Frame 5 GT# Problem with Ratchet & Diesel Engine Overshoot


Thread Starter


I have an issue with the ratchet system of our GE Frame 5 p Gas Turbine. The Gas Turbine model is PG-5361P of 1986.

The ratchet system appears to have weaken over times and during start up, i.e cranking off of the unit, the diesel engine will overshoot.
Please refer to links below 1. Sound of Ratchet of our GT#;
and 2. Start trend from MarkVIe.

Please also note that after reaching standstill position (14HR) after stopping of unit, sometimes we get the Alarm (Clutch did not engage) as the ratchet takes too much time to engage the clutch.

We have checked the inline filters, ratchet motor, pump and relief valve VR5. All appear OK.

Note that the existing torque converter is of VOITH manufacture. If needed I can share the P&I Diagram.

The hydraulic ratchet mechanism serves two purposes: One, to slowly rotate the shaft (approximately 1/8th of a turn every three minutes) during cooldown) to prevent bowing the axial compressor rotor after shutdown/trip from loaded operation.

And, two, to assist the starting means with breaking the turbine shaft away from zero speed during starting. In other words, the starting means, working through a torque converter, usually does not develop sufficient torque to overcome the inertia of the turbine-generator shaft(s) at rest to accelerate them to purging and firing speed. So, when the starting means is ready to start accelerating the shaft away from zero speed (so, in the case of a diesel engine starting means--when the diesel warm-up time is complete and the diesel is accelerated to near rated speed) the hydraulic ratchet is energized to help "bump" the turbine-generator shaft to get it to start turning so the starting means can accelerate the shaft to purge/fire speed.

I'm not able to get the files to download correctly (and, frankly, I'm worried about viruses, too, so I didn't work too hard at downloading them). And your comment about the "diesel engine overshoot." How can a diesel engine overshoot?

So, the hydraulic ratchet self-sequencer mechanism, VH-14 I believe it's typically called, could be the problem. Have you tried replacing it?

When the hydraulic ratchet mechanism is operating it provides hydraulic flow and pressure to the forward piston(s), and then to the reverse (retraction) piston. (There are usually two forward pistons to produce the torque necessary to turn the turbine-generator shaft, and only one reverse (retraction) piston because it doesn't take very much force to retract the ratchet mechanism in preparation for another forward stroke.) The way the self-sequencer does this is by sensing a high pressure when the pistons reach the end of their travel (stroke)--the hydraulic pressure builds up for a couple of seconds, which causes the shuttle valve in the self-sequencer to move and port hydraulic pressure/flow to the opposite piston(s). So, it's common for a brand new turbine and ratchet mechanism to make a scratchy sound at the end of both forward- and retraction strokes--again as the pressure builds up because the piston can't move any further and because the high pressure is required by the self-sequencer to shift the hydraulic pressure/flow to the other piston(s).

The "ratchet" doesn't engage the clutch; there is a separate solenoid, 20CS, which is energized to apply some of the hydraulic ratchet pump flow/pressure to the two pistons which move the non-stationary side of the jaw clutch to close and engage the jaw clutch teeth. The teeth are slanted, so they don't fully engage until the ratchet mechanism turns the jaw clutch to completely engage the jaw clutch. It seems as if there may be some problem with the adjustment of the jaw clutch engagement limit switch, 33CS-1.

It could also be--and this does happen over time--that the pistons of the ratchet mechanism get worn and the seals leak and they don't transmit as much force as they once did. This is typical. Dirty oil is also a problem for the self-sequencer, as well as for the pump and the ratchet pistons.

Other than the above, I don't understand the rest of the conditions/post.

You can post the P&ID, but if it's a typical ratchet mechanism from that vintage with a self-sequencer it's probably pretty standard, though GE usually provided Twin-Disc torque converters (some of the companies licensed to package GE-design turbines did use other manufacturer's torque converters). So, please post the P&ID, and please explain the diesel engine overshoot comment.

Hope this helps!
The only other thing that I can see, other than what CSA has said, is a clutch problem. In two places you have mentioned clutch problems. If the clutch is not releasing on start and staying in, maybe that is what you mean by diesel overshoot? Remember, the clutch should drop out when the turbine is producing more torque than the diesel, if the clutch stays in, the turbine will be driving the diesel (not good). You have also mentioned the clutch not coming in at standstill when the ratchet starts.

I would check the whole clutch assembly, the hydraulic cylinders, the limit switch and the clutch splines for any damage.
Dear CSA

Thanks for the precious advice.
Please check the following youtube link

which shows the extract from MarkVIe for crank off of the unit. You will note that just the moment, the diesel engine breaks off, the 20CS solenoid will chatter, and also the diesel engine speed overshoot. This problem is intermittent. It mostly occurs during cold starts but also did happen during hot starts. Sometimes it is just normal.

