Any wisdom on KTM 790/890 Cam wear problems?

[KrisZ]

I wouldn’t blame emission regulation for this screw up. I hardly doubt the regulations go into such technical details as the amount of oil in the combustion chamber. That’s not how it works. They set an emission standard and don’t care how the manufacturer meets it, as long as it is met.

This is a KTM mess up, pure and simple. If they compromised engine lubrication in order to meet emissions, then that’s 100% on them. Personally I think they simply made a design mistake and didn’t follow up with enough testing.

The oil pump pressure relief fix, to me at least, could be a cheap ticket for the owners. Oil jets rely on oil pressure to deliver the proper amount of oil to the cams. If the relief opens too soon, the oil supply is heavily reduced and thus the oil pressure is also reduced.
Like @ZeeOSix mentioned, the oil pump may hit the relief pressure during cold starts very easily and that warm up period is enough to cause cam wear over time. This would be quite true for EU countries because a lot of them don’t allow idling to warm up the engine. You start and go, no warming up.

 

[ktmDuke890r]

I wouldn’t blame emission regulation for this screw up. I hardly doubt the regulations go into such technical details as the amount of oil in the combustion chamber. That’s not how it works. They set an emission standard and don’t care how the manufacturer meets it, as long as it is met.

This is a KTM mess up, pure and simple. If they compromised engine lubrication in order to meet emissions, then that’s 100% on them. Personally I think they simply made a design mistake and didn’t follow up with enough testing.

The oil pump pressure relief fix, to me at least, could be a cheap ticket for the owners. Oil jets rely on oil pressure to deliver the proper amount of oil to the cams. If the relief opens too soon, the oil supply is heavily reduced and thus the oil pressure is also reduced.
Like @ZeeOSix mentioned, the oil pump may hit the relief pressure during cold starts very easily and that warm up period is enough to cause cam wear over time. This would be quite true for EU countries because a lot of them don’t allow idling to warm up the engine. You start and go, no warming up.

I'll ask him if the way the spring upgrade works is via pressure relief. I suspect the spring replacement is a more fundamental solution than that in this engine design. One of the indicators I refer to is that the spring he is using comes from a version of the engine in another bike that has a different rated spring. I believe the spring in my Duke comes from the RC8C. He is using these in the Dukes and a different spring from another bike for the 790/890 Adventure bikes. He said he does this because the CAMs are different and create different pressures at the same RPM (my naïve interpretation). The springs come from bikes that don't have the pressure issues. This can also be clearly seen in the idle and at RPM pressure test results post spring change. And his isn't making this up. There is no manipulation or BS happening here.

At least that is how I understand it from a brief discussion and it is always easy to miss key points until you go over again in more detail, which I hope to do tomorrow when I pick up the bike.

Regarding the Euro 5 scheme, Honda withdrew the CBR600RR for more than just a few years due to the amount of work and expense they had to go through to pass that bike for Euro 5+. For some reason they decided the effort was worth it and now the EU and UK have the bike again for road use. That is my understanding.

 

[KrisZ]

I'll ask him if the way the spring upgrade works is via pressure relief. I suspect the spring replacement is a more fundamental solution than that in this engine design. One of the indicators I refer to is that the spring he is using comes from a version of the engine in another bike that has a different rated spring. I believe the spring in my Duke comes from the RC8C. He is using these in the Dukes and a different spring from another bike for the 790/890 Adventure bikes. He said he does this because the CAMs are different and create different pressures at the same RPM (my naïve interpretation). The springs come from bikes that don't have the pressure issues. This can also be clearly seen in the idle and at RPM pressure test results post spring change. And his isn't making this up. There is no manipulation or BS happening here.

If the idle pressure is affected by the spring change, that would mean the pump is running in bypass mode all the time, which would be quite unusual.

But who knows these days. Manufacturers try to squeeze every cent and if this means they can use the same oil pump in all of their engines and applications, I can see them doing it. Perhaps they actually have a pressure bleed valve in there for this very reason.

