounded, shielded versus unshielded issue with RS485 wiring for 20 years. Steven says pretty much the same as I have been saying for over 15 of those 20 (I was a slow learner).
Comments:
1) Shielding is always good. This can be proven theoretically and in practice. Double shielding is even better. Single shields must be grounded at one end, double (insulated from each other) are grounded one at each end. This prevents currents from flowing through the shield and inducing currents into the signal lines via the transfer impedance between the two.
2) Bigger may be better but there is a very serious law of diminishing returns as the gauge of the wire gets larger. High frequencies are carried on the surface of a conductor and the more surface area the lower the impedance — this is good. 18ga is overkill. 20ga is OK for very long runs, 22ga is good enough for pretty much every application.
3) There are RS485-specific cables out there. They have a shield & drain wire (ground at one end), a ground wire and a twisted pair for the signal. Belden 3106A is a good example.
I wish I could get more people to understand that carrying a third wire between RS485 devices was necessary. I wish everyone who implemented an RS485 device had put the third wire on the connector so you could connect to it.
Apropos PROFIBUS: Profibus may be RS485, but the reason it works is because the purple cable and connector is EXACTLY DEFINED and full network performance is only achievable using EXACTLY THAT CABLE (and connector). If you grab any old twisted pair cable use it to hook up your Profibus network, you deserve the problems you have.
Great information by everyone! I manage several thousand RS-485 devices in an oil refinery setting (control for Valves, pumps, and alarms). Our original vendor (in 1990) installed the networks without the 3rd wire for ground. I can tell you from experience, ALWAYS install a ground wire or you will have problems. But what if you don’t have a 3rd wire and can’t retrofit and existing installation? I use isolating repeaters and this has solved my grounding problems. Make sure the repeater is “isolating”. These can be purchased from most vendors (www.bb-elec.com, http://www.rs485.com, black box, etc..). You must also bias the RS-485 lines (a-b). Remember, only bias the line at one point on the network. The isolating repeaters have a bias option via dip switch (typically, 4.7k resistors to 5v and GND). B&B Electronics is a very good resource for RS-485 information “http://www.bb-elec.com/technical_library.asp”.
Also, you do not need to add terminating resistors if your baud rate is 9600 or lower (any reflections are dampened out before the center-bit time at 9600 baud). Termination resistors will do more harm than good if your baud rate is low.
Hi guys. This is a great discussion, and I hope one of you (or many!) might be able to help me with a problem I have with a 485 deployment.
I am trying to remote a single serial sensor (232) about 100 meters from a PC. For this purpose, I purchased two 232/485 converters. The PC’s serial port is connected to one, which is connected via 3 wires (A, B and Ground) to the other (and out comes the 232 to the sensor).
I am also passing power to the remote converter and sensor on the same cable that caries the signal (4 wires in total, 3 one 5v supply).
The problem is, that this setup “basically” works. Meaning, it can be up for an entire day, but at some time it will fail. Sometimes a simple power cycling (off/on) will restore communications, and sometimes it will not. Sometimes I have no idea why, it just comes back on! Then, often, it stays off for hours. I really don’t get it.
What might be wrong? This is the first time I am using 485.
Might it be a problem to pass power along with the signal? Maybe the fact that the ground of the power is common to the ground of the signal? (but anyway they are shorted in the converter box so I don’t see how it might mater).
Help!
Thanks guys.
PS – Ah yes, the in-built terminators on the converters are enabled.
PPS – I have not connected the shielding of the 485 cable to ground
PPPS – I am using regular 4 core shielded wire, not twisted…
Wow, a lot of information on RS-485.
Still confused as to what to do -
Do I tie the Chassis of the Transmitter device’s ground terminal specifically for RS-485 to BOTH the 12V power supply negative or ground, and the Shield Terminal located on each of the the 485 Devices? Each device can be connected in Star/Home run or daisy chain with termination at the end of each run (1 for Daisy, 2 for 2 device star, not sure about 3 in star connect…)
What is a foolproof way to ensure there are no issues?
I heard a rumor that if this is done that it pulls the floating RS-485 signal to a common ground.
Exceptions??: The ground’s of each the send device can’t always be tied together, as they may be on separate electrical systems entirely. But in this system, each recieve device is usually tied to the send device so this isn’t as much of an issue, they are treated like 2 separate systems. However, When they are connected as in a network A and network B, it is best to have a autmoatic switch to toggle which device they are connected to at any given time, and the switch should be grounded how? TO both systems?. To one system? Still confused…
Hi friends, I want to connect the two wire R485 to the computer RS232. Which Pins should I use?
