Sign in to follow this  
Followers 0
CanaanP

Wireless Analog I/O?

6 posts in this topic

Greetings, I do not have a specific project in mind but was just wondering what the experts use whenever a wireless solution is required for analog. I am familiar with setting up wireless options for digital i/o, however all of the analog devices I have seen only do one or at most two channels. Does anyone know of a device that can handle at least 4 separate channels of analog for 0-10 or 4-20? In addition, I would be interested in something similar but for thermocouples. Any recommendations would be appreciated, thanks! Also, I have been reading some details of the flex i/o modules from AB. On the Analog, Thermocouple, etc tab on this page, http://www.ab.com/en/epub/catalogs/12762/2...68992/tab7.html it will say something like, eight single-ended inputs. What does single-ended mean? Others will say isolated inputs, not sure what exactly the distinction is. We use three conductor shielded wire for analog, for use with potentiometers and actuator position feedback. So which types of inputs should I be looking for in regards to our purposes? Thanks in advance for the info :) Edited by CanaanP

Share this post


Link to post
Share on other sites
It depends! First off, voltage-based signals have an advantage that there is no capacitance involved so if you need very high speeds such as with motion control (faster than 20,000 samples per second), current loops won't do it. For all other applications, use a current loop. Why? Cable impedance doesn't matter. So you don't have to recalibrate every time you do maintenance on the system. Also since it is off the noise floor (starts at 4 mA), you don't have noisy signals down close to "0". Shielding is nice but essentially all but unnecessary since the current loop naturally has a low impedance and doesn't couple noise signals to any great degree either. Again, something voltage loops which have high impedance followers have trouble with. Current loops can also easily go thousands of feet with no significant signal degradation (no voltage drop) since they are effectively "self-calibrating". A voltage loop is also limited to a single receiver where you can have multiple transmitters with current loops. In fact you can also carry digital signals including calibration over analog loops using the HART protocol. And HART-based loops can even have multiple redundant transmitters. Another disadvantage of voltage loops has to do with voltage drop. Since voltage drop is a fact of life with voltage loops, the receivers are always high impedance designs to mnimize this. The result is that this also makes the whole system highly susceptible to noise since induced voltages (and currents) are similar in magnitude to the intended signals, whereas with current loops, most induced noise signals have a much smaller magnitude since it is difficult to achieve inductive pickup in terms of current onto a system which has no direct connection to the noise source. Another huge advantage of current loops in particular is a two-wire system. 4 mA is usually plenty to power sensors so this allows you to supply power directly through the sensor cabling rather than running a 4-wire system to supply power separately from the signal wiring. Whether voltage or current, the heart of an analog card is always an analog-to-digital converter. Virtually all of them (with rare exceptions) are voltage based. The input card uses a precision resistor to convert the 4-20mA input signal back to a voltage (V=IR) if you aren't using a voltage-based system. The analog-to-digital converter is of course referenced back to a common reference (0 V) on the card and has a fixed input range (usually 0-10 VDC). Most cards multiplex the input signals onto a single ADC converter since the converter is far more expensive than a multiplexer circuit (you can buy ADC converter chips with the multiplexer on board). That leaves just one problem. All of your signals will be referenced to the same 0 VDC common. That's OK if they all have the same power supply, there's no ground loops present, and you don't have multiple receivers in a 4-20mA current loop. This is where single ended cards work fine. Otherwise, watch out because that common reference voltage on the negative (return) side of your card is no longer a good reference. In some cases such as loop-powered devices or isolated transmitters, you can short the commons (returns) together to force them to have a common reference. But if that's not the case, then you need isolation...to allow the common to be different for each channel. You can buy an input card that has this, or you can buy them as external modules such as from FACT Engineering (available through Automation Direct) or Action IO. Usually the most common (and best) analog isolators are optically isolated. This means that it has an LED coupled to a phototransistor, making the link between the circuits 100% optical and providing almost infinite isolation. This isolation problem is true whether it's a current loop system or a voltage system. It's just a more common problem when 0-10 VDC signals are present, especially as these are prone to nearly every form of noise available with the high input impedances present to try to minimize voltage droop. With a little planning though most of the time I can get away with perhaps 5-10% or fewer loop isolators relative to the number of non-isolated loops in a single ended 4-20mA current system, and most of those are 4-wire devices instead of the preferred 2-wire ones (no power cabling to run).

Share this post


Link to post
Share on other sites
Paul, as usual, I only understand a percentage of what you're saying. Just to clarify... You are saying that single ended inputs are for voltage loops? Isolated inputs should be used for current loops, but be mindful of using multiple commons unless bonded? We use the shielded cables because they run very close in parallel to 3 phase 480v. There is a divider in the cable tray to keep power and control separate, but still needs to be shielded. As I stated before, the shielded cables are for potentiometers, used as a speed control, or actuator positioning. We have noticed adverse effects of instruments when not bonding the shield properly. We typically use Red Lion controllers to monitor a temperature and adjust a damper accordingly by controlling an actuator. I have no doubt that some sensors can be powered through a current loop as you said, however we don't do that. All limits and such are powered by a separate 3 conductor + ground cable. One conductor for 110v, another for neutral, and a third for a return hot when the limit is made. So, the analog reads a pot inside the actuator to show what percentage it is extended on a small display in the control house. A 3 conductor cable is also run out to the actuator ( in a junction box) and two conductors are 110v, one for extend and the other for retract. The controller fires either one to obtain the desired position of the actuator which is controlling a damper. The controller watches a temp coming from a thermocouple from another location. I should have stated that we have been doing this successfully for some time with no problems, but doing so by running the cables full length. The control wires for extend and retract still have to be run all the way out, but was just kicking around the idea of having the analog run over a wireless, then come into the plc, which we have not been doing.

Share this post


Link to post
Share on other sites
You are supposed to keep power and signal cables separated but obviously that's the ideal situation, not always achievable in all situations in practice. However, if you converted to current loops, you'd find all your noise problems that you experience now will disappear entirely. It would still be good practice to run shielded cable but the system will run fine even if the shielding gets messed up.

Share this post


Link to post
Share on other sites
Here's some pictures to look at if you still are having trouble. http://icb.olin.edu/fall_02/ec/difsingle.shtml

Share this post


Link to post
Share on other sites
Here's some pictures to look at if you still are having trouble. http://icb.olin.edu/fall_02/ec/difsingle.shtml And more stuff to read. http://www.microlink.co.uk/differential.html Edited by Mickey

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0