• Pages in Chapter 37

Chapter 37 - How to Disassemble a Sliding-stem Control Valve

Chapter 37 - How to Disassemble a Sliding-stem Control Valve

PDF Version

The following collection of photographs chronicles the complete disassembly of a Fisher E-body globe valve with pneumatic diaphragm actuator. This control valve design is quite mature, but nevertheless enjoys wide application in modern industrial settings.

An important safety note when disassembling pneumatic control valves is to first relieve all tension from the actuator spring so that its stored energy cannot harm you or anyone else. These springs may be quite large, exerting thousands of pounds of force during normal operation.

Spring tension may be relieved by moving the spring adjuster until it turns easily by hand without further aid of tools, or in the procedure shown in the following photographs by loosening the spanner nut attaching the actuator yoke to the valve bonnet.

This is the complete control valve, without a positioner attached. What you see here is the actuator (painted green) and the valve body (painted grey), mounted on a steel plate for student learning in a laboratory setting. The left-hand photograph shows the complete control valve assembly, while the right-hand photograph shows a student loosening the spanner nut holding the valve actuator yoke to the valve body:

The next step is to un-couple the actuator stem from the valve stem. On Fisher sliding-stem valves, this connection is made by a split block with threads matching those on each stem. Removing two bolts from the block allows it to be taken apart (left-hand photograph). Nuts threaded on to the valve stem, jammed up against the coupling block, must also be loosened before the stems may be uncoupled (right-hand photograph):

It is very important that no spring tension exists on the stem prior to disassembly of the stem coupler, or else the two stems will slip past each other with great force once the coupler is removed. Spring tension must be released, either by loosening the spring adjuster or by loosening the spanner nut holding the actuator yoke to the bonnet.

A close-up photograph of this stem connector block, with the front half removed for inspection, shows how it engages both threaded stems (valve and actuator) in a single nut-like assemblage. The solid valve stem (below) slides into the hollow actuator stem (above), while the split connector “nut” engages the threads of both, holding the two stems together so they move up and down as one piece:

Once the actuator and valve body stems have been uncoupled, the actuator may be removed from the valve body entirely:

The bonnet is held to the rest of the valve body (in this case) by four large studs. Removing the nuts on these studs allows the bonnet to be lifted off the body, exposing the valve trim for view:

Seats in Fisher E-body globe valves rest in the bottom of the body, held in place by the cage surrounding the valve plug. Once the bonnet is removed from the body, the seat may be removed without need of any specialized tools (left-hand photograph). A view inside the body shows the place where the seat normally rests (right-hand photograph):

With the bonnet removed, the plug and cage may be easily removed for inspection:

The packing follower (between the student’s fingers) has been removed from the valve bonnet, and you can see the upper Teflon packing rings within the bonnet. The student is also holding the packing flange in the same hand as the packing follower (left-hand photograph). In the right-hand photograph, we see the student using a screwdriver to gently push the Teflon packing rings out of the bonnet, from the bottom side. Care should be taken not to gouge or otherwise damage these rings during removal:

The left-hand photograph shows all the packing components stacked on top of each other on the concrete floor, next to the bonnet. From top to bottom you see the following components: a felt wiper, the packing follower, five (5) Teflon packing rings, a coil spring, and the packing box ring. The right-hand photograph shows the same packing components stacked on the valve stem:

Turning to the actuator, we begin disassembly by loosening the diaphragm hold-down bolts (left-hand photograph) and removing the upper half of the diaphragm casing (right-hand photograph). A single bolt secures the upper diaphragm plate to the top of the actuator stem:

In the left-hand photograph you see the student removing the spring seat, having previously loosened the spring adjustor nut. With the spring seat removed, the spring may be removed from the actuator assembly. In the right-hand photograph the spring adjustor and spring seat have been removed from the actuator stem. The student is now pointing at the valve spring, partially removed:

