Wednesday, January 23, 2008

Forward: Qlink Pegasus "Overheating" / Head Gasket

To read the entire story correctly, you will have to read the last post first. The entries are in reverse chronological order.

Use the links to the right to read the blog this, correct, way. (For some reason, blogger doesn't allow visitors to sort by date.)

This is the saga of my 2006 QLink Pegasus 150cc Scooters, purchased new- from an authorized QLink dealer, on September 1st, 2007. It has just now been fixed, by me.

I am the ONLY person who has publicly said they have a DEFINITIVE solution to these series of scooters, which have been described as having "overheating" issues. Rumor has it that about 20 of these scooters were replaced with new ones, very quickly, by QLink. These were the first ones to have trouble. Since then, I have seen no verifiable evidence to support this rumor.

This blog is a result of fairly careful note-keeping. It was something I knew I had to do within the first week of owning my new bike, because I was already starting to get screwed. My original dealer wouldn't answer his phone. A local dealer told me to go to hell, and that he wouldn't support the warranty. He, too, was an authorized dealer for QLink scooters.

I originally posted many of this in my own, closed, membership-only, forum. I did this for the ease of the dealer. I felt it would be simpler for them to view pictures and text together, than to receive email, and that they could do it from any computer. I had never had membership requests for the forum, but eventually had two requests from users in China. This were the first requests in the three years my forum has been up, and it happened a day after the title appeared: "QLink CFMoto Pegasus 150 Scooter."

I kept quiet about many of these details, because popular opinion suggests they have a habit of "voiding" the factory warranty. The owner's manual says that warranties will be voided if work is not completed by an authorized dealer. Well, an "authorized dealer" authorized me. Therefore, a representative of QLink and their products legally gave me permission to do this work without voiding the warranty.

I have tried to protect the dealer who made this arrangement with me. The owner seems like a good guy, the place appears professional, the prices seem fair. He went through the trouble of picking my bike up, driving it to his shop, and working on it for four weeks. I don't want him to get into any trouble over this matter, because I don't feel he deserves it, and I think his hands were tied by QLink to an extent. You will see that communication isn't their strong point, however, especially when plagued by a problem product like an overheating Pegasus. If I were in the market for a scooter, I would see him first. People tell me I'm crazy. People tell me they'd have thrown that bike through his front window. There's always some room for improvement, I guess.

Finally, after documenting MY solution to this head gasket, (note that I do NOT call it an "overheating" problem now,) I'm a bit angry. I lost probably five weeks of riding time- time I planned on saving money in gas over what my big SUV uses- since this started. I've been lied to- right to my face. I knew it, stood there and took it, knowing the other people wouldn't budge. But enough is enough.

Everything here should be considered "opinion." Any emails that were copied are true and complete. I have all email archived including the header files, so there is proof of who sent what, and when. All images are as they were taken, except where obvious notes or color keys were added. I have had a witness to the discussion regarding me doing warranty work. I have two witnesses who observed my repair. I have the opinion of two of the most intelligent mechanical engineers I've ever met. I have receipts, phone records, and more.

I have a growing network of other owners who have gotten the shaft- some far worse than me. These owners are, typically, those who took dealers' advice to keep riding their machine, despite a clear need for taking it out of service and repairing it. Some call this a conspiracy between QLink and dealers. Many facts indicate this could be possible.

My goal here isn't to slam anyone, but to warn others what they're getting into when they choose products that are imported and poorly supported. To warn others to what extent their $2400 investment could mean absolutely nothing for months on end, and to help those who have had these problems fix their machine.

I am fully open to participation in a class-action lawsuit against QLink. I will say no more at this point, except that you may contact me if you wish to do so.

This blog may quickly disappear. I have asked QLink for some things in exchange for information on how, exactly, to fix not only the head gasket issue- but to regain their reputation. I've offered to send them professional-level documentation on the process of doing so- ready for distribution to their dealer network. I've offered to tell the story of how some unscrupulous dealers seem to have given advice to customers that result in catastrophic or irreversible damage to their bikes- probably in the hopes that they would fail. From here, those dealers are absolved of responsibility unless the customers took them to small claims court.