I have checked the clutch splines, limit switch and cylinders and all appear to be OK.

From the youtube video, you will also hear the sound of the ratchet stroke. In my opinion, it is not normal as when I compare with our other Frame 5 GT#, the ratchet sounds shows a weak system and also the crackling sound you will hear at the end of the video, is specific to this ratchet system.
Dear Sir,
Thank you for your comments.
First of all, you may wish to refer to the self-explanatory youtube video

I have posted a little earlier.

The Clutch & hydraulic cylinders appear to be OK. And Limit switch elec wiring is also OK.

The main problem which is intermittent is first that the diesel engine overshoots to crack off the turbine. What appears to be the problem is that just at the moment the diesel engine accelerates, the ratchet stops instantly and restarts 1 second after. And also you will find that 20CS will chatter at this specific moment.

Secondly, again intermittently, but now becoming more frequent, when the turbine reaches 14HR (standstill) position, the ratchet takes a little more time to reengage the clutch. So we are getting the 'starting clutch failure to engage Alarm.

Like CSA rightly stated, i have a doubt about the self-sequencing valve. May be some internal leak.

for the P&I Diagram

Thanks for providing your comments.
Dear Sir,

So the sequence of events during a start is that the hydraulic ratchet is energized--along with 20CS-1--until the unit breaks away from zero speed. At that point the diesel starting means should be transmitting a LOT of torque, which should be keeping the clutch teeth engaged. And the clutch--as described--disengages when the speed of the turbine side of the jaw clutch is higher than the starting means side. Which normally occurs when the unit speed gets above approximately 50-60%. Or, which abnormally occurs when the diesel speed suddenly decreases when 20CS-1 is not energized.

It would appear from the description that the speed detection logic is intermittent--if indeed 20CS-1 is chattering. The diesel should be at or near a constant speed--and remaining at the same speed--during the break-away and the initial acceleration. Because the diesel can't overshoot when it's transmitting torque; the jaw clutch and the inertia of the turbine-generator shaft are keeping the jaw clutch teeth engaged--if the diesel is at or near rated speed (for cranking).

When the diesel is warming up and during it's acceleration the hydraulic ratchet is "ratcheting"--it's going through a forward stroke, and when it reaches the end of the forward stroke if the shaft hasn't started turning (broken away from zero speed) it has to go into a retraction stroke until it can go back into a forward stroke again. USUALLY, the shaft will break away before the ratchet has to go into a forward stroke. And once the turbine-generator shaft breaks away (starts turning), the ratchet stops--and at the same time, 20CS-1 is de-energized, again because the torque being transmitted by the diesel starting means running at near full speed should be more than sufficient to keep the jaw clutch engaged.

I believe 14HR is the logic signal that shuts down the ratchet and de-energizes 20CS-1. So, if 20CS-1 is chattering when the turbine-generator shaft starts turning then it seems there's something amiss with the speed detection circuit.

And, if the jaw clutch is disengaging-based on the information provided--then it's not because the diesel is "overshooting" it's because the diesel is slowing down for some reason. And it would have to slow down pretty quickly because the torque converter will keep transmitting torque for a couple of seconds or so even after the diesel starts slowing down.

You may have problems with the jaw clutch/hydraulic ratchet mechanism when the unit is reaching zero speed--but <i>based on the information provided</i> it's probably not related to the starting issue.

I presume this isn't a new machine, and that the Mark VIe is an upgrade to the original or even the previous control system--AND that the unit has been started successfully for decades and that this problem has just recently started. Would you please confirm or clarify? I failed to ask the most important question when a problem arises on a unit which has been operating fine: What has changed? Or, When did this problem start--after a maintenance outage, or after a trip from load?

After looking at the data in the "video" it's pretty obvious the torque converter isn't transmitting the torque required to accelerate the unit smoothly once the ratchet breaks the unit away from zero speed. The diesel speed won't be perfectly stable as it's accelerating at basic the same time the ratchet is trying to break the shaft away from zero speed. The governor--if one can call it that--on the diesel isn't really trying to maintain a speed, it's just holding the fuel rack in a particular position which roughly corresponds to a particular speed.

But, if the torque converter isn't working properly and therefore the turbine-generator shaft speed isn't smoothly accelerating from zero speed after the "bump" from the ratchet--then 20CS-1 is going to "chatter" when it shouldn't. And that also means the hydraulic ratchet is going to "chatter" since it's stopped/started when L14HR drops out or picks up at the same time as 20CS-1.