Thanks
Considering that most user’s end up with shielded wires which only have a single twisted pair, what would be better:
1) Use the shield wire as the signal GND wire by connecting them to GND terminal on both ends?
2) Just connect the shield wire to ground at one end, and leave the GND terminals on both ends hanging?
485 is a two wire standard. Yes, you measure voltage between two points, but the error in all the comments about having to use three wire is that you assume you are measuring from a signal wire to a local ground. So your arguement is that local and remote grounds have to at the same potential. This is differential signaling, not relative to local or remote ground potentials. The differential voltage is measured from one signal wire to another signal wire, NOT from a signal wire to ground.
The problem here is that for the transmitter side of your device, you are creating voltage sources relative to your local ground potentials when you should be driving the lines by generating the potentials on one signal wire relative to the other signal wire.
You guys need to truly understand “differential signalling”. Heck, you guys will start requiring phone lines to use three wires instead of two. And instead of four pairs of wires in an ethernet cable, you’ll ask for extra ones for grounding.
And you don’t need shielding if you use twisted pair wire. The pair of wires for your telephone are unshielded twsited pair. Even ethernet cables are a collection of four unshielded twisted pairs. These all use differential signalling.
Hi guys,
Great discussion, just to get involved I would agree with Bock and Douglas, surely the whole prupose of differential signalling is to avoid having to reference the ground of the remote device? Wouldn’t joining the grounds together where a large common mode voltage exists cause you to push current between the devices along your RS485 wire?
Couldn’t you optoisolate the incoming data signals anyway?
cheers
ed
Hi again,
I had a bit more of a think about it and I think I can see why you would want the third wire in an optoisolated system where an isolated power supply at the far end creates a voltage relative to the signal ground to power the input opamps and the output of these is then optoisolated.
Surely there is a more elegant solution to this tho? Couldn’t the input signals driver the optoisolator directly? Do any of the guys who think you can do it with two wires have any examples?
cheers
ed
I had a think about the circuit thing myself and decided that if you use a completely isolated power supply (where GND and POWER are derived from a transformer) then you don’t need the ground wire. You just float the output of the transformer/rectifier to within the range of the incoming signal using a resistor to the signal wire. The input stage is then powered by this and the output of that is optoisolated and connected to the rest of the system.
Actually, Douglas is correct. I have some experience in long-distance, multi-drop RS485 among fire panels. Since these can be installed in different buildings – each at a slightly different ground potential – you don’t really want a ground wire. If you had one, you may end up with more current flowing through it than you’s like.
One way to solve the issue is to isolate the power supply to the transceiver and opto-isolate the DI and DO between the transceiver and micro. The data bus then floats, and the devices don’t care.
If you have full control over the power that drives the 485 circuitry in each device on the network – for example by feeding them all from the same supply then your don’t need the 3rd wire. Isolating all the power supply will work too BUT do you have control over how each device powers it 485 circuitry
I have to disagree with Dan. Building to building with ground potentials is the most common reason 485 networks don’t work without the 3rd wire.
i read your article on RS484 and the third wire. though i may be in error, i believe that your reasoning for the third wire is incorrect, RS485 trancievers use a differential amplifier, glorified opamps, which only measures the voltages of the A and B lines with respect to each other. because of that, as long as the voltage between A and B do not exceed the RS485 spec for differential voltage, the only voltages seen on the device is that between the two lines. by introducing a ground reference, in some cases, may actually damage the circuit as the voltage seen at the receiver could now be hundreds of volts above the ground plane of your reciever. voltage is a potential, not an absolute quantity, it can never be measured from one point and never from two points that are not in some way electrically conected. for example, try measuring the voltage between two unconected 9 V batteries. if you took a voltmeter on the positive terminal of battery 1 and the negative terminal of battery 2 the voltmeter will read 0V. the same principle applies to the transmitter and receiver of a RS485 circuit. attaching the positive lead of a DMM to the ground plane of the transmitter and the negative lead to the ground plane of the receiver will yield a voltage of 0V.
I have to respectfully disagree with Jeremy L. As he said – voltage is relative. The difference between 30,000 volts and 30,001 volts in only 1 volt. So according to your argument the remote rs485 device would not even know about the 30kvolts because it is only interested in the 1 volt delta. Not so. I have chosen an extreme example but it’s pretty obvious it would not work. Why ? Because the voltage differential sensing circuit is designed to operate between -7 and +12 volts DC.
