DanW

Testing with a manually controlled signal generator is not a good test for PID because your signal generator is not responding like the process would respond to the PID output changes. If the output is wired to the I/Ps and the I/P's are connected to the damper and valve, then a signal generator can prove the control direction is correct with rise or fall of the input. It would not surprise me if "PID has locked up" means that the output is stuck at either 100% output or 0% output when the input does not reflect the output change.

Here's what the manual says about HART current ratio (page 62) https://www.quicktimeonline.com/assets/images/pdf/Allen%20Bradley/1794-IF8IH%20User%20Manual%20FLEX%20IO%20Isolated%20Input%20Output%20HART%20Analog%20Modules.pdf There is a current fault/fail-safe limit on most field devices that drives the analog signal upscale beyond the overrange 20.5mA level (like 21.4mA0 or downscale below the 3.7mA underrange level (like 3.3mA).   But the fail-safe limit happens when the transmitter determines that its internal functioning can not report a valid process variable.  It's telling the controller, "my process variable is not trustworthy".   Which brings up the point, what does the HART variable do when the transmitter is in fail-safe fault condition?  If the HART variable goes off-scale, too, of then there's no deviation between the fail-safe analog value and fail-safe HART value. It might be for dealing with a situation where the 4-20mA analog output is a subset of the full range of the transmitter.   For instance, pressure transmitters can be capable of measuring 0-1 bar but can be ranged 0-0.25 bar.  I've never checked for this but it could be that the instrument senses and can report a 0.40 bar signal that would saturate the 4-20mA signal at 20.5mA or thereabouts while the HART signal could conceivably report a reliable 0.40 bar value.  The current ratio tells you that the HART signal is valid but the analog is not ? ? ?   

I've been mulling your question for a couple weeks now because I'd never heard of current ratio in relation to HART. If I interpret your conclusion correctly, the HART digital variable is compared to the analog value and if the analog value wanders or drifts beyond the percentage limit, then current ratio alarm is triggered. The analysis function would have to be smart enough to know what the 4-20mA signal represents, in eng. units, in order to make a valid evaluation and the card is probably smart enough to know that.   But what if the HART variable is the secondary or tertiary variable, not the primary variable?  Then there's no direct correlation between the analog value and the HART value. And, what's the purpose?   Detecting drift between the digital and analog variable?    When I started this field in the late 1970's, the analog (non-smart, zero and span pot) pressure transmitters would drift several percent in 6 months, necessitating periodic re-calibration.  But the electronics have improved so much over 40 years that today's' 4-20mA's don't drift much, if at all;  they're remarkably stable, and they don't typically drift to the over 5% range.   There is a possibility of ground loops occurring and although that's usually detected during installation, events like the moisture intrusion into junction boxes over time can create a ground loop or increase the magnitude of a ground loop.  So analysis could detect a ground loop affected current signal that doesn't match the HART variable.  So an analysis might catch a ground loop error. But I suspect that the current ratio refers to signal strength, noise or signal to noise ratio of the HART signal. A web tutorial/primer on dB (deciBel) states: Typically the deciBel, dB is used for defining amplifier gains, component losses (e.g. attenuators, feeders, mixers, etc), as well as a host of other measurements such as noise figure, signal to noise ratio, and many others. https://www.electronics-notes.com/articles/basic_concepts/decibel/basics-tutorial-formula-equation.php and the primer continues on to mention current ratios. So I suspect that the current ratio is a means of determining the signal strength of the HART signal, an FSK voltage superimposed on the mA current signal. But how that works out to percent and what the practical real world implication is, is beyond me.

The panel under consideration has a compressed air supply of 120 PSIG which runs through some Parker N 1/2" OD nylon tubing rated to 250 PSIG.  Parker Prestolok Push-to-connect tube fittings are used throughout.     Should one of the tubes fail or the seal on a push-to-connect fitting fails, the supply air would leak out and fill the panel.  I assume that that door gasket would leak at some point, probably leaking at the same rate as the 'fault' or the door gasket would tear, opening a gap that exhausts/vents the pressurized air from the enclosure.  But this is speculative on my part. Is it a standard practice to install a relief valve to open and vent compressed air in the event of a air component failure inside the panel?