The hydraulic ratchet normally "pulses" as it's going through a forward stroke, then is pretty "smooth" during the retraction stroke, then "pulses" again as it goes through the next forward stroke. The little "chirp" at the end is not even unusual, but does seem a little more "louder" than normal. When the ratchet mechanism completes a forward stroke and then completes a retraction stroke it starts another forward stroke--to try to ensure the jaw clutch remains engaged between operations (the three minutes between forward strokes during Cooldown operation). I've always believed that's what causes the "chirping" sound at the end of a normal ratchet operation--the switching from retraction stroke to forward stroke and the sudden build-up then decay of pressure as the ratchet both the ratchet pump is shut off AND 20CS-1 is de-energized.

I still believe the ratchet engagement problem during the start of Cooldown and this starting issue are not entirely related--and not caused by the Mark VIe. Unless someone has recently changed the turbine shaft speed pick-up gaps.... or these problems started after the commissioning of the Mark VIe....

The Mark VIe passive speed pick-up inputs are EXTREMELY sensitive and have been known to "detect" speed when the jaw clutch closes and when the ratchet is going through a forward stroke. So, if this "problem" just started after either the Mark VIe was installed or the speed pick-ups were recently replaced or re-installed or re-gapped to something less tha approximately 0.050 inches the "over"-sensitivity of the Mark VIe speed pick-up inputs could be part of the problem.

But from the data it seems like the torque converter isn't working properly to keep the turbine-generator shaft accelerating once it is broken away from zero speed by the torque converter.

It may well be more than one or two problems--and often things like this are.

Please write back to let us know the answers to the questions above--and how you fare in resolving the problem(s)!!!
Dear CSA,

Thank you for your reply.

In fact, the problem definitively started after installation of MarkVIe. But frankly speaking, it was around 2 Years after that I noticed the problem of DEngine overshooting, and the problem of delay in Jaw clutch engagement after stopping of the unit & reaching 14HR started few months back.

This gas turbine was commissioned in 1987.

THe Gas Turbine had in recent years experienced around 10 trips, but this was mainly during start up or at full speed to load. This happened because of a problem in the Aux LV source change over CB and this has been cured. The unit almost never tripped from full load.

However, I have a few points to clarify.

You are right, in fact we face "parasistic" / noise pick up during operation of the gas turbine. Actually, during running of the GT# at part load OR full load, we have observed a parasitic signal being induced thg the diesel engine speed pick up. I had discussed this problem with GE Controls & I was told that this was due to the fact that the earth connection of the MarkVIe panel is connected to the main/common earthing grid of the GE unit & I was recommended to connect the MARKVIe to a separate, dedicated earthing system. Unfortunately, this is something difficult to implement of our site.

Secondly, regarding the Torque Converter, I understand that if it was having a problem with it, we would probably find a problem with the disengagement during start up. In fact, during commissioning, the starting clutch was disengaging at 68% speed. Nowadays it is disengaging at 69.8% speed.

Thirdly, I wish also to emphasize that the overshooting of the Diesel Engine is an intermittent problem. Sometimes It can happen during a cold start. After you stop the unit, and carry out after 3omins,a hot start, it will behave ABSOLUTELY normal. Sometimes it will start normally during a cold start and afterwards overshoot during a hot start.
FOR THIS PURPOSE, I AM PROVIDING the following links which shows a completely normal start of the unit. Hence in my opinion, a problem with the torque converter would have resulted into the overshooting of the diesel engine for all starta. (no need of download)

And lastly, I will try to carry out tomorrow or on wednesday, a start up of the unit and include the Toolbox trend, some additional data. I will see if I can remove the clutch cover & make a video movie of the start up. I think this will enlighten us further in resolving this problem & also is a good case study for other frame 5 users of this forum

niz 123

I'm not buying the ground thing GE is trying to sell. The main reason for separating safety earth from instrument earth is to provide isolation primarily from damage and upsets as a result of lightning strikes. Of course if there are grounds from motors and other electrical equipment and machinery on a single earthing system that can have a negative effect on the control system if it's not properly designed.

But improperly constructed instrument earthing systems (too small; damp soil; close proximity to safety earthing systems; etc.) can cause serious instrument and control system problems. But worse, few electrical contractors and few electricians and fewer plant designers properly design and connect instruments and control systems to instrument earth systems, and frequently interconnect devices and instruments to the wrong earthing system. "Earth is earth," or so I've been told many times by electrical supervisors and electricians (usually when they're in a hurry). And that's NOT true for separate safety- and instrument earth systems. And that's where the lack of training and experience and skill make--or break--the job for years to follow.

