Hi Friends

I want to know if there is any difference in SRV, GCV, LFBV and IGV LVDT calibration method in GT MKV/MKVI/MKVIE. Please furnish in detail

Thanks
AK

 

Hi "friend",

What you really need is formal training, followed by practical experience. I do not understand how you can expect answers "in detail" without providing *NONE* yourself.

And please, use the search feature. You owe everyone that courtesy before you can assume you can be spoon-fed answers.

I am normally a silent spectator, but I just had to drop into this one. For obvious reasons.

 

silent_majority,

What kind of information do you expect should be provided with a question of this nature?

 

For starters:

-Unit Size? (Frame)

-Control system? (which one does he work with on site? the procedure differs slightly depending on Mark-V, Mark-VI, and Mark-VIe. He has merely listed all of them, I doubt his site has all of them.)

- Relevant detail of site specific hardware? For Example, for the LFBV, does it have an LVDT? (if it does not have an LVDT, it cant be calibrated).

- Specifics? it is just too vague asking "whats the difference in calibration", and then expect it to me permuted for three different control system offerings. A much better question would be "what is the difference between the calibration of the GCV and the IGV for a Frame 9E machine with a mark-VIe control system. An even better one would seek answer to a particular aspect of that calibration procedure. The "procedure" should be in the control specification (RTFM). If something is unclear, people here can help. "Help me to help you".

 

Okay, silent_majority. Let's say the original poster is working on a Frame 5 with DLN-I combustors driving a generator using a Mark VIe control system.

Can you provide a procedure?

If not, what control system could you provide a procedure for?

If you want the poster to provide more detail, please tell the poster what you require.

By the way, a part-time responder to this site, ProcessValue, had previously uploaded a Mark VI procedure to a web-sharing site and posted the link here. I don't know if the procedure is still available or not. I do caution it was quite weak on some points.

Here's the link to the document:

http://www.2shared.com/document/TSYug7uX/sevo_calibration.html

Here's the link to the control.com thread which the original posting was made:

http://www.control.com/thread/1289667103

I have become aware of some "fishing" going on here on control.com. Some contractors and others (including GE employees) are looking for step-by-step procedures for LVDT calibrations for various devices on various Speedtronic control systems. So, a generic request like this would certainly seem to fit this bill, wouldn't it?

Note also that the originator checked 'Other' under Vendor. It wouldn't surprise me to find other turbine control suppliers and their employees looking for information on control.com to provide with their competing products and services. And that's fine; it's a free world and everyone gets their information from different sources. (We know GE doesn't readily provide such information for the equipment they provide.)

I'm going to state the basics of calibrating LVDT feedback in GE-design heavy duty gas turbines, which are true for any version of Speedtronic turbine control system. The basic concept is to make the LVDT feedback equal to the actual measured position of the device the LVDTs are attached to. It's that simple. [Note that most people don't actually measure device position, and just ass-u-me that the LVDT feedback is equal to the actual position, which may or may not be true.]

1. Ensure the null bias current values are set to the Control Specification value (default value) prior to any calibration procedure. That's -0.8 mA for SIMPLEX applications, and -0.267 mA (both converted into percent, as necessary) for TMR applications.

2. Ensure the polarity of the current being applied to all coils of the servo-valve of the device whose LVDTs are being calibrated is correct.

3. Ensure the LVDT zero-stroke voltage setting is within the allowable range for the type of LVDT being used.

4. Use some measuring device as appropriate (dial indicator, calipers, machinist's protractor, etc.) to independently measure the device's position as necessary. This could include measuring the total range of travel prior to any calibration, and certainly includes measuring the device's position when verifying the accuracy of any calibration.

5. Determine, as necessary, the 100% stroke range for the device who's LVDT feedback is being calibrated. (In other words, many gas valves were not designed to be 0-100% from full closed to full open; the 100% stroke range was supposed to be something less than full stroke, but most sites don't use that value. And the turbines run just fine.)

6. Establish hydraulic pressure as necessary after reviewing the Piping Schematics (P&IDs) to determine what solenoids might be required to be forced and which pumps need to be running.