Sliding the actuator diaphragm, plate, and stem out of the actuator assembly from the top of the actuator makes it easy to remove the large actuator spring (left-hand photograph). The right-hand photograph shows all the moving actuator components re-assembled in their proper order outside of the yoke:

The left-hand photograph shows the lower half of the actuator casing, with the student removing six (6) hold-down bolts joining this casing half to the actuator yoke. The right-hand photograph shows the actuator casing half completely removed from the yoke, revealing a gasket and the bronze stem bushing (which serves to both guide the actuator stem and seal air pressure, since this is a reverse-acting actuator):

A circular spring clip holds the stem bushing in the yoke casting. The left-hand photograph shows the student using pliers to squeeze this spring clip and remove it from its groove cut into the metal of the yoke. In the right-hand photograph, we see the student using the wooden handle of a hammer to gently tap the bushing out of the yoke. The bushing has rubber O-ring seals between it and the yoke casting, so a small amount of force will be necessary to dislodge it. Using the hammer’s wooden handle to drive the bushing instead of a metal tool protects the relatively soft bronze bushing from impact damage. Note how the student’s right hand is waiting to catch the bronze bushing when it emerges from the hole, to protect it from falling against the hard concrete floor:

The final photograph shows the bushing removed from its hole:

  • Share
Published under the terms and conditions of the Creative Commons Attribution 4.0 International Public License
  • G
    GMM November 15, 2019

    In addition to the standard languages, some PLCs, e.g. the GE Fanuc Series 90, enable the use of C code/applications to program non-standard actions.

    Like. Reply
  • makwanavicky December 20, 2019

    Thank you for this neatly explained article.

    Please let me know if my understanding below is correct as the Modbus slave is a server side endpoint:
    1. All Modbus Slaves must be installed with a valid SSL certificate.
    2. Customer has to buy ssl certificates for all the IP addresses of the slave devices (wildcard will help minimizing the cost).
    3. Current age Modbus Slaves need to be replaced with the slaves that are capable of allowing installation of SSL certificates.
    4. Securing the Private Keys of each device’s certificate.

    Like. Reply
    • Bruno March 09, 2020
      You cannot buy certificates for private IP. Usually it will involve to have a Public Key Infrastructure (PKI) in your company. (Active Directory can do it, or other products. Anyway the difficulty is usually about the processes and who to trust)
      Like. Reply
  • M
    mtpattie664 January 28, 2020

    Can I download this?

    Like. Reply
  • Eric_W January 29, 2020

    Looking forward to Nokia 5G!

    Like. Reply
  • B
    Bijan Ghofranian February 02, 2020

    It is somewhat misleading advertisement because they want you to stay on their website and read it. But I was able to download this way. First click the pop out icon near top right hand side to open it in a new window. Then click on the open original icon on top right on the PDF window. Now the download icon becomes enabled. If it did not then try to print PDF and then save it as PDF.

    Like. Reply
  • U
    uchechukwu February 07, 2020

    I can not view the document right now.

    Like. Reply
  • A
    AR Lakshmann February 08, 2020

    Thank you Bijan Ghofranian

    Like. Reply
  • B
    Bromano February 11, 2020

    As a business owner from 1998 to 2014 and now as the engineering manager of an OEM that makes high throughput assembly machinery, the workforce gap has been an issue since the late 90’s.  Trying to find talent during this time has been somewhat difficult and about 4 years ago I started looking around at regional and then national colleges and universities that had a 4 year program, teaching the specific aspects of a control systems engineer.  I found all the typical EE and CompSci programs that maybe had a class or two for PLC programming, but no real comprehensive program that taught many of the aspects covered by a skills matrix I had for my Control Systems Engineers at my place of work.  The problem with hiring an EE or a CompSci major out of college is that one knew the coding methods and the other, some of the hardware design aspects.  But hiring either one meant a long period of OJT where I would lose efficiency of a tenured person to work with the new hire.  As a small business owner, I couldn’t afford that and even at my current place, we are so busy, I need to keep a very high level of efficiency and competency and my new hires need to hit the ground running.  That usually means that we need to hire control systems engineers from other positions at other companies and that comes at a premium as all the good engineers are gainfully employed.  So, I went back to my alma mater, The University of Hartford, two years ago armed with data, statistics, pay information and showed them that this line of work is not going away, in fact it is growing, and the skills gap and workforce shortage is growing.  I also came in with a skills matrix and a proposed curriculum of what I expect a Control Systems Engineer to have as a base knowledge.  The welcomed me with open arms and this past September, an Automation minor and an Automation concentration to the existing Electo-Mechanical Engineering program was rolled out.  There are two students that recently declared the minor and will be graduating this coming May!  To help out with this new program, UHart asked me to be an adjunct professor.  So now after nearly 40 years in the process and automation controls industry, starting as a junior in high school, I feel like I am giving back to help teach the next generation of control systems engineers!