More clarification would definitely be appreciated.

 

[bulwnkl]

I find the wording used by the shop ktmDuke890r is using curious where they essentially say the engine was "designed" for higher oil pressure and flow. The "design" includes the oil pump, pressure relief system, and oil viscosity grade spec so this seems a potentially oxymoronic statement. I may not press them on it much, as one often says things a bit loosely for all kinds of reasons or conveniences. So long as a pyramid of incorrect assumptions are not made upon the errors of such a convenience, it needn't be problematic.

I do see a potential materials or workmanship defect in 890r's engine based on his shop's report, though. The high-rpm oil pressure may be below service manual limit, though the rpm at which said high-rpm test was performed was not displayed so I'm not positive. It also _appears_ from the post-spring-change pressure readings that the oiling system was in bypass all the time. That assumes, of course, that the shop made no other changes between the readings they took.

I don't buy for even a moment that these cams are eating themselves up over low-temp, essentially start-up, oiling. I _may_ be open to the argument that inadequate high-rpm flow caused by a too-low oil system bypass pressure is contributory, but no more than superficially so.

890r's shop says his cams are fine, yet they found presumably-out-of-spec (inadequate) oil flow to the cam squirters. That undermines the claims (or speculations) of failures rooted in oiling inadequacy, whether at idle or high-speed. The same shop claims, though, that the failed cams they've inspected didn't have metallurgical deficiencies. So it must be neither?

 

[ZeeOSix]

If the idle pressure is affected by the spring change, that would mean the pump is running in bypass mode all the time, which would be quite unusual.

It does sounds like the oil pump.is in pressure relief at idle if changing the spring changes the pressure at idle. If the oil pump relief is only controlled by a spring and not computer controlled in any way, then this means the pump has the unregulated volumetric output potential and doesn't sound like it's physically under sized.

The oil flow going to the heads seems a bit choked down. Wear like seen in the valve train has to be caused by lack of lubrication if all the materials are top notch and not a factor. Lack of lubrication in any engine is the main cause of high wear like seen in this KTM issue.

 

[ktmDuke890r]

Certainly an interesting subject. One other thing the mechanic said is that if there is wear on the CAMs it is always the left (if I recall correctly) which is second to receive oil. I'm guessing that sometimes both are worn but if only one it is the last to get oiled in the flow direction.

Clearly the point of contention here is the role that the spring plays, it's function. From the terms used, such as "pressure relief", it seems that typically this spring operates in a secondary role and not in a primary role. If it is secondary it makes no sense that it would fix the primary oil pressure issue. This is the question I will put to him.

 

[ZeeOSix]

Certainly an interesting subject. One other thing the mechanic said is that if there is wear on the CAMs it is always the left (if I recall correctly) which is second to receive oil. I'm guessing that sometimes both are worn but if only one it is the last to get oiled in the flow direction.

In the one YouTube video linked earlier, at around the 12 minute mark where the guy is cranking the engine but it doesn't start, there's actually a decent volume of oil coming out of the oil jets for the engine just at cranking speed. I'd like to see what those oil jets do when the engine is running ... they would be flowing a lot more for sure. All the components in the head look coated pretty good with oil. The guy is wrong when he thinks the valve train will be hurt if the engine cranks over 5 or 7 times before oil starts flowing in the head. There's always enough oil film to protect parts in a start-up like that. Every engine in the world cranks some revolutions before full oil flow is established.

I'm thinking if the left cylinder valve train wears first or wears the most, it's an oil volume problem. Since KTM claims it's only been reported that 0.5% of the bikes have this issue, I bet it's a manufacturing issue where oil paths are not open enough from manufacturing issues, and that's why not every bike does this. I'd bet the bikes that do it have some kind of issue in the head oil path system that chokes oil volume down before it get to the oil jets that squirt oil onto the cams and followers. On those specific engines, the oil jets delivering oil to the cams and followers is probably choked off and don't deliver enough oil unless the engine is revving pretty high. At low RPM, they may not be supplying enough oil to keep the cam lobes and followers properly lubricated. Lack of oil is a bad situation for any engine or moving machine.