6. Execute an LVDT calibration using the method specific to the version of Speedtronic being used.

7. Verify the accuracy of the LVDT calibration using the appropriate measuring instrument.

8. Adjust the null bias current only if necessary or required to make the actual measured position equal to the feedback (which should be equal to the reference), plus or minus 1-2% (for a device calibrated in percent of stroke; 0.5 degrees for the IGVs). !!!<b>DO NOT</b> adjust the null bias current solely to make the LVDT feedback equal to the reference!!! Remember, there are thousands of GE-design heavy duty gas turbines that are running perfectly with the default values of null bias current. If you can't be bothered with measuring the actual position of the device to compare against the LVDT feedback, <b>leave the null bias current value at the default value.</b>

When "re-calibrating" LVDTs, one should check to see if they need re-calibrating, by performing and documenting the as-found condition. If no change is required, then no action is required. (Just like for pressure switches, and pressure transmitters, and the like. LVDTs are no different; if the as-found condition is good, then no adjustment is required.)

For the Mark V, there are some descriptions for using AutoCalibrate. If one understands the sequence and the process, one can read the Help files for LVDT calibration in Mark VI Toolbox and Mark VIe ToolboxST to get the particulars for using those applications to stroke the devices and to calibrate the LVDT feedback.

But, most sites don't really understand what is being calibrated (LVDT feedback, not the SRV or the IGVs or the GCV or the LFBV--the LVDT feedback from these devices). And they don't understand how the Lube Oil and Hydraulic Oil and Trip Oil systems work together (and are controlled by the Speedtronic) to get hydraulic pressure to the actuator through the servo for these devices equipped with LVDTs. And, many Frame 5s are very different from other Frame 5s; some have AC motor-drive Hydraulic Pumps, many do not. Some have Trip Oil, some have high-pressure Hydraulic Trip circuits instead of low-pressure Trip Oil circuits. Every Frame 5 is not the same as every other Frame 5. Yes, they all compress air and burn fuel and produce torque, but they do so with different auxiliaries and systems and require slightly different procedures (step-by-step procedures) for LVDT calibration.

But the basics, outlined above, always remain the same: To make the LVDT feedback and the actual position match each other as closely as possible.

And if the actual position isn't being measured, then the whole process is just a guess and doesn't deserve much attention to detail. And there are thousands of turbines around the world which get very little attention to detail, and the actual position is never compared to the LVDT feedback, and they run just fine producing torque and income and profits.

Other than the above, any step-by-step procedure should be produced by personnel on site who are familiar with their equipment and operation and settings. They should be a part of the Standard Operating Procedures, and/or the Maintenance Procedures.

And one size does not fit all machines and applications.

 

Hi CSA

I have exposure in mark V but not in mark VIe and Mark VI, my doubt is with the regulators used in mark V during LVDT calibration for SRV, GCV, IGV and LFBV are not used in Mark VIe. Are they configured internally in software and change automatically or we need to change the regulators to position control loop before starting manual calibration.

I understand rebooting RST is not required in Mark VIe after calibration.

Please can you clarify me

thanks

AK

 

Using the Mark VI Toolbox or Mark VIe ToolboxST feature to stroke and calibrate servo-operated devices equipped with LVDTs does not require changing regulator types.

Once the LVDT calibration is completed with Mark VI Toolbox and Mark VIe ToolboxST I believe all that's necessary is to upload the new calibration parameters to "back-up" the new calibration parameters.

If one uses the Mark V AutoCalibrate feature, it's also not necessary to change regulator types.

 

Hi

Thank you

AK

 

CSA;

I have been thinking about your post about the LVDT %100 calibration position. As you state it should be measured and compared to Control Spec. rather than just using the full opening of the valve.

I reviewed our Control Spec and it states for the Gas valves to Enter a Setpoint of %125 and use this as the %100 value for calibration.