There are stand-alone single circuit calendar timers that operate on sunset/sunrise that are intended for lighting purposes.  The one I saw fit in electrical wall box.  I saw the electrician messing with the configuration.  It came in a retail box with printing on it.  Sorry, I didn't catch a name, brand or model and since I didn't buy it, I don't its cost.  But maybe the cost of loal timer control would be offset by the extensive wiring needed to do central control. 

short answer, from a Honeywell authorized distributor. For years, the HC-900 development software was called Hybrid Control Designer (HCD) but is now called ControlEdge HC900 Designer (not to be confused with ControlEdge PLC Designer, an altogether different platform).  The current Version 7 is backwards compatible through all previous versions,  Be aware that there is a 'utilities' version that only allows setpoint programming batch edits; other logic can not be edited.   There is separate software for Honeywell CS900/CR900 branded HMI panels called Station Designer. Both are licensed products, available only through Honeywell authorized distributors. Also be aware that an HC-900 CPU has optional password security that if enabled, must be dealt with. Updates within a version are still free downloads from the Honeywell public web site, but will not install unless a licensed version is detected.   (software tab at this link). https://www.honeywellprocess.com/en-US/explore/products/control-monitoring-and-safety-systems/Scalable Control Solutions/ControlEdge HC900 Process and Safety System/Pages/ControlEdge-HC900.aspx All the manuals are downloadable from the documentation tab at the link above. Delivery of licensed software is either an electronic download or a DVD.  The DVD costs more. Honeywell contact information for the Americas, who can presumably put you in touch with a distributor: Honeywell Process Solutions (Sales) 1-800-343-0228   Email: (Sales)    FP-Sales-Apps@Honeywell.com

Industrial communications involves - a hardware bus, - a protocol (set of rules for communication) and - software that can handle the data at either or both ends. A primary requirement is to determine what you are trying to accomplish with the communications. - program the logic? - get process data from the PLC into the PC?  Log the data?  Display the data? - send recipes from the PC to the PLC? - have the PC act as an HMI for the PLC? - get process data from the PC into the PLC? - something else Hardware bus: Both devices have to support the hardware bus or you'll need a converter, like RS-232 to RS-485 or RS-485/Ethernet Modbus generally does not run on USB.  That's not to say that it can't but that would be more of a special case than a standard situation. Protocol: This is biggie because the PLC states that it runs Modbus or some proprietary protocol.  Which one is the best choice depends upon what the proprietary protocol is designed to do (and with what software on what platform) and how that matches up to what you are trying to accomplish. Software: Do you have softare that does what you are trying to accomplish?  What hardware and protocols does it use?

I'm a Modbus guy, I've never done Ethernet/IP, but I have had a minor run-in with the UOC.   The UOC is a Honeywell ControlEdge 900 controller running the Honeywell control execution environment (CEE) configured with Experion Control Builder, the UOC development software, which is a licensed software product.  Honeywell calls its versions 'releases'.   I do not know if a later release is backwards compatible with an earlier release.   Make sure you have access to the development software.  It's pricey. A Google search will locate the 50 page manual for the ControlEdge family of CPUs (of which UOC is one), EtherNet/IP User's Guide, document # RTDOC-X548-en-160A, dated from Jan 2020, link below: https://www.honeywellprocess.com/library/support/Public/Documents/ControlEdge-PLC-EtherNet-IP-Users-Guide-RTDOC-X548-en-160A.pdf The Function Block section of the manual with the Ethernet/IP FB's is attached. It isn't clear to me how two 'controllers' talk to one another, as opposed to a controller talking to a field device, like a flow meter.   With two controllers, who is consumer and who is producer?   Is that where an 'adapter' is used?   The Honeywell Ethernet/IP manual might shed some light on the 'adapter' situation.    Enet-IP FB's from ControlEdge-Builder-Function-and-FB-Configuration-Reference-Guide-RTDOC-X286-en-151A.pdf

My opinion is that it is it best to power the sensors from the monitoring station end using a single large DC power supply.   My experience is that using the power from a remote device to power local sensors is very likely to produce ground loops, an electrical artifact due to the fact that earth ground potential between different areas is not the same and current flows when connected between points at different potentials.   And devices eventually reference earth ground.    When a central power supply is used with isolated field transmitters like analog 4-20mA 2-wire loop powered transmitters, ground loops are rare.    Powering device with the remote device power supply produces ground loop, common mode problems devices like, - unexplainable offsets that do not appear when a battery powered analog source is substituted for the field device, - analog signal levels that drive off-scale - formerly functional, working signals that either suddenly offset or drive offscale when an additional signal is added to a multi-input module, or - in extreme cases, analog input burn-out due to excessive common mode voltage.   The reason for the development of 2-wire loop powered devices was to - use a common DC power supply at the monitoring end where all the signals end up - use the 2 signal wires to provide operating DC power to the field sensors - eliminate ground loops     