Clearly the point of contention here is the role that the spring plays, it's function. From the terms used, such as "pressure relief", it seems that typically this spring operates in a secondary role and not in a primary role. If it is secondary it makes no sense that it would fix the primary oil pressure issue. This is the question I will put to him.

If the pressure relief of the oil pump is from a simple coil spring, then that's the only thing controlling the pump output pressure. I'm wondering what #15 ("Force Pump") is since #16 is identified as the "Oil Pump Unit". The "Oil Pressure Control Valve" must be where the spring is located.

 

[midobo]

It does sounds like the oil pump.is in pressure relief at idle if changing the spring changes the pressure at idle.

Reading between the lines, they also changed the oil from 10w50 to 15w50. Given the pressures reported, the readings where not made a full operating temp. At full temp, the pressure readings via the process outlined in the repair manual are basically right on the minimums listed.
Consider that the Motorex 15w50 is a step down in quality and price from the recommended 10w50 Power Synt/Cross Synt. The Formula line is a semi synthetic.
There are 2 oil pumps in the semi dry sump. A suction pump and a 'regular' pump. Some things get lost in translation from Austrian.
There's so much well intentioned half truths and speculations with these cam failures that it's literally taken on a life of it's own. The Facebook group dedicated to it is an echo chamber where you're banned if you have a sound grasp of mechanical theory.
As much as it's preached, replacing the stock cam chain tensioner with the Dirt Tricks unit does not cause the oil pressure sensor to fail. The Dirt Tricks unit is the same design as OEM.

 

[ZeeOSix]

Reading between the lines, they also changed the oil from 10w50 to 15w50. Given the pressures reported, the readings where not made a full operating temp. At full temp, the pressure readings via the process outlined in the repair manual are basically right on the minimums listed.

Those are both a 50 grade oil when hot, so not sure why they changed to a 15W-50 grade. All that does is make the oil a bit thicker at low temperatures during start-up and warm-up. And if the oil pump is already in pressure relief in a cold start, it's going to make the oil pump cut back the oil supplied volume to the engine. Oil pressure doesn't always mean more oil flow volume if the oil pump is in relief, and it's oil volume to the right areas in the engine that results in adequate lubrication. Inadequate oil volume to the moving parts is not a good situation and causes accelerated engine wear and damage.

 

[midobo]

One of the indicators I refer to is that the spring he is using comes from a version of the engine in another bik The springs come from bikes that don't have the pressure issues. This can also be clearly seen in the idle and at RPM pressure test results post spring change. And his isn't making this up. There is no manipulation or BS happening here.

The spring being used comes from the LC8 engines. They're from the 75 degree V twins used in the 1090, 1190 and 1290 engines. There's absolutely nothing in common between the LC8 and LC8c (790/890 inline twin) thus you can't draw the conclusions the spring will work the way it is suspect. Would it be fair to make a connection between a Ford Aerostar and a Mustang?

 

[ZeeOSix]

The spring being used comes from the LC8 engines. They're from the 75 degree V twins used in the 1090, 1190 and 1290 engines. There's absolutely nothing in common between the LC8 and LC8c (790/890 inline twin) thus you can't draw the conclusions the spring will work the way it is suspect. Would it be fair to make a connection between a Ford Aerostar and a Mustang?

If the spring is the right size to replace the stock spring, and the spring constant is correct to make the oil pump's relief pressure increase to some desired point above the stock level, then it wouldn't matter what other engine it came from, or if it was an aftermarket spring custom made by a vendor to increase the oil pump relief pressure.