Our machines are 7FA, MKV controls, DLN 2.6

 

MPAC,

What you have written sounds like it was taken out of context. Without being able to see the Control Specification I would not be able to make any comment on the information you are providing.

There are control valves that have been used by GE that have a physical stroke that is greater than the 100% stroke value specified in the Control Specification.

There are also control valves that have been used by GE that have a physical stroke that is less than the 100% stroke value specified in the Control Specification.

But, I can't understand what you have written from the Control Spec. It just does not compute.

In my experience, the valves used for controlling gas fuel flow to GE-design heavy duty gas turbines are mechanically adjusted such that 0% stroke is the fully closed mechanical stop and 100% stroke is the fully open mechanical stop. The flow through many of these valves are not generally directly proportional to stroke, meaning that 50% flow does not occur at 50% stroke. So, there is a function in the Speedtronic that compensates for the non-linearity.

Now, <b>I'm NOT saying</b> that the valves used at your site are like the description in the paragraph above. I'm just saying that the majority of the valves I've seen and worked on are like the ones in the paragraph above.

You say you have a Mark V. If AutoCalibrate has been used for calibrating LVDT feedback, then you can look in ACALIB.DAT to see what values were used for the 0- and 100% stroke setpoints.

Now, I have a request of you. Why is it felt it is necessary to calibrate, or re-calibrate, LVDT feedback? The only reasons to calibrate, or re-calibrate, should be when an LVDT is replaced, or when the valve is disassembled for maintenance/repair and reassembled.

Prior to "calibrating" or "re-calibrating" LVDT feedback do you perform any kind of check to see if the calibration requires adjustment? In other words, do you manually stroke the valve, stopping at intermediate points, and check to see if the actual position matches the feedback/reference?

Please take a couple of minutes and answer the questions in the two paragraphs above. It will be very helpful to me, and when trying to help others with this process.

Thanks!

 

> I reviewed our Control Spec and it states for the Gas valves to Enter a
> Setpoint of %125 and use this as the %100 value for calibration.

I know what that means; it's telling you to put in a manual stroke (position) value of 125% when you're measuring the full range of travel (stroke) of the valve. That should cause the valve to go to full open mechanical stop, and that is the 100% stroke position. Then you measure that position and that's what you use when verifying the accuracy of the calibration (at, say, 25-, 50-, 75- & 100% stroke references/positions). In other words, when you put a manual reference setpoint of 25% you should measure approximately 25% of the value you measured at 100%, and so one for the other values during your verification. I believe that most LVDTs GE uses are supposed to be accurate with 1% of the range of travel of the LVDT.

So, the Control Spec seems to be saying that the valves should be calibrated such that the fully closed mechanical stop is 0%, and the fully open mechanical stop is 100%.

Which seems to match my experience on most gas valves used on GE-design heavy duty gas turbines equipped with DLN combustors. (Not that that means anything, except that it agrees with what I wrote in my previous response.)

 

CSA;

Sorry, I guess I misread what you had said in the paragraph below. When I check calibrations or calibrate valves I just enter -25% for the zero stroke LVDT voltage and 125% for the %100 LVDT voltages. When I read your statement below I thought I should be looking for an exact valve travel and set the 100% valve travel with a caliper. I also remember them talking about using the 100% stroke measurement from the control spec when I took the MKV maint. course a few years back. I have watched GE TA's perform the calibrations a couple of times and have only saw them take a measurement went checking IGV's not GCV's.

(5. Determine, as necessary, the 100% stroke range for the device who's LVDT feedback is being calibrated. (In other words, many gas valves were not designed to be 0-100% from full closed to full open; the 100% stroke range was supposed to be something less than full stroke, but most sites don't use that value. And the turbines run just fine.)

>You say you have a Mark V. If AutoCalibrate has been used for
>calibrating LVDT feedback, then you can look in ACALIB.DAT to see what values
> were used for the 0- and 100% stroke setpoints.

I have never used the AutoCal. feature and am a bit hesitant as I have never seen it used on site so I'm not sure it has been set up correctly or not. I have no problem calibrating manually using the AutoCal screen buttons and that seems to be what the TA's that I have seen do.