If I understand the situation, the reference to the "Diagnostic Code" is on the HART Device Info tab of the HART module properties. Being a field device diagnostic code, not a HART/AI module/card diagnostic, its interpretation would come from information from field device documentation. There's probably a long detailed section in the Rosemount 87 flowmeter manual on HART, or even a separate manual.

You're not color blind.

Honeywell’s Hc900 has both Modbus Master/client and Modbus Slave/server functionality as standard features in its CPU.   Modbus is implemented with Function Blocks.  Tags are used to identify the relevant data.  There is default mapping of slave data to Modbus address registers or a custom mapping option. I’ve seen theads on internet fora like this one regarding Labview having Modbus, but I do not recall the details. So, yes, NI’s Labview should talk Modbus to the HC-900.

Yes, I believe that there is a gateway type product that might help you out.  It can take the inputs it gets from its from its slave side from the existing master and map these so that the master side writes as it needs to, to accomodate changes in the former slave architecture.  The slave side can pretend that a single port is mulitple fixed slave ID node addresses, on one single port.  Try the email jschulze at iccdesigns dot com and ask him about the ETH-1000 (or whatever product) for your sitation.  He's the resident expert on control.com for Modbus situations like this and is product specialist at ICC Designs.  You can read a thread on PLCs.net where a somewhat similar situation existed.   http://www.plctalk.net/qanda/showthread.php?t=128109&highlight=intelligent+modbus+gateway  

I have used instrumentation that has had an internal I/S component, like a radar level unit with an I/S connection to the radar antenna, where the unit was rated EX and that was required by the agency(s) doing the approvals. I suppose you could put an IS barrier in the safe area for connection to the I/O points in the EX enclosure to prevent a fault in the safe area from providing an ignition source in the EX panel, but to my knowledge it is not required and I don't recall panels I've been in doing such.  People use I/S to avoid the EX enclosure, conduit, fittings, seals, etc.

What kind of switches?  Managed Ethernet switches?  electrical distribution switches? Are you logging the data read via Modbus?  Does the datalog show no data for when the page is closed and some data when the page is open?

Modbus is a project.  Here's some tips on the Omron end. 1. The spec sheet says that the Omron E5AC uses Modbus RTU over RS-485.  The Omron is a Modbus slave. Modbus/RS-485 is a communications option, make sure your device's model number decodes as having the Modbus option. The PLC will need an RS-485 port and whatever firmware it needs to run as a Modbus master. 2. The Omron user manual tells you where to - select the Modbus protocol (not the Omron CompoWay protocol) - configure the serial settings for RS-485 (baud rate, parity), which have to be the same on both ends. - the communications data length (data bits, presumably) defaults to 7, must be changed to 8 for Modbus RTU. - I recommend 1 stop bit because some devices do not handle two stop bits correctly. - configure the slave node ID number (Communications Unit No.) - wire the RS-485, terminals 13 B(+) and 14 A(-).  The Omron lacks a signal ground terminal which can create problems requiring an RS-485 isolator, but try it and see how is works.   Definitions of A/B or (+)/(-) vary from vendor to vendor.  It's supposed to be A to A, B to B, or (+) to (+), (-) to (-), but sometimes the definition is opposite and the A/B wiring needs to be swapped at one end. 3. There is a separate communications manual for Omron devices.  The comm manual will tell you which value is in which Modbus register. 4.  Data formats Omron calls a 16 bit Modbus register an "element".   Omron calls a one element, 16 bit integer "two byte mode" Omron calls a two element, 32 bit floating point values "four byte mode" Each value has both a 16 integer and a 32 bit floating point value, mapped to different locations, so you can pick which data format is easiest to deal with on the PLC The data is most likely in 'Holding Registers' that the Master uses Modbus Function Code (FC) 03 to read, or FC 06 or 16 to write values to. 5.  The Omron being the slave is essentially passive once you configure and wire it.  It waits for a command to provide data or change a data value when the command it receives is a Modbus write command. 6. If you're new to Modbus, it sometimes pays to get a generic Windows Master and a USB/RS-485 converter and experiment communicating with Omron with the Windows Modbus master because it is so quick to change settings (don't need to download to make a change).  Modscan, Modpoll or Simply Modbus are the Modbus Masters I frequently see people use. 7.  Get the Modbus comm manual and work on the PLC Modbus master end.