Proper wiring practices, proper wire type selection, proper level separation and proper shield drain wire termination can correct a lot of noise problems and control system nuisance issues. But when poor construction practices aren't followed--mainly improper level separation--that's VERY difficult to correct after the plant has been commissioned and is in commercial operation without a lot of analysis and time and re-routing of cables--all of which can be very costly.

There are tens of thousands of plants and control systems that use a single earthing system for instrument- and safety earthing that run just fine. They were built properly using skilled labor with experienced supervision and sound design practices. All of which are in short supply in the modern world. Those struck by lightning are more prone to tripping and nuisance problems, but two earthing systems are NOT required for proper Mark VIe operation.

As for your comments noise on the diesel speed pick-up wiring during normal operation of the gas turbine (when the diesel is not running)--that's only really possible when the diesel isn't running because when it is running it's speed pick-ups are generating a solid signal which is not affected by induced electrical noise.

Have you confirmed that twisted, shielded pair wiring is used for the entire circuit length from the pick-ups to the Mark VIe input terminals--and most importantly that the shield drain wire(s) is(are) properly terminated along the entire length of the circuit???

Now, on to the torque converter, it can slip at high torque transmission (such as during break-away) and still keep spinning at the upper end of turbine shaft speed range when the jaw clutch is still engaged. There isn't much torque being transmitted as the speed of the turbine side of the jaw clutch over-runs the speed of the torque converter side of the jaw clutch. (I think many of us are guilty of saying the torque of the turbine side becomes greater than the torque converter side of jaw clutch--which probably isn't technically true; it's more about the speed differential.)

The way the turbine speed is fluctuating when the shaft speed is very low just tells me that the torque converter is slipping. That's not likely noise on the turbine shaft speed pick-ups; it's more likely the turbine speed is actually fluctuating. But the turbine shaft speed pick-up wiring and shield drain wire termination should be checked.

And, it could be that when the oil is cold and the torque converter is cold that there is slippage, but when it warms up the slippage is greatly reduced.

You seem to have exhausted just about every other possibility--with the exception of changing the ratchet self-sequencer (VH14-1)--so since troubleshooting is often a process of elimination the torque converter is looming larger as the cause of the issue.

How long has it been since the torque converter was refurbished? And, there have been two threads on about refurbished torque converters being bad immediately upon reinstallation--the point being that not every refurbishment is equally good. Or lasts as long as every other refurbishment.

Please continue to keep us informed!
Dear CSA,

Based on the various possibilities of root cause, that was discussed in this forum, we decided to carry out new start up tests of the frame 5 unit in two situations.

First was a cold start
Second was a hot start

Results have now cleared out most of the questions.

In fact, the problem is a too abrupt acceleration of the diesel engine during the cold start. The ratchet remained into the retraction cycle and could not shift into the FORWARD cycle, to BUMP off the rotor. In fact, if you carefully analyze the plots from the cold start (l33hrf and L14HR), the cranking off of the rotor has been realized solely by the diesel engine & torque converter. This explains the speed drop.

In case of hot start, you will note that the diesel engine takes a little longer to accelerate & is smoother. Hence this gives time to the ratchet to complete its retraction cycle and enter the forward cycle to bump off the rotor.

At this stage, I can eliminate the possibility of torque converter slip and rather concentrate on the diesel engine which itself may have a cold start issue.

I leave it for you to appreciate. Please watch the video`s carefully & focus on the diesel engine acceleration and the moment the shaft starts to rotate.



I know need to introspect the possible problem with the diesel engine. could it be a problem with 20DA or the governor?

Thanks for kindly advising
I think you are stressing a bit too much on the difference between hot vs cold starts and the results. I've seen machines which can break away without the ratchet operating, it's pretty marginal on a diesel start. I am guessing that this is a 600 HP Detroit not a 500 HP Cummins diesel? On your cold start there is definitely a ratchet problem, on the second ratchet cycle you don't get a full forward stroke so the ratchet will not get a reversing signal and will stall.To be honest, both starts look OK but not perfect.

As for 20DA, it must be working OK or the diesel would not accelerate at all, the governor is a more interesting question. Many failed starts are caused by low diesel speed, this can all be checked at standstill by idling the diesel and forcing the 20DA to see what diesel speed you get, if my memory is good it should be around 2200 rpm but check your Control Spec.

When did the diesel last have a top overhaul?

