Fuel Pump Diagnostics ( ST1100 ) *

[KoTAOW]

(original web page here)

Fuel Pump Diagnostics

By Paul Kolbo:

Have you ever found yourself trolling along on your ST, enjoying the scenery, enjoying the ride, only to have the moment ruined by the feeling of a motor struggling and starving for fuel? Or have you noticed that your ST seems to be misfiring at higher RPMs while doing the ton? I have and, although it wasn't chronic enough to stop riding, it was something I always had in the back of my mind and wrote about on occasion.

The story starts in the year, 2000. On the way to WESTOC Durango, I suffered a stall condition on the highway between Ouray and Silverton. It was a hot day and we were experiencing altitude changes just prior to the stall. I was able to limp down to a gas station in Silverton to top off with fresh fuel and some roadside diagnostics. After disconnecting the fuel line to the fuel filter and dumping out the filter contents, the bike started up with a hard choke and ran fine the rest of the vacation. While in Durango, I bypassed the vacuum shutoff and replaced the fuel filter but was not able to pinpoint any obvious defect or plugging in either component.

Two years later in August of 2002, another hot day, the bike stalled again. Another session of roadside diagnostics did not reveal any major issue and after 15 minutes the bike fired right up and ran fine. In 2003 the stall happened twice on the way to WESTOC Moscow. Both times were in very hot ambient temperatures and rapid altitude changes. Curiously, the stalls happened immediately after descending a substantial altitude. This time I was able to determine that the pump was not pumping fuel by disconnecting the fuel line to the filter and running the bike off the choke. I was then able to make the pump functional by blowing back through the fuel line into the tank/pump. This led me to a theory that I was experiencing a vapor lock situation that would stop fuel flow. I reasoned this situation was precipitated by altitude changes and hot weather. While both did play a role in the final outcome, it wasn't bubbles or inadequate tank ventilation that led to the pump failure.

Fuel starvation on ST1100s is not all that rare. There have been plenty of stories of stalling problems. Discussions with other owners with similar issues led to the realization that something common might be the root source of these mystery stalls. These discussions wound up yielding a list of possibilities that could be eliminated by diagnostic techniques.

Possible Cause Diagnostic Solution
 
   
1.  Contaminated Fuel  - Replace filter, drain and inspect tank.
2.  Vacuum shutoff valve - Bypass fuel line from tank direct to filter.
3.  Tank Vent Blockage - Inspect vent line for blockage
4.  Fuel Cap - Remove cap and check airflow in both directions.
5.  Fuel Pump - Flow rate volume check, flow pressure check.

After spending the last three years eliminating all five, I was perplexed as to the source of my issue. I credit Rob Parker for promoting the idea that, while my flow rate met the Honda minimum spec, perhaps my pump was weaker than others, or intermittent. Rob rigged a low pressure gauge with a piece of fuel line with splices for easy installation in between the pump outlet and filter inlet. We then began collecting flow rate and pressure measurements on STs in our area.

Year   Flow Rate (15 Seconds)  Pressure (Fuel Cap off)
 
'92 500 ml 16.5 oz/in2
'98 na Erratic (This pump is a known failure)
'99 700 ml 20 oz/in2
'91 500 ml 17 oz/in2
'94 250 ml 11 oz/in2 (This is my '94)
 
The results from the last bike listed, my '94, are different from the others. Bear in mind that the flow rate still meets the Honda minimum spec of 650 ml/minute. (More on this later.) Ideally you would want to collect many more samples that would also include a fair number that exhibit the stall failure. We don't have that luxury at this time and, with WESTOC 2004 approaching, I decided to bite the bullet and purchase a new pump. (Service Honda - $273)
 
Note the pump pressure, not high at all. I could blow about 20 oz/in2 with my lungs.
 
'94 550 ml 18.5 oz/in2

It appears the new pump is more in line with our limited sample average. However there still is no evidence the old pump was bad. It flowed the fuel above the minimum spec until some condition (heat and/or altitude) would stop the fuel flow. So the next step was to analyze the old pump and see if we could find the "smoking gun" that causes it to stop pumping fuel.

An afternoon in Rob Parker's garage would yield the most likely explanation. First we verified the pump was flowing the same rate on the bench with an external 12-volt source. This essentially eliminated my ST1100 as the source of the weak pump (power source, pump relay or other). After further disassembly of the pump from the inlet screen, tubes and mounting fixture we confirmed the pump component, removed from the rest of the fuel pump assembly, was still weak. This eliminated the lines and filter inlet.

I made the first significant discovery during the pump diagnostic. While orienting the pump inlet to the water, I heard rattling from inside the pump housing. I could make this rattling come and go by changing the pump orientation. This rattling was not present when the motor was not running. The rattling sounded like one of the rotating parts (impeller or motor windings) was interfering inside. Despite the unsettling noise, I could not make the pump lock up.



The second discovery was not an accident. We felt heat was involved so we directed hot air from a heat gun into the pump inlet. We had the heat set to the low position and kept it about 3 inches from the inlet to avoid melting things. After about 30 seconds the motor would start to slow down and eventually stop completely. Dunking the pump into a pail of water would immediately revive the pump motor. Voila!!! I repeated this several times to prove the failure was consistent. We didn't have a temperature probe but I felt confident that it wasn't an unrealistic temperature. There wasn't any melting of plastic, nor did we impart any permanent change in the behavior of the pump stall. The failure was always about 30 seconds after directing the hot air into the inlet.

The final step is total pump disassembly. It took a series of hacksaw and Dremel swipes to separate the housing. It is worth noting that before the housing was cut open, I noted the impeller had perhaps a 1/16" of fore/aft movement (not lateral). Because the pump inlet screen is fixed, there is no way of inspecting another pump to see if this is normal. The pump internals did not yield any broken pieces but the magnets did have corners chunked off. The impeller is only about the size of a quarter and it didn't appear to have any worn or broken areas. Nothing about the internals yielded the clue to why this pump would stall when heated or why there was play in the impeller axle. I wonder if this motor wasn't missing a bushing or spacer to begin with.





Conclusion: This was a fun exercise and I am satisfied with our post analysis that leads to the following summary. My current theory is when the fuel tank heats, the pump can fail when the fuel reaches a very hot temperature. The orientation of the pump motor, that causes the rattle, might increase the probability of interference in the pump. Sufficient cooling, or blowing back through the pump, would allow the pump to function again. It's worth noting that the tank pressure would exceed 20 oz/sq in at operating temps when the gas cap is secured. Shallow throttle operation would allow the pressure to maintain a minimal pressure. Heavy throttle operation would deplete the tank pressure and lead to fuel starvation. In other words, the pump may have locked at some point while descending and fuel starvation wouldn't occur until ascending the next hill. This is why I took the pressure measurements with the gas cap off.

To those who suffer the occasional fuel starvation issues: Eliminate steps 1-4 first. When 1-4 is eliminated, check the flow rate first for step 5. I believe the flow rate/volume is directly related to the pressure. So if you're pumping around 500 ml/15 secs (remember, with the gas cap off) you're probably OK. If you find the flow rate to be very low, then check the pressure.

Thanks go out to Rob Parker who provided his knowledge and pressure gauge and John Oosterhuis for his pictures.