It is my impression from minor exposure 5 years ago that the development software version has to match the CPU firmware version.  The patch is not likely to load without the presence of the development software.   A document I found indicates that the current software version is R170. Honeywell ControlEdge Builder development software is licensed software, available only through authorized distributors or sales offices. Software purchase includes a software license with a multi-digit license code, that is necessary for downloading updates or for purchasing upgrades.  Be certain you get and save the license code. One can purchase the media kit-DVD version that includes the software on a DVD or media kit-ESD which provides a web link through which the software can be downloaded. The software license does not include documentation.   Documentation can be downloaded from the Documentation tab on this page: https://www.honeywellprocess.com/en-US/explore/products/control-monitoring-and-safety-systems/Scalable%20Control%20Solutions/Pages/ControlEdge-PLC.aspx I have no idea what export restrictions might apply to the availability of the software. You really need the advice and support of an authorized distributor to get the right combination of software/firmware to get a functional PLC.   

1. Schneider claims a non-re-wiring migration path to Foxboro EVO https://download.schneider-electric.com/files?p_enDocType=Brochure&p_File_Name=ID%2342_998-19865506_GMA-US_A4_WEB.pdf&p_Doc_Ref=PAS_63680_CPM16104&_ga=2.203063126.699307783.1532672213-458034261.1531877776 2. This Emerson RA Delta V blurb claims there is some means of 'saving existing wiring'. https://www.emerson.com/documents/automation/brochure-emerson-migration-solutions-for-moore-systems-pss-en-67768.pdf 3.  Honeywell  claims "a phased migration can also use existing wiring" for its PKS DCS.    My limited exposure to Honeywell is that the full blown PKS will be overkill for an APACS replacement.  The distributor supported ControlEdge UOC is probably more in line with APACS system performance, but I suspect that there are no wiring adapters. https://www.honeywellprocess.com/library/marketing/brochures/Competitive-Displacement-eBook.pdf 4.  Siemens hosts a User forum for the PCS7.  You might try an inquiry there for experiences with a migration from APACS.   Registration is required for the site is required. https://support.industry.siemens.com/tf/us/en/threads/135/?page=0&pageSize=10

Gotta admire performance for 40 years.   I, too, look forward to your reports on progress.

Noise. Almost every flowmeter has a setting called something like Low flow cutoff.   It is a signal level at which any flow below that level is reported as zero flow, primarily to cover this circumstance.  Find the low flow cutoff, figure out what units it uses, (% of span, counts, engineering units) and put in a number slightly higher than your existing low flow values. As a side note, unless the flow meter has a tight shutoff valve on both ends, there can be actual flow even though the pump is not running.  Bleed level flows happen as the sytem equilibrates after the pump turns off and gravity works on the remaining liquid in the pipes, sometimes even siphoning flow.

> i dont have any rs485 card or com port on my PC to directly connect to it. It pays to have both a USB/RS-232 and a USB/RS485 converters for a PC when you have to deal with industrial serial communications. Those with an FTDI chipset have served me well.  They are not very expensive.    I prefer flying leads (individual wire conductors) on the RS-485 converter because the connection is almost always screw terminals.

http://www.plctalk.net/qanda/showthread.php?t=127931&highlight=joystick

I'd suggest you look at the Blog at the link below, Understanding Control Signal Jargon, aimed at the control of electric actuators. https://blog.belimo.com/Blog/bid/47275/Understand-Control-Signal-Jargon I suspect that your actuator is in the group called "3-point, Tri-State, Floating Control" (to which I can add: position proportional, Three position step control, PAT, valve motor) You need to understand exactly which control mode your actuator requires.