I completely agree with glenmorangie about the undue emphasis on hot- versus cold starts. It's pretty unusual for machines with electric motor starters to be able to break the shaft away from zero speed without the assist of the ratchet, but I have seen diesel starter motors be capable of doing it without a ratchet assist--but it's still not the norm and the ratchet assist does help with breaking the shaft away from zero speed faster than without the ratchet assist. I've only seen the Cummins diesels on Frame 5 machines (an inline cylinder arrangement--not the v-arrangement shown in the first video).

I don't believe the diesel "overshoot" is the cause of the problem. It's a symptom of something else. I can't explain why it has an "overshoot" (your term--not mine) during a cold start but not during a hot start--that may be something to do with the diesel engine fuel rack/governor mechanism operating sluggishly when cold, and not so when warm. But, it's not causing any starting problems--in my opinion. As glenmorangie said, just looking at the video both starts were "good"--the shaft kept accelerating once it broke away from zero speed and didn't seem to go back to zero speed/fluctuate at all, not like it seemed to do in the previous trend you sent. (Actually, in my opinion it doesn't seem like the diesel needed the ratchet assist to break the turbine-generator shaft away from zero speed in either video....)

The fact that it's taking more than one forward stroke to break the shaft away from zero speed is unsettling to me. I'm not accustomed to it taking more than half or three-quarters of a forward stroke to break the shaft away from zero speed. Certainly having to wait for the ratchet to complete a retraction stroke seems excessive. I wonder if the diesel is building up speed fast enough, or if there is some problem with diesel fuel flow being restricted--say the fuel filters are dirty or have gotten wet from water in the diesel fuel.

Even if you're correct in your assumption that the diesel is "bumping" the shaft during a retraction stroke if it breaks the shaft away from zero speed at all it should then be accelerating it to purge speed--and it seems to be doing just that, not at all like what the early trends you provided showed.

I think both videos are "okay" as far as the break-away goes--once the shaft starts turning it doesn't stop or fluctuate--which is what you/the first trend seemed to be saying happened. I'm back to that speed "chatter" possibly being caused by noise on the wiring from the turbine-generator shaft speed pick-ups.

I think the diesel acceleration from idle to near rated is a little slow. It could be a fuel flow restriction (as above), or it could be a sluggish diesel hydraulic actuator or a dirty solenoid valve or dirty oil in the diesel hydraulic circuit (some seemed to have their own self-contained hydraulic oil system; others used engine oil, I believe).

The engine in the video seems to be a Detroit Diesel. so a higher horsepower engine would likely be more capable of breaking the shaft away without an assist from the ratchet.

I also have the same comment about the first video--it doesn't look like the second forward stroke completed at all. The way the ratchet works--when on cooldown--is that when it is started it should already be in a forward stroke (keeping the jaw clutch halves engaged) and when it completes a forward stroke it then goes into retraction, and at the completion of the retraction stroke it then goes into a forward stroke again for a couple of seconds--this to try to ensure the jaw clutch halves remain engaged during the period when the ratchet is off (three minutes between strokes), and then the ratchet pump motor is de-energized. So, when the ratchet pump motor starts after three minutes it should already be in a forward stroke with the clutch halves engaged.

During starting, though, I believe the ratchet goes through a forward stroke, then a retraction, then a forward stroke, then a retraction stroke--with no pause as it's trying to assist with breaking the shaft away from zero speed.

It would appear there is, indeed, more than one issue at your site. There appears to be something amiss with the diesel acceleration and/or it's ability to hold speed "when cold." It's not clear why the ratchet didn't complete a forward stroke in the one video--could be because the turbine-generator shaft speed was "noisy" and L14HR was "chattering." Again, there was no evidence of "chattering" during break away--once the shaft started turning (whether it was because of the assist from the ratchet, or the torque from the diesel) it kept turning and accelerating. And the resolution of TNH and tnh_rpm in both Trends is too low to be able to see any bobble in speed.

20DA is just a solenoid-operated valve that changes state when energized. It's nothing complicated. You should be checking diesel fuel filters, diesel air filters, the fuel rack/hydraulic actuator mechanism, possibly even the diesel compression--though if it were low on one or more cylinders the diesel exhaust would be white. It should be dark, even black, when it's breaking the turbine-generator shaft away from zero speed--as the load is greatest then.
Dear CSA,

Thank you very much. Adding Videos and other information to the post, definitely helps a lot in finding root causes of problems & I would strongly recommend the forum members to do so as this will render the exchanges more interactive.

I am new to this forum but I am already confident that the knowledge you will gain and give here will not available elsewhere, specially in the GE training or even from mainstream consultants.

The diesel engine is a Detroit diesel engine.