ONAN twin cylinder head temperatures

[JackC 616]

so if lower head temp = higher carbon buildup

and in the Onan, the front builds carbon faster, which stresses the rod in the rear cylinder

is it arguable that the FRONT cylinder should be made to run HOTTER to match the rear head, Yes, belt guard vents MAY cool the rear cylinder, but will additional cooling only serve to accelerate the buildup in the rear head?

I was wondering the same thing.

It may be best to run the Performer Series engines hard and hot to avoid carbon build up which appears to be one of the main failure mechanisms for those engines. On the John Deere 318 and 420 tractors, the Performer Series engines are more enclosed and they may run hotter with fewer failures.

[Save Old Iron 1,563]

Just make certain you investigate the agreement of probe to probe when making temperature comparisons. Always take the opportunity to swap the T1 and T2 probes at the meter to see if your delta remains the same. You would not want a $6 poorly manufactured thermocouple to bias all your readings and compromise this entire investigation.

[JackC 616]

Just make certain you investigate the agreement of probe to probe when making temperature comparisons. Always take the opportunity to swap the T1 and T2 probes at the meter to see if your delta remains the same. You would not want a $6 poorly manufactured thermocouple to bias all your readings and compromise this entire investigation.

Good points. I think I will pull out my Craftsman Infrared Thermometer next and compare the readings to the ones I took with the made in Hong Kong unit although I am impressed with it so far without having validated its readings.

Actually, the Infrared unit should be easier to take date with.

[Save Old Iron 1,563]

Jack,

I finally received a 14mm ring mounted thermocouple for experimenting with head temps directly at the spark plug.

I'm going to correlate this sensor with the sensor I have mounted to a calibrated Fluke 52 thermometer. Then I will do some cross checks from the same sensors in the same oil bath to the $23 meters we both purchased. Hopefully all will correlate well.

Here are a few pics of the initial setup. The head shown in the 3rd pic is a high compression LP head.. I'm not sure I will use this head or just a standard Kohler head while performing the Kohler K experimentation this winter.

I'm wondering it the additional 0.050 sensor thickness will change combustion characteristics given the spark gap will be further out of the combustion chamber?

[Don1977 604]

When you get what should be the average head temperature on flat head engines. I would be interested to know the head temperature for the over head valve V-twins, seems it should run hotter with the exhaust exiting through the head.

[JackC 616]

The ring configuration looks like it provides one of the best ways to secure the thermocouple to the engine.

I am not sure if mounting it under the plug and raising the plug a little will make any difference.

I personally would mount it under the head of a nearby head bolt before I would put it under the plug. A smaller diameter ring could be mounted under one of the screws that secures the engine tin to the head. It would be interesting to find out how much the mounting location affects the temperature reading.

For any application, once a reference temperature is established, I would think the deviation from that reference will be the important information in monitoring the status of the engine.

For any engine, the deviation could help to see how well certain lubricating oils are performing over others or how different fuels (octane levels) affect operating temperature. If the temperature rises a certain amount above the reference, then a serious problem would likely be indicated.

For a twin cylinder engine, once the reference temperatures are established for each cylinder and a reference delta established, the thing to look for would be that delta changing and going above a certain level. That could suggest an imbalance where one cylinder is not performing as it should thereby putting a load on the other cylinder that could lead to a broken rod.

Temperature readings may be able to provide early warnings before a failure. If that is the case, then that would be valuable to have.

[tiggercl 28]

Jack,

I finally received a 14mm ring mounted thermocouple for experimenting with head temps directly at the spark plug.

I'm going to correlate this sensor with the sensor I have mounted to a calibrated Fluke 52 thermometer. Then I will do some cross checks from the same sensors in the same oil bath to the $23 meters we both purchased. Hopefully all will correlate well.

Here are a few pics of the initial setup. The head shown in the 3rd pic is a high compression LP head.. I'm not sure I will use this head or just a standard Kohler head while performing the Kohler K experimentation this winter.

I'm wondering it the additional 0.050 sensor thickness will change combustion characteristics given the spark gap will be further out of the combustion chamber?

That spark plug seems a bit shorter than normal and equipped with a washer... I have replaced NGK and Champion spark plugs on two 520H with Onan P220G's, they seem longer and never had washers (see pics).

The replacement NGK plugs were not equipped with washers either.

Cheers,

Paul

[mattd860 161]

Hey everyone - I have a Mastercool infrared temp gun that i use when doing A/C work and I decided to use it on my 520 after mowing 2 acres with the 48" deck & vacuum bagger. The fins on the rear cylinder read approximately 9 degrees cooler than the front cylinder.

This sort of debunks the theory that the oil filter causes additional heat on the rear cylinder. This even debunks the theory that a vented belt guard is needed. So..... why does the rear valve seat go bad on so many of these models??? My theory is that many 520 owners rarely clean out the fins on these tractors (even a simple hose down after mowing) and somehow the grass/dirt/grime collects more on the rear cylinders than the front. IF the rear cylinder is cooler because more air is running past it than the front, this will explain why more grass/dirt/grime collects there rather than the front cylinder.