 

[ktmDuke890r]

Picked up the bike and asked the mechanic how the secondary role of the spring impacts the primary function of the pump. He said the existing spring rating impacts the oil pressure directly by essentially activating too early, causing the oil to never reach optimal pressure. This means that the pump rating is reduced.

By increasing the spring rating the "relief" pressure is raised to a point where the oil flow reaches all parts of the engine.

Put another way: if blowing into a long hose going up to the ceiling that has water in it, the water will be pushed up to the ceiling if there is no hole in the hose. If a hole is introduced below the water in the air pocket (relief) the water may never reach the ceiling. The weak spring is that hole.

He had an Adventure bike in the shop getting a spring installed and it still had the oil pressure gauge attached. He said the owner had specified a 4 Bar idle pressure. I was surprised at that value and asked him "the owner specified that pressure?" -- yes, he said. He added that this pressure was unnecessary but it is what the owner wanted even after discussion.

I showed him my Chinese de-Cat link-pipe and he laughed at the fake Cat bulge in the middle. I mentioned that the elbows on the Arrow link-pipe were more curvy unlike the Chinese product which had a 90 degree elbow. He said he has the Arrow on his Adventure bike and it is fully flow tested. I agreed. He said still it will be a lot better than the Cat which causes a big problem with heat soak into the engine, adversely affecting oil temperature.

 

[ZeeOSix]

Picked up the bike and ask the mechanic how the secondary role of the spring impacts the primary function of the pump. He said the spring rating impacts the oil pressure directly by essentially activating too early, causing the oil to never reach optimal pressure. This means that the pump rating is reduced.

As I mentioned above, it sounds like the oil pump relief is purely controlled by the mechanical coil spring. There is no so called "primary and secondary" control. It's purely just the mechanical loaded relief spring that's controlling the pump relief pressure, which in turn determines the pump output flow volume, which determines the oil pressure when the other factors are held constant.

 

[KrisZ]

The oiling system diagram above refers to the oil pressure control valve, which to me means more than just a pump relief valve. The function of the relief valve is pretty much for safety and system protection from over pressurizing the system. This valve never opens unless very high pressures are hit during engine operation.

An oil pressure control valve's function is likely much more complex and it would operate under a variety of conditions. KTM probably decided to use such a valve instead of a two stage or variable pressure pump for cost savings.

This would make sense that changing the spring rate on this control valve affects the operational oil pressure right from idle speeds.

 

[ZeeOSix]

The oiling system diagram above refers to the oil pressure control valve, which to me means more than just a pump relief valve. The function of the relief valve is pretty much for safety and system protection from over pressurizing the system. This valve never opens unless very high pressures are hit during engine operation.

An oil pressure control valve's function is likely much more complex and it would operate under a variety of conditions. KTM probably decided to use such a valve instead of a two stage or variable pressure pump for cost savings.

It may just be their terminology - a valve can simply be a spring loaded bypass valve. If that "Oil pressure control valve" (#1 in the schematic) is not electronically controlled by the ECU in anyway, and it only has a coil spring in it to control the pump relief, then it's a purely mechanical controlled design. That's what it sounds like if these KTM techs are swapping out the relief spring to change the pump relief and output at the same RPM. I haven't seen anything that they are flashing the ECU to change how the oil pump relief works.

This would make sense that changing the spring rate on this control valve affects the operational oil pressure right from idle speeds.

That points to it being just a mechanical coil spring that controls the pump relief pressure. As mentioned earlier, it also shows that the oil pump is physically not under-sized if it will put out more flow at idle with just the pressure relief spring changed.

 

[ZeeOSix]

The oiling system diagram above refers to the oil pressure control valve, which to me means more than just a pump relief valve. The function of the relief valve is pretty much for safety and system protection from over pressurizing the system. This valve never opens unless very high pressures are hit during engine operation.