    Like. Reply
  • J
    Jivan February 15, 2020

    Simple and clear

    Like. Reply
  • N
    Nyein Thet Wai February 17, 2020

    I’m learning PLC

    Like. Reply
  • R
    rajdua February 18, 2020

    Most of equations in table and some text are not rendering correctly in any browser.

    Like. Reply
  • K
    Kifa April 07, 2020

    I have heard about Solid State Relays but never understood the concept until i read this article. It is by far the best read i have had when it comes to understanding the concept and also the advantages of large load devices using solid-state relays instead of traditional mechanical relays. Thank you, Mr. David for this informative blog.

    Like. Reply
  • Akin April 08, 2020

    pressure relief valve,pressure safety valve

    Like. Reply
  • S
    scadametrics April 10, 2020

    Is special 2-wire cable required?

    The reason I ask is that there are many industrial I/O systems that have extensive 2-wire cables feeding a variety of RS-485 (Modbus/RTU) and 4-20mA instruments… I’m wondering whether the existing 2-wire cables can sometimes be re-purposed to implement Ethernet-based I/O instrument replacement?

    Another potential application is where many in-building analog intercom systems are based on 3-wire cables.  Replacing with an Ethernet-based intercom system would be desirable, but replacing cables behind drywall would be prohibitive.

    If the distances are short, perhaps the cable characteristics are somewhat forgiving?  I could see many neat applications for this—especially in retrofit applications, if existing cabling could be re-purposed.

    Like. Reply
  • J
    Jonas Berge May 16, 2020

    The ultimate user guide to Ethernet-APL (Advanced Physical Layer) / Single Pair Ethernet (SPE):
    -Technical description
    -Design & installation
    -Integration & architectures
    -Application protocols, interoperability, interchangeability
    -Commissioning & management
    -Control & safety


    Like. Reply
  • R
    rameshwar May 17, 2020


    Like. Reply
  • BIN95com May 22, 2020

    A lot in this article about what is AC. I would increase its value by adding more about ‘How Do You Measure AC Voltage’. For example instead of just informing to place COM lead on the reference point, I would explain the difference between using ground and neutral as your reference point, and voltage drop. Etc.

    Like. Reply
  • H
    HLCPP June 03, 2020

    pl add below Tags in Exemples
    ZSO - Position Switch Close
    ZSC - position Switch Open

    Like. Reply
  • R
    Rob Miller - Balluff July 01, 2020

    This is a long well-written article you have written. Proxy sensors most sensor manufacturers bread and butter. Here at Balluff, we offer a huge range of these on there own.

    I hope you do not mind me sharing this video but it shows sensors being used for object detection well. https://youtu.be/gM3JrM-2024

    Like. Reply
  • R
    Rob Miller - Balluff July 01, 2020

    An interesting new product they have produced there.

    Like. Reply
  • G
    Gsteel July 13, 2020

    How does ground side (neutral) motor control circuit switching square with NFPA NEC 430.74?  Consider a typical control circuit wherein the Hot conductor leaves the controller enclosure, runs to a remote start station, and returns to the enclosure either to the starter coil then OL device or to the OL device then starter coil.  In this circuit a grounded conductor either trips the short circuit / overcurrent device instantly or when the start circuit is enabled, depending on which conductor is grounded.  In either case the motor does not start.  The value of not preventing an orderly shutdown, alarm signaling, etc. is clear.  However, 430.74 would seem to allow ground side switching only if there are no remote stations.  I am, of course, referring to US installations only.