What blows 485 ports? According to your argument it would be an oversize voltage differential. According to my argument it would be a line at a voltage outside the designed operating voltage range. If your argument is correct then the point of adding isolation is what ? to limit the differential ?
“They are -7Volts to +12Volts. What happens if you have two devices and a ground potential exists between the two devices of 24 volts ? You can see that one of the devices will be operating outside the specified voltage range.”
This is a problem if the transceivers are not allowed to float — that is, if their power supply or the logic lines operate relative to their local ground, and the local grounds of the different end points have a potential difference.
The right way to do this is to have the RS-485 interface ‘float’ away from the local ground. One way this can be done by making the entire device float (battery powered, or powered through a transformer and not having a chassis ground connection that connects to earth ground).
The other way is to build the RS-485 section of the device have its own independent floating power supply, and connect to external logic via isolators (optoisolators, normally).
The key thing to remember is that the sensing circuit operates between -7 and +12 volts, relative to the circuit’s “0 volt” point. That 0 volt point isn’t necessarily chassis/earth ground. It’s a virtual ground.
Specifically, Peter C asks — what if there’s a 30kV difference between transceivers… Well, I certainly wouldn’t want to ground-equalize two buildings with a 30kV potential using a single itty-bitty grounding conductor! Moreover, if such a potential were to exist, it’s probably there for a reason, and so you don’t want to get rid of that!
The point of adding isolation is so that the RS-485 transceiver’s virtual ground can float up/down as needed to meet the RS-485 network.
Gents,
I have read your articles with interest as I am about to impliment a 485 network for access control at a retirement home, which I have not done before. I will have one controller with about 15 readers, in a point to point configuration. The distance between the controller and the furthest reader will be about 200m. Each reader will have an independent 12V DC (battery with charger) power supply. I have on hand, some 6 pair (twisted, unshielded) cable which I was hoping to use. One pair for the access control, and the others for intercoms, and a gate release trigger.
Is this feasible? Would you foresee any problems?
Your comments and advice would be greatly appreciated.
Thanks
Allan
Just to add my two cents,
The UK Highways Agency uses RS485 as the primary communications interface between roadside signal devices and their controllers. These use a two-wire RS485 interface with opto-isolation of up to 2kV from local ground and an additional lightning protection circuit to clamp high-voltage transients and large differential voltages.
This system has been in place for at least ten years (I’m not sure of the exact length of time as I’ve not been in the industry that long) and has never required or operated with a third wire for grounding. The trick here is that a differential bias is placed on the line at the controlling end, and absolutely every device on the line has the RS485 signal circuitry isolated from the local ground.
As a number of people here have pointed out, this allows the RS485 circuitry to operate entirely differentially, as the the 0 volt reference specified in EIA-485 now becomes a “virtual” 0 volts. In addition it’s known that the “absolute” voltage from the local ground on these lines can be up to +/- 120V in places. However with the opto-isolation and lightning protection circuits that are specified the devices can operate perfectly under these conditions.
I hope this helps to show that RS485 can be operated as a purely 2-wire protocol without undue destruction of devices, if the proper isolation is designed into the circuit.
That a great insight. Thanks. In a way your point adds weight to the need for the 3rd wire. I say this because, you comments demonstrate the kind of care that must be taken to avoid issues addressed by the 3rd wire.
Wow – very strong differing opinions here. So what method is correct? It seems to me that it mostly depends on the control you have over all the devices connected in the system.
If EVERY device in the system is designed with proper isolation (i.e. the UK Highway Agency as described above), then the 2-wire scheme should work.
However, if some of the devices in the system are isolated and others are not, it seems as though there could still be issues. So what is the bullet-proof implementation what works no matter what type of device (isolated vs non-isolated) is connected to the system?
Nice discussion, but no one seem to address on of the question and I am aslo curios. Can the shield wire of the STP to be used as the third wire?
Hi,RS485.
I think you can not ude shield as common wire because the shield should be grounded at one end only, so you will have a potential difference at every device of oyur network.
If you ground the shield at both ends, you will have current flow through it, which is not good.
Hi everybody.
I have a problem here: 3 temperature controllers (RS485), a RS232 to RS485 converter and a PC.