[JackC 616]

Matt saw 9 degrees cooler than the front on his ONAN 20 hp, and I saw 25 degrees cooler than the front in my ONAN 16 hp with 705 hours on it..

That is 2 for 2 that the back cylinder runs cooler and not would you would think.

There are two failure modes that we hear about, the valve seat coming loose and the rod breaking.

As far as I am aware it is usually on the back cylinder.

Matt, how many hours on your 520 engine?

[JackC 616]

Matt, what temperatures were you reading? On my 16 hp the front cylinder was averaging about 270 degrees Fahrenheit and the back cylinder was averaging about 245 degrees.

[mattd860 161]

Matt saw 9 degrees cooler than the front on his ONAN 20 hp, and I saw 25 degrees cooler than the front in my ONAN 16 hp with 705 hours on it..

That is 2 for 2 that the back cylinder runs cooler and not would you would think.

There are two failure modes that we hear about, the valve seat coming loose and the rod breaking.

As far as I am aware it is usually on the back cylinder.

Matt, how many hours on your 520 engine?

My 20hp Onan has 1200 hours on it and is all original. It is from a 1996 520H. I also have a 20hp Onan with 620hrs from a 1992 Onan that has a bad rear intake valve seat.

Matt, what temperatures were you reading? On my 16 hp the front cylinder was averaging about 270 degrees Fahrenheit and the back cylinder was averaging about 245 degrees.

My temps were higher. The rear was 296 and the front was 305. But I measured the temps after mowing with the deck/vac-bagger for 2hrs. Also, when I took the temps, the throttle was up about 3/4.

[Save Old Iron 1,563]

This sort of debunks the theory that the oil filter causes additional heat on the rear cylinder. This even debunks the theory that a vented belt guard is needed.

We cannot be certain if cylinder temp does or does not contribute to valve seat failures just from this limited data. It is interesting, isn't it? Once you start making measurements, old wives tales start to fall apart. Since you are not reading the difference (or absolute) head temps under actual operating conditions, its hard to tell how fast the heads heat and cool during a "workout".

What we need is a real time data logger to record an hour or two of actual second by second head temps - absolute and differential. Fortunately, the new micro controllers I'm working with make this very easy to implement. The data points can be stored on a small SD memory card and data can be easily imported into a spreadsheet format and then graphed as temp vs time.

http://www.ladyada.n.../lighttemp.html

Looks like a I have a winter project all planned out and Jack will be my twin cylinder guinea pig.

As far as valve seat failure, I still think the difference in expansion rates of steel vs aluminum can come into play if an operator fails to allow the engine to "come up to operating temperature". Slamming and jamming the PTO on to snow blow a driveway on a 0 degree morning after a 30 second warmup may make life hard on a valve seat.

Its good to get the conversations started. Some beliefs will be debunked, some will be reinforced. Very interesting.

[Save Old Iron 1,563]

The replacement NGK plugs were not equipped with washers either.

Cheers,

Paul

Paul, thanks for the info on the sparkplugs. I never really took the time to look closely at the position of the plug tip in relation to the bottom of a K cylinder head. We will see how this all evolves. :eusa-think:

[JackC 616]

We need to better understand the reasons behind the valve seat and rod failures on the twin ONANs, especially on the rear cylinder.

Temperature cycling could certainly be a factor for the valve seat failures. I bought a 1990 416H with a broken rod on the rear cylinder. When I took the heads off there was a lot of carbon build up which may or may not have been the cause.

I wonder if ONAN ever did a failure analysis report and published anything that we could get a hold of. Kohler did produce a failure analysis document that is available on the internet and they advise the engine mechanic to always try to determine the cause of the failure to avoid a repeat.

The 9 degree difference that Matt measured sounds a lot better than the 25 degree difference that I measured on my engine. With the limited data that we have so far it sounds like 275 to 300 degrees could be the normal operating temperature.

[JackC 616]

Here is an ONAN manual. See page 9 for service intervals. Sounds like running the engine under light load is not a particularly good thing.

Cleaning the cylinder heads of deposits looks like it should be done as frequently as 200 hours.

Note the 1,000 hour interval for adjusting valves on the P216, P218, and P220.

http://www.pneumaxcafs.com/manuals/Support%20Docs/Onan/Onan%20Ops%20%20P-216-218-220-224.pdf

[JackC 616]

Here is a little discussion on spark plugs for what it is worth. I have had good luck with Autolite.

http://www.casecoltingersoll.com/forum/viewtopic.php?f=8&t=6728

[JackC 616]

Here is a mechanic that likes the ONANs. We just need to figure out how to avoid failures on the ones we have.

I like this quote:

"I will admit the smoothness of an Onan is unparallelled"

http://www.mytractor...d.php?p=2131028

[dsholler 54]

Not being an engine guy, and not having an Onan I do not know if this makes sense, but being an analyst I do have a couple of observations here.

The temperature has always been stated as the culprit.. do we have any idea where this notion came from? Did someone do some research about it in the past that has been forgotten?