    Like. Reply
  • E
    Erric Ravi July 14, 2020

    Industrial IOT is always best for industrial automation

    Like. Reply
  • pnachtwey July 31, 2020

    The integrator gain is not unitless.  It has units of output/(error*time)

    Like. Reply
  • pnachtwey August 01, 2020

    The diagram shows a PID with all the gains acting on the error.  This introduces zeros into the closed loop transfer function which will result in overshoot.  Some controllers have an option for the derivative gain to act only on the changes in the PV or feedback.  I call this a PI-D.  If the proportional gain also acts on the feed back there will be no zeros.  I call this I-PD.  Bandwidth is reduced but there are not zeros that cause overshoot.  Another advantage is that since all three gains are integrating.  The integrator is integrating errors.  The proportional term is integrating changes in the PV and the derivative gain is integrating changes in the rate of change in PV.

    Like. Reply
  • A
    aktifnet August 02, 2020

    You have provided very good information about AC Motor Braking, but you have not given much information about Dynamic Braking Resistors. I would like to leave a source link on this subject. https://aktif.net/en/product-and-services/motor-control-and-braking-resistors/dynamic-braking-resistors

    Like. Reply
  • pnachtwey August 13, 2020

    Processing power isn’t much of an issue these days.
    It takes a little skill do to a proper system identification to get an accurate model.
    Something that wasn’t mention is that the longer the dead time, the further one most calculate into the future.  This increases the need for CPU power.  CPU power can be reduced by updating at intervals that are twice as long if the application doesn’t require high frequency changes in the set point or target trajectory.

    Like. Reply
  • R
    RUBEN August 15, 2020

    very informative

    Like. Reply
  • pnachtwey August 16, 2020

    In the case of the RMC controllers the feed forwards gains are determined by the inverse of the open loop transfer function.  The motion trajectory and the feed forward outputs must be computed on-the-fly because often commands are issued while the actuator is still in motion.  I have NEVER been provided a model ahead of time.  The feed forwards gains can be estimated but usually they are found empirically by trial and error or by using an auto tuning program

    Feed forwards can be used for temperature systems too.  The control gain usually has units of degrees/% control output.  So if you want to estimate the output for changing the set point to 50 degrees above ambient you would multiply the 50 degrees by the inverse of the the controller gain which will have unit units of % control output per degree.

    The big advantage of using feed forwards is that it will estimate what the integrator term will eventually reach quickly instead of letting the integrator term wind up over time , over shoot then unwind.

    Like. Reply
  • pnachtwey August 28, 2020

    Modbus TCP and ProfiNet are two very different protocols.  What they have in common is that they use the same hardware but after that they differ.
    The Modbus TCP stack is very small and simple and is only good for transferring basic PLC types.  The byte and word order is a problem for 32 bit values because it isn’t really defined in the specification.  Manufacturers of equipment that Modbus TCP do what they please and sometimes this results in incompatibilities or 32 bit data must be re-arranged after receiving to be in the desired format..  We have seen this way too often.  We checked our code by testing it with a Modicon PLC to make sure we could send and received 32 bit DINTs and REALs in the right order without having to shuffle bytes or words afterwards.  Another problem with Modbus TCP is the TCP part.  You want to use TCP to send data such as parameters that must be right.  However, when sending basic I/O back and forth,  TCP is a problem because if there is an error then TCP will force the re transmitting of old data.  It would be best if there was a standard Modbus UDP because UDP wouldn’t force the re-tranmission of old I/O data.