At the time of purchase, very little negative information existed regarding the model, and the few internet discussions out there made the bike look good. One negative comment mentioned an overheating issue that a forum owner decided was "not credible," and banned the user from his forum. It was said that 'there are people out there just trying to make trouble' and that no one should trust this person's account of events. (Note that this forum also is a heavy supporter of QLink, and has their banner ads on the forum and main site. The owner, on more than one occasion, has said that he met the "general manager" of QLink, a "Mr. Johnny Tai.")

Nowhere in this saga was Mr. Tai available for discussion, despite repeated attempts to contact him.

Please note that much of what is presented here is speculative, in terms of the actual corporate policy of QLink. It is, rather, an account of actual communication with them, where it is referenced.

Feel free to post a response, or contact Tim at
qlink_pegasus [at] yahoo [dot] com.

Monday, January 21, 2008

Troubleshooting: Verify you've got a blown head gasket!

Much has been said about troubleshooting a blown head gasket. Some true, some not so true, and some are just impossible to come by for many people.

In the case of QLink scooters, it really doesn't behoove someone to go ahead and verify- even with 100% accuracy- a blown head gasket. Why? Qlink doesn't care. They will NOT send you a head gasket, despite a very clear indication that this is the problem. They are NOT in the business of getting people back on the road quickly. If you disagree, do a quick internet search for it and read the accounts yourself.

The way I looked at it was this: if I verified, hands-down and for sure that there was a head gasket leak, I'd still have a big headache arguing about all the stupid shit they forced me to wait for. I'd rather simply suspect the problem, give them the chance to straighten it out, wait forever to get parts, try them, and ask again... until it finally got fixed; than to know and end up waiting just as long. I know; I called them myself and spoke to them. Don't believe me? Try it yourself.


In the meantime, here are the checks you need to perform to have a reasonable indication that you've got a leak; at least in the same capacity I had mine.

  1. Does your scooter run hotter than the screw on the temperature gauge? Even in cooler weather? Does it blow coolant when you accelerate suddenly? Is your radiator coolant level low? (Like down to 1/3 capacity.) Is your recovery tank level low, despite adding more? Is your tank level high all the time? If you answer yes to these, keep going.
  2. Stop the scooter. Allow it to cool. Take off the front shroud using a phillips screwdriver. Remove four screws on the rear side of the front fairing, and two at the bottom by the fender. Take off the COOL radiator cap. Don't burn yourself. If the tank is low, add coolant to the top. Start the bike. Once warmed up, goose the throttle to 1/2 or 3/4 full, very quickly, and let off. Do it again. If coolant pushes out of the filler neck, consider yourself 90% sure that you've got the problem. Check it a couple of times. It'll throw a shotglass or more of coolant every time you do it.
  3. Remove the lower head bolt that is at the lower left hand side of the head. (LEFT = LEFT WHEN FACING THE FRONT OF THE BIKE.) You can get to this bolt without removing the valve cover. If it has ANY grey, rubbery, RTV like sealant on it, this indicates a scooter that was suspected of having leak issues at the plant. Perhaps they used it through all years, perhaps even after the fix; I do not know. What I do know is that it was on this scooter. Can you see this grey sealant around the hole? Oozing from the head/cylinder joint? Just another way to check.
  4. Did you drain the coolant? Did you see black flakes of paint in the coolant? If you get flakes from your coolant, can you bend them without them snapping in two? If the latter is the case, and they're stringy, rubbery, or do not break- this is the coating from the head gasket. It comes off as a result of exhaust gas and high pressure passing over it. Some of the coating is within the combustion chamber, before the seal point of the gasket. This will be expelled through the exhaust. But what you're seeing is the coating that was blown from the gasket and into the coolant, as hot, high-pressure exhaust gas cooked it off and blew it there. Seeing this should make you very suspicious that you've got a gasket leak.
  5. If you've changed coolant and flushed the system, and have run the bike for two hours or more, does the coolant still look nice and green? Does it look kind of stained? Brownish? This could be exhaust gas staining your coolant.
There is more to come.