The observed differential does not seem to account for much at all. Even if we discount measurement errors, a 25 degree difference does not sound like much. I do not have an Onan manual, but other engine's spec sheets often give much larger ranges for normal operating temperature.

Are the front and rear rods and valve seats identical? (or other parts of the mechanism? ) Maybe the issue is a design or manufacturing one.

We also have two different things that happen, the valve seat coming loose and rod breaking. While the temperature/carbon theory would explain both effects, there is nothing that says that these two things are related. Maybe there are two issues that result in different failures?

This list is not based on any knowledge of anything, just in listening to the discussion here. it might be worth some interested party's time to make a list of the questions, and the group can collectively figure out how to answer them. Afraid I am not going to volunteer, because as I said, I do not have one of these.

[JackC 616]

dsholler, We are all speculating at this point. Some kind of early warning system for an engine failure would be good to have.

Maybe temperature measurements of some form will be helpful.

The temperature/carbon theory is all we have to work with at this point. I wish I could locate an ONAN engine designer to get inputs from.

I thought I read somewhere that engines built during a certain period were more prone to failure but I have not been able to substantiate that.

Back to the temperature/carbon theory, I wonder if carbon builds up faster on the aluminum head compared to the iron valve seat?

Could that cause a sufficient temperature differential to cause a problem?

The more i read the more i am believing that the de-carbon procedure is important for the twin Performer Series ONANs.

[JackC 616]

dsholler, The fact that the valve seat problem is not uncommon does suggest that there is indeed a design or manufacturing deficiency.

However, since it is a fact of life, what do we need to do to avoid it and is there any early warning system that we can come up with?

[Save Old Iron 1,563]

9 to 25 degrees F head temp differences could easily be accounted for by fuel mixture differences between cylinders.

[dsholler 54]

I am just speculating here, and was curious about a subject I know very little about.

I realize that saying it is a manufacturing issue doesn't help predict whether the problem will happen to you, but there may be a way to get some data that might help. For example, maybe if you could collect the spec numbers and whether the engines have suffered from this particular type of failure from as many people as possible on this forum (and on those that cater to other tractor types with the same engines) you might be able to find a pattern.

I also realize that the temperature test may be indicative, but (and here is where my naivete is showing) is there any other reason a valve seat could come loose other than temperature? some kind of mechanical stress?

[Hydro 130]

When my 416 rear valve seat came out of the block at 450 hours I was told the cross hatching for that cyclinder was also badly worn too so the conclusion was it overheated. I used a leaf blower to clean out the air intake but that is because I used to see it plug up in a matter of minutes depending upon grass conditions and the amount of dead grass left on the lawn. In the fall when I would semi mulch leaves the debris from the leaves would completely cover the air intake so I would stop and clear it...several times in fact during the cut of 1 1/2 acres. In hindsight my wife would also mow and she knew about how the flywheel cover would get completely covered but I have no guarantee she cleaned it as well as me nor do I know what temperaure fluctuations this would cause when it wasn't clear and over what period it continued but I will bet every Onan Wheel Horse owner who cut grass would have experienced what I am talking about because the left front wheel would pick up the dead grass and the engine would draw it in.

I have to wonder now just how long this condition could go on before the temps became problematic and if that is a contributing factor then why just the rear cylinder?

In the tests being talked about now they do not take into consideration an obstructed flywheel cover.

[JackC 616]

"In the tests being talked about now they do not take into consideration an obstructed flywheel cover."

Good input on what happened to your engine.

I have experienced the fly wheel cover blocked with grass condition often but I usually also check often to clear it when mowing.

If that is one of the conditions that causes overheating then the temperature monitoring we are talking about may be the early warning system that we are looking for. If we can come with a device that constantly monitors both cylinder temperatures and the delta and alarms when certain numbers are exceeded, then that may allow the operator to stop and investigate. For example if the monitored cylinder temperature exceeds 325 degrees an alarm goes off or if the delta exceeds a certain value like 50 degrees an alarm goes off.

I think there is real potential here.

[Save Old Iron 1,563]

"In the tests being talked about now they do not take into consideration an obstructed flywheel cover."

If we can come with a device that constantly monitors both cylinder temperatures and the delta and alarms when certain numbers are exceeded, then that may allow the operator to stop and investigate.

Very easy to implement using straight forward comparator circuits. The circuit board could be easily powered off 12 volts and hermetically sealed in a small enclosure to be tucked away in the hood stand area somewhere. The user could be given the choice to select a trigger temperature anywhere between 300 to 350 degree's F. Each cylinder could be set independently to different trigger thresholds if desired.

I'm not sure the "delta" would need to trigger an alarm. If any cylinder would go beyond 325 to 350 degrees, the alarm would trigger regardless of which cylinder it was.

The most important question would be what type of sensor to use. Most automotive thermistor based sensors are rated for around 340 - 350 degrees F. These sensors usually have 3/8ths pipe thread fittings. We may need to investigate a sensor held in place by friction or a small clamping device.

A second question would be how the operator is alerted to the overtemp condition. Lights? Beeper?