The systems seems to be ok, at least the first week after finishing cabling and parameters setup.
Now, sometimes the controllers change from PROGRAM to RESET mode automatically, and nobody understand why.
I guess it could be a RS485 net problem.
The details are:
-The is no ground connection because the RS485-232 converter doesn’t have a ground terminal. But controllers do have.
-The cable used is a profibus (shielded, 2 wires, not twisted)
-The converter has a built in terminator resistor, and it is enable.
-The are no bias nor terminator resistors in the net.
Can somebody help?
Do you need any extra data?
Best regards.
Dear All, I have read with a lot of interest the whole discussion. I am currently finishing the development of the interface board for a transformerless solar inverter. Since due to its topology the inverter’s internal GND will be at -400V with respect to protection earth, isolation transformer and optoisolators are a must in the design to achieve SELV (Safety extra low voltage) conditions with 3750 Vrms isolation. I have employed shielded optoisolators to avoid problems due to the very high dV/dt which is issued when the GND of the inverter and PV generator assumes its potential with respect to protection earth due to the grid connection, which in the networks this product will operate is TT (neutral tied to ground in the transformer station).
Coming to my issue: I had to give up the possibility to have a terminal block for the RS485 connection to comply with the mechanical design of the product so, I am using two RJ45 shielded connectors to connect the RS485 bus (one shield has local PE connection, the other doesn’t to reinforce the shield grounding at every node). This, at least, have an advantage: it gives me 8 wires available and always connected by the end user, so I can tie all of the floatings SELV GND to the same reference. Due to another mechanical problem (cable glands to achieve an IP65 product), I have been asked about the possibility to use unshielded cables, standard ethernet cables, in place of STP cables because the unshielded connector is slightly smaller than the shielded connector and can pass through an M20 cable gland, while the shielded one can’t.
So, my question:
do you think I could use unshielded cables, since I have carried around the SELV grounds and tied them all together?
As far as I know, the shield protects the wires from electric fields, which produce common mode noise, but in my case this noise will also move the SELV GND, so the transceivers wouldn’t see it.
Always af far as I know, operating differential as RS485 does is a further protection about common mode noise.
The twisting is supposed to protect against magnetic fields, which produce differential voltages on the line and can’t be blocked by any shield.
The typical cable lengths for this kind of bus connection are up to 1Km, the typical baudrate is 9600 bps, the typical environment is industrial or rural.
Today I’ve been testing an unshielded twisted cable twisted around the DC input cable of the solar inverter, which was temorary fed by a 1kW flyback converter (which is something like an EMI noise generator through it’s air gap). I was seeing some 25 ~ 100 ns spurious signals on the MAX3080 output, but the communication was fine… however the distance was very short.
By the way, about tests, I have been trying also SN65HVD22 and in my application it is unusable: it’s extremely sensitive, thus collecting all the EMI noise coming from the inverter itself and the surrounding environment.
About biasing, since in my application it can’t be known in advance how many devices will operate on the same bus, it is not possible to design in advance the biasing network, so I have decided to use idle-failsafe devices, which output an high level if bus is idle.
Hello, today I’ve been making some tests in my laboratory in order to better understand what is the best way to get rid of communication errors on RS485 for my application.
In order to perform my test I’ve set up
- a computer with a non isolated USB to 485 converter, which GND is tied to protection earth by design
- one of my boards which is fed through an isolated AC adapter and employing the ultra sensitive SN65HVD22 transceiver (terminated)
- two twisted pair shielded data cable
- 1kW flyback converter as source of noise
- 1 solar inverter as load for the flyback, part of the noise generation and slave device to check the communication.
During the tests I’ve been observing the output pin of SN65HVD22 while there was no activity on the bus and I’ve been performing my experiment playing on three variables:
- bus bias
- grounding of the floating GND
- shielding
The shield and one of the four cables inside the data cable were permanently tied to ground on the PC side.
Here follow my results:
1: No ground brought to the board, no biasing: the transceiver was picking up all the noise possible and getting locked due to its hysteresis.
2: Ground brought to the board through wire, no biasing: the transceiver was picking up a smaller amount of noise, but the switching frequency was visible at the output.
3: Ground brought to the board through cable shield, no biasing: no noise at all.
4: No ground brought to the board, bus biased on the terminator so that the thevenin resistance seen by bus is 120Ohm and standby voltage on bus is 1V: the board was picking up an amount of noise comparable with case 2
5: as case 4, but grounded through wire: no noise at all
6: as case 4, but grounded through shield: no noise at all
7: shield disconnected from the PC and tied only to the floating GND: no noise at all.