What do I need to make this repair? General Introduction.

First of all, you need a general piece of advice. The single greatest gift that a mechanic can have is that of consistent and careful observation. Look for inconsistencies in what you see in front of you. Look for inconsistencies in sounds an engine makes, or in how hard one bolt might be to remove than another.

Machines are designed to fit together, and work together, as a system. When one part of that system fails, it can cause a chain reaction that is difficult to assess, understand, troubleshoot, and repair. When you're talking about opening up an engine, know first that anything you do will have a marked effect on the longevity of your engine when you put it back together. If you have a doubt at all about anything you're going to do, DON'T DO IT. Stop, and seek advice from someone you can trust, and someone you know could have experience in what you're doing. Even if you have to simply get an uneducated but simple second opinion from the neighborhood handyman.

Second, in the case of QLink motorscooters, what you're about to do will- in their opinion- void the warranty. It will not matter that you've successfully accomplished a repair- even when they themselves made you wait months without solving it. It will not matter your ability, knowledge, or even if you are a licensed mechanic. In my experience, and from talking to others, I fully believe they will void your warranty, just as they would void mine.

It doesn't matter that I am the only one to offer a definitive solution to the problem. It does not matter that I have a relatively high mechanical aptitude. It would not have mattered one iota if I had videotaped the repair for their approval, as I have offered to do so, and have been told right from the source- no. They wouldn't let me change a thermostat- something no major automotive company would never void a warranty for.

Just keep this in mind- you're now on your own. Make certain you have everything you need to complete the repair BEFORE attempting it, as parts right now are only available through a dealer, and that very dealer will have your warranty voided based solely on the part(s) you order from them.

For this particular repair, I would recommend that you fully troubleshoot your Pegasus prior to even considering repairing it yourself. This now said, I also recommend that you do NOT take the word of your dealer alone: if coolant is pushing out of the system only on acceleration, I'd call any dealer a liar for denying the possibility of a head leak.

Since troubleshooting is important, I will address this in another post.

For now, understand that most people who have done a bit of work on cars probably have 90% of the tools required to perform the repair. You will need a torque wrench that will handle 35Nm, 7Nm, and 10Nm. Don't worry yet- I'll translate that in the tooling post.

You could invest in a compression test kit, ($30-$60 on average,) to assess your scooter prior to repair. You may have to purchase a few specific metric sockets and tools, as not everyone generally has all of these in a general kit. A definitive list will come up soon, but for now- don't let tools discourage you. In a bind, you can always borrow them from someone, or rent them from an automotive store.

How can I believe Tim and Savage? Taking Measurements

Incidentally, here is the bolt from the head that goes in the bottom hole surrounding the timing chain cavity. The grey stuff is rubbery RTV type sealant. This is a weak spot- clean off the RTV well, hit it with acetate solvent, add some blue (medium) Permatex locking liquid (Or medium loctite) and reinstall.
















Savage taking measurement.

Protrusion / height of alignment sleeve above head flange surface.

This is the depth that MUST be in the rocker arm assembly, or the pin is too long!









The height of the sleeve above the mating surface of the bolt hole flange is 0.3695"

This means the rocker arm assembly counter bore MUST be 0.3695" or greater in depth.










This is the measurement on the UN modified sleeve, overall length.











This is the depth of the counter bore in the rocker arm assembly. WAY TOO SHALLOW!

THE PIN IS TOO LONG!

File or grind the sleeve down to a final measurement of around 0.465". Remember to put a slight taper on the outside of the pin. This may help center the counter bore of the rocker arm exactly where it should be.

How can I believe Tim and Savage? Cylinder Head Close Inspection!

These pictures were taken from 10x to 30x magnification, through a microscope. Initially, I was trying to identify what small black dots were on the face of the head. I discovered a load of pits and voids across the face of the head, in the combustion dome by the valves, etc. Most were not of a troublesome size or location. However, it should be noted that it is extremely risky to take a head like this, and re-flycut it to remove material in a warped-head condition. This is because you could potentially open up a big void- it's like trying to cut away at the opening of a cave. You could make matters far worse.