    ProfiNet is a big beast.  The stack size is huge.  The ProfiNet stack size is bigger than the Ethernet I/O,  Omron Fins,  Mitsubishi’s protocol and Modbus TCP combined.  ProfiNet has a lof of code that is run early on that will make define connections or packets.  ProfiNet allows for the I/O data to be transferred back and forth very efficiently.  It is just too bad that the start up code is so big.  It is good thing that memory is cheap.  ProfiNet is a HUGE improvement over Siemens’ previous Ethernet protocols.

    ProfiNet devices must be certified to claim ProfiNet compatibility.  This is a big deal for the customer and the vendor.  Obviously the cost of the testing is eventually past on to the end user but it also means there are no problems with the byte ordering that occurs with Modbus TCP.

    We have implemented many Industrial Ethernet protocols on our product so we are quite aware of the development time,  the amount of resources, difficulty to implement and support that each protocol requires.

    Like. Reply
  • M
    murad_kuri August 31, 2020


    Like. Reply
  • P
    philipmarshall1 August 31, 2020

    Modbus/TCP was the 1st industrial Ethernet protocol, its basically Modbus on Ethernet.  PROFINET is at least two generations of functionality improvement over Modbus; 1st generation being PROFIBUS.  The 1st question that was asked in creating PROFIBUS was “how do we make this better than Modbus”, vastly that is diagnostics.  Diagnostics to reduce downtime thus improving productivity.

    Like. Reply
  • X
    XiaoJun September 10, 2020

    The Autonomous mobile robots is applied to express industry.

    Like. Reply
  • R
    rogerval September 18, 2020

    Thanks for the article, it was interesting.
    However, I have a query about the photo which is supposed to depict a “linear power supply”.
    Looking at the number of chokes, and what appears to be a high frequency transformer instead of an iron cored mains transformer, I suspect that this is a switchmode power supply of some sort.
    Kind regards, Roger Valmadre.

    Like. Reply
  • V
    Vietnamtransformer September 26, 2020

    that’s a good article, thanks

    Like. Reply
  • johnand October 01, 2020

    We are using several level sensor for our industries for a long time and they still working well.

    Like. Reply
  • pnachtwey October 01, 2020

    The author didn’t say “Why bother?”
    SFC are good for machine control.  SFC make it easy to organize ladder or structured text in steps or states.

    SFC also makes the program more efficient because only the active step or state gets scanned instead of scanning the whole ladder.

    Like. Reply
  • M
    mon foret October 06, 2020

    hello there;
    if i have continous real signal (4-20 ma) from pressure smart sensor, so to get high precision should i choose input module with high resolution? then i should use instruction ladder to make scaling of signal but. so what i do to this signal to make my pc to read this signal by modbud tcp/ip with saving of real signal in high precision.

    Like. Reply
  • K
    khirab01 October 27, 2020

    This site is the best for me

    Like. Reply
  • callaghanpump November 04, 2020

    A metering pump passes a specific volume of liquid in a defined time period providing an actual volumetric flow rate. Delivery of fluids in specific adjustable flow rates is sometimes called metering.

    Like. Reply
  • pnachtwey November 16, 2020

    Ladder logic is a poor tool for implementing state machines.  Obviously it can be done but SFC or sequential function charts are much better.  In each step one can use ladder ( LD ) or structured text ( ST ).  I prefer structured text.  On top of that, a good SFC implementation will high light the current active steps.  Better yet is to log the transitions from one step to another.

    Like. Reply
  • patrickduis November 20, 2020

    Excellent article. I also program a lot of state machines in Pro-face HMI/SCADA touchpanels. The big benefit of these Japanese touchpanels is that they also contain a ladder logic soft PLC. Even the small GP4116 ones also. When the customer makes a new machine and is not 100% sure how things should work, you should use this way of implementation. It makes sure that your program is servicable in the future and it can be modified very easily when the customer comes with changes lateron.
    It looks like a lot of work for a simple solution, but when more and more changes come this way of implementation actually saves a lot of time and money at the end. And what is most important: everything works like it should, and no strange things happen (if you didn’t make a mistake).
    I shared the article on my LinkedIn page.

    Like. Reply