> Now, I have a request of you. Why is it felt it is necessary to calibrate, or
> re-calibrate, LVDT feedback? The only reasons to calibrate, or re-calibrate,
> should be when an LVDT is replaced, or when the valve is disassembled for
> maintenance/repair and reassembled.

I have only perform calibrations on SRV's as we have changed out a couple of these as they seem to be prone to fail leak checks. I have also had to calibrate Steam valves as steam turbine won't start when valves reads >1% open. The TA's calibrate these valves cold and then when they leave and the valves warm up they need to be adjusted.

> Prior to "calibrating" or "re-calibrating" LVDT feedback do you
> perform any kind of check to see if the calibration requires adjustment? In
> other words, do you manually stroke the valve, stopping at intermediate points,
> and check to see if the actual position matches the feedback/reference?

Like I said I have not calibrated LVDT's on GCV's, I have verified positions though as well as Null Bias. I only entered setpoints and checked LVDT position, not checking actual position with a caliper. I have done this to all our gas valves and submitted data to GE for a PAC case, and was not asked to measure actual position.

I was told by a TA that if calibration of GCV's is modified that unit need to be tuned. I'm not sure if this is true or not or if he was just making sure he was called next time.

Thanks for the Clarification

 

If one doesn't take a physical measurement of the actual physical stroke (travel) of the valve or device, how will one verify the accuracy of the calibration at 25-, 50-, 75%? It's important, wouldn't you say, to know what the physical travel is, so that one can calculate 25% of it, and 50% of it, and 75% of it so that when the device is told to go to 25% and the actual, physical position is measured and compared to the calculated value one will know if the calibration was accurate or not.

<b>Most</b> people never bother with measuring anything. Ever. They just bypass those steps, because the Speedtronic is just <b>that good</b>--it always calibrates LVDT feedback <b>perfectly</b>.

If that's the case, why are those same people always re-calibrating LVDT feedback, and never bothering to measure anything, ever? Any time there's a problem with starting (establishing flame), or shutdown, or loss of flame trip, or instability, or anything which could <b>possibly</b> be attributed to a servo-operated device with LVDTs, these same people rush to re-calibrate the LVDT feedback--and <b>***NEVER, NEVER, NEVER***</b> bother to determine if the LVDT feedback even required re-calibration. And they <b>***NEVER, NEVER, NEVER***</b> bother to measure and verify the re-calibration when they've finished. They just "re-calibrate" and pronounce everything fixed. (Though it's usually NOT.)

It's simply, and completely, asinine. People just don't understand the basic fundamentals, and they go blindly on what they've seen others do or been told by others to do. They waste huge amounts of time and lost production, for very little, or, in most cases no, gain.

Do you agree? <b>The whole purpose of calibrating LVDT feedback is to make the LVDT feedback equal to the actual, physical position of the device the LVDTs are attached to.</b>

If one doesn't make any measurements and comparisons (verification), then how the <b>HECK</b> does one know if the LVDT calibration is accurate or not<b>?????</b>

You said you use -25% reference to obtain the 0% voltage (at the fully closed mechanical stop). Well, if you set the zero on a dial indicator, or took a reference measurement with a dial caliper at this position, that would be the zero position. You said you then use a reference of 125% to get the device to go to the fully open mechanical stop (ostensibly, the 100% stroke position) and take a voltage measurement. If you read the dial indicator, or took another dial caliper reading from the same reference point and subtracted the first dial caliper reading from the last, you would then know what the actual, physical stroke (travel) of the device is. And you could use that information to verify the accuracy of the LVDT calibration. And one should only be adjusting the null bias current when verifying LVDT calibration. (In other words, that crap in the Control Spec about how to adjust null bias current is just that: pure, unadulterated crap. But, that's been covered in other threads in control.com.)

GE TAs are not always good role models. A lot of what GE TAs "know" was obtained from what's called "tribal knowledge". That means that information is passed, verbally, from one person to another to another to another. There are lots of problems with "training" people like this, or allowing myths and legends and falsehoods to persist through tribal knowledge. This means that many GE TAs have very different methods and "knowledge", which is not very good. They should all be using the same methods and knowledge to perform their work.