The surface roughness and flatness of the surface were very good on this head, in my opinion. Again, the flatness was within 0.0005" across the entire head surface, which is kind of surprising considering everything I expected to see.

20X magnification
Coolant passage opening, edge. The discoloration is burnt coolant and carbon. Note the fracture-looking pits. Further- and of great importance- note how the sealing feature of the gasket dug into the head.

In most areas of the head, this mark looks like a rope. In this area, and along the leaking side of the head, the mark looks like a continuous groove. This is a result of the movement of the head gasket when running. The rope or cross-hatched pattern of the mark is caused by the high and low spots of the tooling that cut the head. The flycutter cuts a series of circles across the head, these circles are alternating high and low spots in the aluminum, very small indeed. As the head gasket moved, shifted, and vibrated; the tool marks were worn down so far that they are no longer visible in the problem, or leaking, areas of the head. This is important to know when you inspect your head, as it will point out very readily where the gasket was actually loose.


30x magnification

Further pitting and apparent fracturing at the coolant passage opening edge.

Note again the depth of the head gasket seat groove in the head aluminum. It is not only continuous in this photo, but it is also significantly deeper.

Again, the discoloration is burnt coolant and carbon- not black gook from the head gasket coating.



Roughly 15x

This is a further view of the bottom of the timing chain side coolant hole. Tool marks are not visible in the head gasket sealing groove, caused by vibration and shifting of the gasket.











10x
And still further away.














30x magnification


This image shows a void under 30x magnification. Again- this could be a real problem for some heads- causing leak paths, potential starts for fractures, and could make flycutting for warpage a dangerous and risky operation.

How can I believe Tim and Savage? Cylinder Head!

2006 QLink Pegasus 150cc Cylinder Head

This is the condition of the head prior to cleaning:

The red identifies the remaining RTV / Sealant on the problem side of the head.

The orange shows that there are no leaks going into the timing chain cavity. This is important, as if a particular bike did not seal well, coolant could be forced from the coolant passage into the timing chain cavity. This could mean coolant (and lack of lubrication) flowing into the timing chain, and upper end of the motor to the rockers, valve stems, and camshaft.

The yellow in the following image shows the actual leak path between the combustion chamber and the coolant passage. Again- the black residue found here is a combination of burnt antifreeze and carbon deposits. This is a huge image that will show, in an upcoming post, how the aluminum began to suffer fatigue as a result of the hot, high pressure gas, flowing over the sharp edge of the coolant passage opening.

The green circle shows where you should look if you click on the image and view it full size. The blue highlight did not do well in the photo, but if you look at the larger version, you'll clearly see when comparing it to the first image- there is a load of RTV in this hole.

Following a good cleaning, this cylinder head was placed atop a granite inspection table. Using standoffs that are within 0.0002" in height, the entire surface of the head was inspected for warping. This head measured ONLY 0.0005" different from corner to corner! Incredible! No head warping, and it had a few hundred Km put on it (gingerly) since the gasket first lost its seal!

How can I believe Tim and Savage? Head Gasket!

2006 QLink Pegasus 150cc
Head Gasket Analysis

Note that pictures are labelled directly with what they are showing. There are two sides to this gasket, and both are in this post.

Take special note to the shiny area surrounding the leak path into the coolant passage. This is the passage closest to the timing chain cavity. Also notice the stringy appearance of the black coating as it comes off of the gasket. The single reason why this coating came off like this, is due to high-pressure, hot exhaust gas, continually passing over the gasket. It literally cooked the coating off.

Now, understand that the coating will come off in other areas due to combustion, but all of this coating should have been expelled via the exhaust valve and exhaust pipe; NOT via the coolant.

This black coating is what I found at the very beginning of the ordeal, when I was flushing out the coolant!!!!!!!!!!!








How can I believe Tim and Savage? Cylinder!

...huh? Proof? OH!

Here's the proof.