8: no bias, no terminator on the board, grounded through shield: no noise at all.
So, in conclusion I can say the best way to achieve a stable RS485 communication is: to only ground the master, use optoisolated devices for the rest of bus, bring the master GND around through the cable shield, bias the bus on terminator only.
Of course no other connections to PE should be done on the shield at any point.
If unshielded cables must be used, it is mandatory to bias the bus and bring around the ground on a wire.
I think there is difference between shield and cable ground because, while all of them are subject to the same noise inducing electrical field, in strong electric field, the cable distance from each other matters in the magnitude of the induced voltage.
While the shield surrounds the cables and ideally is a Faraday cage, which forces the inner potential to zero, so even if the ground is far away and the potential is not anymore ground in a certain portion of the cable, at least it is forced to be equal on all the conductors.
My 2 cents.
Tomorrow I’ll try what happens with another source of noise, producing massive dV/dt.
Errata corrige:
I’ve just reviewed my report and reusults of experiment 1 and 2 are swapped: in case 1 I had distubs lasting 1us and with 52kHz frequency, in case 2 I had a 50% square wave with a frequency of 55kHz.
Case 4 had a disturb lasting for 60ns, always at 52kHz
Test results: also dV/dt disturb effects can be eliminated by connectig the shield to the floating GND of the slave device and using it to carry the master GND. In this caase also bus polarization alone solves the problem.
Acrually it is possible to avoid using 3-rd wire.
first, both wire must be protected against transients with diode clamps in both directions, VCC and ground. And then they should be forced to remain in the required point (half of VCC) with resitor voltage dividers. These will also work as terminators to avoid reflections.
unless noise is extremely high it should work fine.
if full isolation is required 1:1 transformer can be used for that purpose.
[...] No you can not use a T connection, MSTP requires both the positive and negative wires. Note: MS/TP actually requires 3 wires. #call_to_action {padding: 0px 5px; margin: 10px 0px 10px 20px; background-color: #E0E3EF; border: [...]
Hello tadas. I see two problems in what you propose:
this kind of bus with voltage dividers is fine for just a point to point RS485 connection because there must be just two terminators in the bus, but what you propose should be done in every node.
The transformer idea can’t work because RS485 isn’t a zero average voltage standard: no one warrants that the total duration of zeros is equal to the total duration of ones, leading to a DC voltage component, which in turn saturates any transformer. Even if the transformer is not saturated because of the line impedance limiting the current which can flow through it, in case several zeroes or ones form a packet, they would be lost because the transformer works on voltage variations.
[...] Chipkin Automation Systems – RS485 Cables – Why you need 3 wires for 2 (two) wire RS485 [...]
We are using rs 485 network for our Energy meter commnunication.we are facing the some problem this network having totally 24 Nos enegy meters.
All are Rs 485 2 wire,Modbus,9600,shiled twised pair cable out of 24 Energy meter the following meter are not communicating Meter ID 11,14,20&23.
We are using Terminating resistor also (120ohm 0.25W) for both ends.We are using Rs 485 to Rs 232 Converter.
I am check individually for Energy meter ID 11,14,20 & 23 separately cable that only one meter that time these meter are working.
Please help me.
Thanks
Sri.V
can i shied both ends of control wire.
Hi everyone !
In our power plant we are having problems to establish Modbus RS485 connection between Microscada and Beckwith M-3425 generator protection relay.We use CAT5 cable and in beckwith module i have inserted 200ohm termination resistor.Does anyone done this job before?
We are service provider for metone Air born particle counters.
i have one problem at site.
the counters are connected through rs485 communication cable belden 9841.
it is observed that counter shows counts even if there is no particle present at that time & this is for only 1min duration which is sampling cycle of counter.
i have doubt that this may be due to spike in communication cable.
is there any way to moniter the spike OR remove the spike from comunication cable.