As for not using AutoCalibrate, well, you've been listening to GE TAs or people who were also trained by "tribal knowledge" from GE TAs. There was a very influential TA back in the early Mark V days who went around telling everyone who would listen, and even those who wouldn't, that AutoCalibrate didn't work. That individual didn't know about ACALIB.DAT and how it was used by AutoCalibrate and how it had to be manually configured by the user. And, this same individual also "thought" that AutoCalibrate should not require a download and re-boot after completion--in other words, it wasn't "Automatic" if the user had to manually put information in the I/O Configurator, download to the control processors, and re-boot them to "complete" the AutoCalibration process.

As a result, many, many, many people were "poisoned" against AutoCalibrate. And that myth continues to this day.

The only time I have <b>ever</b> had difficulty using AutoCalibrate is when the IGV actuator is very, very worn or the bolts holding the IGV actuator stationary are loose and worn and in one case two of the four bolts were even missing (YES! Missing!), and the machine's owner refuses to repair the problem. In that case, one is forced to manually calibrate the IGV LVDT feedback.

Yes; I do also subscribe to the theory that if the LVDTs of gas control valves of machines equipped with DLN combustors are re-calibrated without any kind of verification prior to or after re-calibration that it's pretty likely that the machine may require re-tuning.

 

CSA;

Yes I would agree that the purpose of Calibrating LVDT feedback is to ensure the LVDT feedback matches the actual position.

So, yes it would be a good idea to verify actual position with a dial indicator. This is stated as the last step in the Control Spec, but like I said I have never saw a TA perform this task.

From reading your post I also find it interesting how seemingly unfamiliar most GE TA's seem to be with the equipment they work on.

The only other question I have that you brought up is the difference in checking the null bias. The procedure in the control spec. would have you calculate servo current by measuring servo voltage, then averaging the values from all three cores. The method you describe has you adjust the null bias to make the actual position match the LVDT feedback. I don't understand how changing the null bias would change how the actual position relates to the LVDT feedback. I would think that changing the null bias could make the LVDT feedback (and Actual Position) come closer to the Reference. I must be missing something but I don't understand how null bias could effect how LVDT feedback relates to actual position.

Could you please explain.

Thanks

 

MPAC,

It's not the fault of the GE TAs; it's the fault of their management for making some very false assumptions about procedures and documentation, and for having an organization that is so fragmented that the wrong division is in charge of training TAs on equipment produced by another division that is applied to equipment produced by yet another division.

But, enough on that.

As for the LVDT calibration "procedure" in the Control Specification, it's just a very high-level "check-list" that purports to be something it is not. Remember that the same procedure is generally provided for any and all turbines equipped servo-operated devices with LVDT feedback, and in many cases the devices are very, very different.

And, if the Control Specification procedure for calibrating LVDT feedback says to verify the LVDT feedback against the actual, physical device position, why do people choose not to do that portion of the procedure? Is it permissible--for anyone, because you are also selectively performing, or not performing, a portion of the procedure--to selectively perform or not to perform any portion of a procedure?

It's been said for many years on control.com: When you read any written documentation or procedures provided with GE-design heavy duty gas turbines about GE-design heavy duty gas turbines, read it for INTENT, unless you wrote the document or you know who wrote it. In other words, read any document or procedure, then re-read it, then re-read it, and after considering what it tells you in conjunction with what you know about the equipment at your site, use everything available to you, including the document or procedure provided, to develop your own document or procedure that is specific to the equipment at your site.

Now, moving on to the the question of null bias current and null bias current adjustment. I would like to develop this topic by asking questions of you and working together to come to an understanding of what null bias current does and why it might be necessary to adjust null bias current and how null bias current might be adjusted and how to know if the adjustment is satisfactory.