Here is the bare cylinder. Sorry for the poor pic. In the images that follow, you'll see evidence of RTV/sealant that was applied only to the left hand side of the cylinder face. You will also see the leak path, which is very visible here, from the cylinder to the coolant passage / coolant jacket.

The blue area shown here is where RTV / sealant remained on the face of the cylinder housing. The areas where it is not visible are due to bad photography, or that they have stuck to the gasket upon removal. The sealant is a rubbery, RTV-like film. It is thin in most spots, grayish in color, and had to be carefully scraped from the cylinder during repair.

The following image shows a pink area that represents the very visible leak path shown in the first picture. Though there is a lot of black colored residue on the cylinder in this example, most of the black residue is from the gasket coating. However, in the area denoted below, this darker color was carbon deposit- completely different in composition than anything else surrounding the cylinder and water jacket, except where the gasket is normally not sealed. In this area it is identical.
Please read the posts that follow for more examples.

What caused the problem in the first place?

  1. CFMoto applied an RTV or Sealant on the cylinder and head in an attempt to remedy a known head leak:
    1. The RTV was applied only from the centerline of the cylinder, to the left side of the head and cylinder housing. This means it wasn't a "for good measure" fix, it was intended to fix a specific problem, or for a specific purpose.
    2. The RTV is on the same side of the cylinder that the leak occurred.
    3. The RTV is on the same side of the cylinder as the inherent assembly flaw or engineering flaw.
    4. The RTV applied to the head-cylinder-gasket junctions was counterproductive. It caused more problems than solving them.
      1. The RTV was put on sloppily, and oozed into the lower timing chain head bolt cavity.
      2. This RTV prevented the proper torque being applied to the lower left head bolt, and allowed torque to relax in use.
      3. This may be an individual scooter's issue, or could be commonplace. In any case, this did not solve the problem, and in my case it made it worse.
  2. The Stock "acorn nuts" have adequate clearance for the cylinder studs, due to copper colored spacers atop the rocker arm assembly. However, the internal friction between the nuts, the spacers, and the studs is such that prevailing torque precluded a proper final torque value, on the order of 30% of the required torque value.
  3. The Cylinder bolts are weak. Because of the above condition with the stock hardware, the internal friction adds shear stress (twist) to these weak studs. Under torque called out in the CFMoto Echarm service manual, the bolts torque forward something on the order of 90 degrees with the stock "acorn nut" hardware. This means that when torque is applied and released, the tool will move 90 degrees forward when tightening, but will relax back 90 degrees when the wrench is released. It is possible that some units- even beyond the 2006- could experience gasket leaks due to studs stretching over time. But only time will tell.
  4. The left hand alignment sleeve is 0.064" too long. The protrusion of this pin above the surface of the head is something like 0.370". The cavity left in the rocker arm assembly is only 0.306". This means that the head gasket will NEVER become fully compressed. There is a slight lead angle on the sleeve, that may enter into the counter bores a little bit more, but regardless- this is the major reason why the head gasket leaks.

The long sleeve(s) caused the problem. The RTV made it squeak through inspection, but amplified the final problem. The weak studs and incorrect hardware make the entire affair impossible to correctly torque. All together, these three things MUST be remedied in order to properly address the cylinder head leak. Look forward to the next posts for a detailed process and procedure on how to do this.

What really happened here?

Here is my interpretation of the entire situation, from CFMoto manufacturing, to the owner(s) "Voiding their warranties" by continuing to ride the bikes... according to QLink.