The Moisture Meter supports only Function Code 03, Read Holding Register (this device is limited to one read or one write at a time) and Function Code 06, Write single value to a Holding register. From that one might deduce that K0 means that the register addresses use zero based addressing and that K5 and higher are read-only registers Meter  zero-based  one-based           Type K0      0000             (4)0001        R/W, FC03 or FC06 K1      0001             (4)0002        R/W, FC03 or FC06 K2      0002             (4)0003        R/W, FC03 or FC06 K3      0003             (4)0004        R/W, FC03 or FC06 K4      0004             (4)0005        R/W, FC03 or FC06 K5      0005             (4)0006        R, FC03 K6      0006             (4)0007        R, FC03 remainder are read only The data values are most likely 16 bit integers, where K5 is a signed integer with a range of -5 to +5.   The other registers could be signed integers, too, but they all appear to positive values. The K1 value seem to be truncated to the two high order decimal digits: 11 = 1100 25 = 2500 It's not my moisture meter so I only know what the sheet says,but K0 appears to be a control status word. the value 0 = stopped, the value 1 means it is in manual mode, value 2 means it is in auto mode.  This is most likely a read/write register so that one can remotely run the device in manual or auto mode or stop it by writing the appropriate value. Regarding K2 Moisture setting address K1, it appears that this device somehow changed moisture content, from the statement, "When the measured moisture vlue for three consecutive times is less than or equal to the set moisture value . ".  The 'set moisture value' appears to be a control setpoint, hence K1 register (4)0002, is the moisture setpoint value.  Again, not my box . . . K3 is the the type of grain that is being measured. K6 Actual Measured Address probably means the measured moisture value, residing at register address (4)0007   One uses zero or one based address values depending on what the master uses or requires.    

1. valve/actuator Realities a. A 3 way valve has 3 ports, either one inlet and two outlets or two inlets and one outlet. Why are you using a 3 way valve for steam?  How does that work?  Do you waste steam when the valve is in the 50% position? Normally a heat-only application uses a 2 way valve, one inlet, one outlet. b. On/off or modulating Depending on the actuator a valve can be controlled as either - open or closed, or on-off, or - modulating, where the actuator partially opens or closes the valve over the range of 0% (closed) or 100% (open). Questions: Is your valve/actuator that you are working with an on/off valve actuator or a modulating valve? What input signal does it expect? Is it driven by 4-20mA, one relay or two relays?   2. Control modes a. heat (reverse action) Heating action uses only one control output, either - a single relay or - a single 4-20mA, - the special application of driving a modulating electric actuator with two relays, where one relay drives the electric motor to a closed position, and the other relay drives the electric actuator to an open position. Honeywell calls this position proportional when there is slidewire feedback from the actuator or "Three position step control (TPSC)", an open loop control mode with no feedback for motor position. TPSC uses timing to estimate the motor position.   The controller's output for PID Heating action is 0% is closed, 100% is open.   b. Heat/cool control mode uses 2 control outputs.  One output is dedicated to the heating loop.  The other output is dedicated to the cooling loop.   When the PV (temperature) is below setpoint, the heating output is controlled and the cooling output is OFF, closed, at 0%. When the PV (Temperature) is above setpoint, the cooling output is controlled adn the heating output is OFF, closed, at 0%.  There is a setting, typically called deadband, which is a range around the setpoint where both heating and cooling are disabled (at 0%) because there is rarely an application where heating and cooling are needed to maintain setpoint.   3. All that said, what's going on? a. it isn't clear whether your valve is modulating or on-off: >"valve is 50 mm , step is 10 mm per minute"  An electric actuator takes time to close or open a valve but what is the intended design of your valve/actuator?  on/off or modulating? b.  I suspect that what you call a 3 way valve has an actuator that expects to be controlled with two relays, one to drive the motor open, one to drive the motor closed.  Is your term '3 way' related to the number of wires on the actuator or the number of ports on the valve?   c. >"when temp is going up  Delta DT320 is operating at the same time OUT1(because temp is lower then 13) and the same time OUT2 is try to close, This sounds like your are trying to drive an electric actuator with two relays, one relay to close the valve, the other relay to open the valve, which is the position proportional or 3 position step action described above. That's not the action that the relays are performing - the action that the relays are performing is heat/cool action: the heat relay is enabled (on a time proportional (a form of pulse width modulation) basis while the cooling relay is OFF. The control of two relays to drive an electric actuator is a special function.  Only one relay is enabled at a time.  The ON time must be calculated and controlled.  That motor control function is not mentioned in Delta DT3 instruction sheet, at least, I could not find it. Page 8 of the instruction sheet describes the control outputs and their allocation as Heat, Cool or Heat/Cool control outputs or as alarm outputs.  No mention of actuator control.