If you really want to get to the "bottom" of this topic, I will work with you to get there, but you're going to have to be part of the process. I have described it several times and I must not be doing it correctly, so any further attempt is going to require the participation of the questioner.

AND, it's going to have to start from a very basic point, perhaps unfortunately for you, but I suspect that a large part of the confusion and questions about null bias current adjustment stem from the fact that there is a fundamental misunderstanding about what null bias current does and what effect changing null bias current can have versus what changing null bias current actually does.

So, let's get on with it.

First, let's come to the same understanding about what null bias current does. What do you, MPAC, think null bias current does?

When you follow the null bias current adjustment procedure in the Control Specification when do you feel that the null bias current adjustment is satisfactory? In other words, when you make an adjustment to the null bias current, what changes in LVDT feedback or actual, physical device position do you see and when do you consider that the null bias current has been satisfactorily adjusted?

Please describe how you perform the null bias current adjustment following the procedure in the Control Specification. Be as specific as possible. Do you make an adjustment, then check to see if that adjustment was satisfactory (and what do you deem satisfactory) and make another adjustment, or do you just make an adjustment to the null bias current then continue on, ignoring at least one subsequent step (the verification of actual, physical position versus the LVDT feedback)?

How do you perform a null bias current adjustment? What is the starting condition? Do you start with the null bias current set to the factory recommended value and then make adjustment(s) as necessary? Or, do you start from the last null bias current adjustment value and make any further adjustment(s) as necessary?

And, when do you deem that the null bias current adjustment is satisfactory and does not require further tweaking?

So, to sum up: What do you think null bias current does?

What do you think adjusting null bias current should do?

How do you perform a null bias current adjustment? For answering this question, please be as specific as possible.

What is the starting null bias current value when you perform a null bias current adjustment?

When you make a null bias current adjustment, how do you determine if the adjustment was satisfactory? What are the parameters you use to judge the success of the null bias current adjustment?

I look forward to your complete answers to the above questions in order to continue.

 

Hello.

I need some help with the GCV autocalibrate in a MKV simplex.

My inquire is if there is some procedure as in GCV autocalibrate in a MKV TMR. Since in a TMR we have three cores <R> / <S> / <T> and in each one connected one servo coil, but how would it be in a simplex where we have only one processor <R>. According to what I read de GCV servo has two coils connected to <R>.

How is the procedure to carry out the autocalibrate in this case where the interface is IDOS and a conventional combustion system (no DLN).
thank you very much in advance.

 

http://www.2shared.com/document/TSYug7uX/sevo_calibration.html
this link isnt working please re share..

 

@Gtlucifer,

Interesting pseudonym.

The person who posted that link hasn’t been heard from in many years. He may still read threads, or visit once in a, but he hasn’t contributed anything in a very long time (unless he’s changed his pseudonym…!).

The Mark* IV Maintenance Manual had some procedures and maybe had one for IGVs (is that what you’re looking for—a Mark* IV IGV LVDT calibration procedure?).

It’s been a very long time since I’ve worked with a Mark* IV. And writing a step-by-step procedure for anything involving a GE-design heavy duty gas turbine with a Mark* turbine control system is really hard to do without a copy of the Control Specification document and copies of several of the P&IDs. And even then it’s not easy. For a Mark* IV one would also the Mark* IV Speedtronic Elementary. All of these documents are specific to the machine in question.

Do you know what logic signals need to be forced to get hydraulic pressure to the servo & actuator? Do you know how to establish hydraulic pressure? (Some machines in the Mark* IV era didn’t have auxiliary hydraulic pumps.). Because all of these questions are important parts of the procedure.

The Mark* IV Maintenance Manual has ONLY the instructions for "stroking" the hydraulic-actuated devices (and only the ones with LVDTs), it will not have any instructions for forcing logic, establishing hydraulic pressure. And, with Mark* IV-equipped GE-design heavy duty gas turbines it is really necessary to understand how to measure actual physical device positions.

Tell us what you know, and what you are trying to accomplish, and perhaps we can help you. To my knowledge, there are no detailed written procedures anywhere for calibrating LVDT feedback.