SPECIAL NOTE: LEFT = LEFT WHILE FACING THE VALVE COVER OF THE ENGINE. (Like sitting on the seat backwards.)
  1. CFMoto was aware that there was a problem with the sealing of the head gasket in [certain] 2006, 150cc, water-cooled engines. These engines were destined for QLink, perhaps for the ECharm, but I have no evidence to support the latter.
  2. CFMoto attempted to fix the problem by adding a small amount of RTV or sealant to the left side of the cylinder, and left side of the head.
  3. CFMoto may have performed a QA (Quality Assurance) test on the engine prior to shipping, and it is feasible to assume that- if they did- the engine passed a seal test. I believe this could have happened simply because so many scooters were test ridden, and ridden following the sale, prior to mechanical failure and the coolant leak occurring.
  4. CFMoto failed, for some time, to properly identify three key problems that were the root causes for the cylinder head gasket failure.
  5. QLink may- or may not have- been aware of the conditions above.
  6. The customer test drove the scooter, bought it, used it, and it eventually failed.
  7. Some customers sought advice from their dealer.
  8. Word has it that many dealers attempted some minor troubleshooting and repair. To my knowledge, not one single dealer fixed the issue after it occurred.
  9. Some dealers advised their customer to keep riding the scooter, saying it would simply continue to run hot.
  10. It is my impression that these dealers had no clue what they were doing, OR, that they wanted absolved of all responsibility in the matter. They probably assumed that the engine would eventually fail, and that QLink would either fix it, or their warranty would be voided. In either case, the dealer couldn't lose, unless the customer took them to small claims court and won. It may be that QLink was not aware that this was happening, and saw, only, a customer with hundreds of miles on their scooter following declaration of a problem.
Some of what is above is speculation on my part, but the following post shows facts.

Things checked, observations, and conditions prior to repair

Please again note to read this factually for one of two bikes.

Symptoms and Observations:

Most all of these symptoms and conditions existed until the repair, at which time they were never again observed. All of these were communicated to dealers and QLink, both of which continued to push non-fixes to the base problem.
  1. Upon draining coolant from system to replace with new, small black flecks of paint came out of the coolant drain plug with the antifreeze. This happened the first three times the coolant was entirely drained while working on the system. At these times, it was assumed that the flecks were from inside the radiator fill neck, where the cap's gasket rubbed some of the radiator paint off. However, this is not the case, and though I didn't realize a very important part of the observation, I figured it out later. This piece of the puzzle was that, when the flecks were bent in half, they did not break. They were pliable, not brittle. This should have told me that they weren't from the radiator, because the paint on the radiator is brittle enamel-like paint. At the time, I wondered if the water pump was made of rubber, so I removed the cover to inspect for play, wear, or any other obvious visual cue. I found the pump, of course, to be made of metal.
  2. The problem came on without warning, when accelerating quickly under load. Specifically, going uphill and accelerating around a turn at 45 mph.
  3. Another symptom of the problem was when the radiator cap was off. Sitting still but running, I "goosed" the throttle quickly to 1/2 to 3/4 full range. As I did this, the coolant would raise in level violently, and pour over the filler neck.
  4. If the bike was sitting as in #3, and you GRADUALLY increased the throttle, the coolant became turbulent and airy, but wouldn't surge out and over the fill neck. When the throttle was released, it appeared as though the coolant level dropped a bit. I quickly realized that most of this was a result of coolant going up just high enough to overflow into the recovery tank, but not high enough to spill over the radiator cap neck.
  5. Following test rides where coolant blew out, when the bike cooled down after stopping, the radiator would only be about 1/3 full, and the recovery tank antifreeze level would be between the bottom limit line and 1/2" lower than the lower limit line.
  6. The bike would run normal if you did not goose the throttle, or demand a great deal of acceleration under load. In this condition, the fan would cycle normally, no coolant would blow out of the recovery tank when it filled up, and the bike would run about 3/4 of the full amount of the white zone on the temp gauge.
  7. No appreciable amount of smoke ever left the tailpipe. Not when goosing the throttle. Not when starting up. Not when throttle was snapped back to zero.
  8. Some steam / gas was evident at the radiator when pushing coolant out, but was deemed normal due to the apparent temperature of the coolant and ambient air.
  9. Exhaust smell, if any, was always masked by "new vehicle" smell. Exhaust gas smell was not readily apparent because it was a new bike.
  10. Exhaust gas analysis did not take place, because I couldn't readily find a source.
  11. No oil was readily apparent in the coolant, though it did appear stained fairly quickly. It turned from bright green to a more brownish color soon after replacement. I was not sure if this was a result of contaminates being cleaned out, or if it was, indeed, exhaust gas.
  12. The spark plug never looked any other way than completely normal.