Gentlemen, there is alot of confusion here on RS485 and adding ground references into the mix only confuses the issue. RS485 is a bidirectional, half-duplex, communication bus structure consisting of a single Master and at least one Slave. The maximum number of slaves can vary widely from system to system however; most manufacturers cap the max number of slaves between 16 and 32. Most RS485 signals operate ideally on a DC bias of 5 volts. The signals are driven alternately from each other. That is, each line operates inversely from the other and each is also referenced to the other from an electrical standpoint. The receiver looks at the difference, not the absolute voltage value, of the two signals. This is referred to as the “Line Bias” and it is critical in RS-485 applications. A bias difference of higher than 0.3v is generally accepted as valid but can be as high as 0.7v, depending on the system. Absolute values below this are considered “undefined” or “gray” and may result in either a high or low reading by the receiver. In many non-isolated applications, a “ground” or “shield” connection is provided in addition to the 2 data lines however, this is not necessary as the signals are referenced to each other and not to absolute ground. It is only provided as a ground point for the communication cable shielding. It is important to note that the shield ground should only be connected at a single trunk end in non-isolated applications. If grounds are tied together in non-isolated systems, even a slight voltage difference, between the absolute ground points, will create a “ground loop” condition which can cause serious damage to the equipment. Note that in isolated equipment, the above does not apply as there is no direct connection between the ground points however, in all applications, proper cable shield grounding should be practiced to eliminate and/or reduce electrical interference.
All devices, including wire, have a resistance to electricity. Although resistance changes slightly with temperature and age, it remains the same at all times, no matter what electricity is doing. As stated earlier, RS485 communication utilizes fluctuating signaling across a pair of conductors which creates an RF frequency. The baud rate determines the RF frequency at which the signals operate and the lower the frequency, the higher the impedance. Impedance is a measure of the overall opposition to current flow and only exists where there is AC, or fluctuating DC signals. In other words, Impedance is the measure of how much the circuit impedes the flow of alternating current. It is like “ohmite” resistance, but it also takes into account an additional property called Reactance. Reactance is the opposition to alternating current due to either inductance (inductive reactance) or capacitance (capacitive reactance).
Inductive and Capacitive reactance are exactly opposite in their effects in AC circuity and will cancel each other respectively. Inductive Reactance is the electrical resistance caused by the build-up of electric or magnetic fields and increases as the frequency gets HIGHER. Capacitance Reactance is the electrical resistance to a signal’s ability to charge or discharge to a voltage state and increases as the frequency gets LOWER. Thus, if a circuit contains 150 ohms of inductive reactance and 125 ohms of capacitive reactance in series, the net reactance, is 150 – 125, or 25 ohms of inductive reactance.
The key truth of all communication lines is that the source impedance must ideally be equal to the load impedance in order to achieve maximum power (signaling) transfer and minimum signal reflection at the destination. Serial communication operates most efficientially when the source and load impedance are matched to 120 ohms. Unfortunately, when higher baud rates and/or longer transmission lines are implemented, the effective load impedance tends to change causing a mismatch. Keep in mind that the ideal impedance of 120 ohms only describes the intrinsic impedance of the transmission line and it is not a function of the line length. As a result, the transmitted signal will not be completely absorbed by the load and a portion is reflected back into the transmission line, causing a condition called “cross-talk”. As the source and load impedance become more equal these reflections are eliminated.
There are several different ways to counter cross-talk, however, the most practical solution is to create an additional load to absorb the excessive reflections. This can be done by simply placing a “termination resistor” in parallel, with the communication medium. In RS485, a termination resistor may be added across the receiver’s “+” and “-” lines in order to bring the load impedance back into compliance. Keep in mind that resistor termination can potentially add heavy DC loading to a system and overload the source. Therefore it is best to confer with the hardware manufacturer before adding any line termination. That being said, a terminating resistor of less than 90 ohms should not be used and termination resistors should only be placed at the trunk ends of the transmission line. No more than two terminations should be placed in any system that does not use repeaters.
Many manufacturers suggest utilizing a termination resistor valued between 100 and 150 ohms; however, they fail to mention that if proper communication cabling is utilized and the baud rate is not excessively high, a load termination resistor may not be needed at all in shorter cable runs. Remember, if improper communication cabling is utilized, all rules are thrown out and it is only a “hit-n-miss” if the application works at all. Belden 9841 or 9842 is EIA-485 cabling is made specifically for paired, half-duplex, multi-point RS485 serial communication applications
[...] other times) and finally you risk re-installation. For a more detailed discussion read this article http://www.chipkin.com/articles/rs485-cables-why-you-need-3-wires-for-2-two-wire-rs485. The more power sources used to power devices, the greater the physical separation of devices, the [...]
Y’all need to learn what “differential signalling” means.
Common ground references indeed…!