In this case, and in many cases on a PD oil pump using a simple spring loaded pressure relief spring, the relief valve may be cracking open at relatively low RPM, it then opens more as the RPM increases. That's how the simple spring controlled relief valves work on the oil pump on GM LS engines. Even though the relief is opening more with RPM, the pump is still sending more volume into the oiling system with increasing RPM because a spring loaded relief valve can't control the pressure and flow to a set constant level. You would need a more complex computer controlled system design in order to control oil pressure to a constant level with increased RPM.

Here's some flow and pressure graphs of the Melling PD oil pump performance, and you can see that even after the relief valve starts opening, the flow and pressure still increases with increased pump RPM. This shows how the pump performance changes when a different pressure relief spring is used in the pump. The purple line is with a relief spring that makes the pump essentially put out +25 PSI more.

 

[KrisZ]

In this case, and in many cases on a PD oil pump using a simple spring loaded pressure relief spring, the relief valve may be cracking open at relatively low RPM, it then opens more as the RPM increases. That's how the simple spring controlled relief valves work on the oil pump on GM LS engines work. Even though the relief is opening more with RPM, the pump is still sending more volume into the oiling system with increasing RPM because a spring loaded relief valve can't control the pressure and flow to a set constant level. You would need a more complex computer controlled system design in order to control oil pressure to a constant level with increased RPM.

Here's some flow and pressure graphs of the Melling PD oil pump performance, and you can see that even after the relief valve starts opening, the flow and pressure still increases with increased pump RPM.

I agree that some relief valves were designed to start partially open at lower RPM, but I don't think they will do it right from idle speeds. From the charts above that looks to be around 3k RPM.

That's why I'm saying the valve is probably more complicated on the KTM engine than a normal relief valve. This can be done without computer control.

 

[ZeeOSix]

I agree that some relief valves were designed to start partially open at lower RPM, but I don't think they will do it right from idle speeds. From the charts above that looks to be around 3k RPM.

Not all oil pumps and engine oiling systems are designed the same - so the RPM when the pressure relief starts opening may be all over the place. The oil pump needs to be relatively matched to the oiling system as a whole system. In the case of this KTM engine, from what I'm reading in this thread the idle oil pressure went up with just a pump relief spring change. Could be the oil temperature was different between tests which would skew the results - the oil viscosity needs to be held constant to see the true effect of changing the relief spring. I'd have to see a much more detailed test procedure reported to know.

That's why I'm saying the valve is probably more complicated on the KTM engine than a normal relief valve. This can be done without computer control.

I don't think it could be done without some kind of computer control unless it's maybe a mechanically controlled variable volume PD oil pump, which I don't think this is on the KTM. If the pressure control valve is indeed ECU controlled, it would be shown in the service manual. But the guy in the long YouTube video never mentioned anything about the oil pump being computer controlled.

 

[twX]

Here's from the factory repair manual. The relief spring isn't the issue as the pump doesn't achieve relief.

An oil pressure curve like that indicates that the oil pump relief valve opens well below 6,000 rpm. When the PRV is closed, oil pressure will be directly proportional to rpm. So if the oil pressure is 20 psi at idle, say 1,500 rpm, it would be around 80 psi at 6,000 rpm if the PRV was closed. The relief valve is certainly opening below 3,000 rpm, and it may even be relieving some pressure at idle.

 

[ZeeOSix]

An oil pressure curve like that indicates that the oil pump relief valve opens well below 6,000 rpm. When the PRV is closed, oil pressure will be directly proportional to rpm. So if the oil pressure is 20 psi at idle, say 1,500 rpm, it would be around 80 psi at 6,000 rpm if the PRV was closed. The relief valve is certainly opening below 3,000 rpm, and it may even be relieving some pressure at idle.

When the oil pump is not in pressure relief, the pressure vs pump PRM curve will actually be more like a squared function as shown in the Melling pump graphs in post 76. The flow volume vs pump RPM will be linear until the PRV opens. The oiling system basically acts like an orifice on the output of the oil pump. I've marked on the curves in post 76 where the pump relief valve starts opening. Look at the shape of the curve before the relief valve starts opening.