Peering into the dragon's mouth, we look for clues as to why he squirtsfuel at high speeds.
First, the diagnosis (this usually happens on four-cylinder engines): Ifyou are having trouble with the engine running up to about 5,000 rpm ormore then flattening out with sputtering and popping, then the followingmight be your problem. To check, start by removing the air cleaner.Crank the engine and rev it up. If it is not backfiring through thecarb, then position yourself to be able to watch accelerator pumpdischarge nozzles with the engine revved up. If these squirters continueto spew fuel well after the throttle is opened, or when it is at asteady high rpm rate, you have properly diagnosed your problem.
The use of a Holley carburetor, either 500-cfm or 350-cfm sizes, onvarious four-cylinder engines has sometimes led to these problems.Often, the squirters work overtime even when they should not. They cancause the engine to run so rich that it won't even reach maximum rpm.After fighting this problem for some time, I have found the solution.
These are some of the things I used while looking for the cure for whatis called "pullover." You see a spring and ball, an antipulloversquirter, and the epoxied power valve. The last thing I tried was thesquirter with tubing attached. When it would continuously shoot fuel forseveral feet at 5,000 rpm, I was convinced "pullover" was not theproblem. It had to be pressure. This could only come from one place: theaccelerator pump.
The problem has been that the accelerator pump squirters keep on workinglong after the accelerator pump is wide open and should have stoppedpumping. This seems to happen to a greater extent up the rpm scale, say,from 5,000 to 7,000. I have watched as the squirters sprayed fuel intothe venturis at such a rate that the engine went flat at about 6,000 rpmand began popping and sputtering. All of this occurred in the shop withthe throttle held wide open. This action is typical of an engine that isdead rich. I found that if I fired the warmed-up engine and then let itidle for 30 seconds before giving it one quick blast to 6,000, it wouldturn a new spark plug black as coal in an instant.
These are the parts of the accelerator pump. Notice that there is a redrubber flapper valve in the pump cavity in the float bowl. There are twoinlet holes under this. With negative pressure in the chamber, theflapper opens to suck in fuel. The spring assures the diaphragm willretract when the throttle arm releases the pump arm. When the pump armdepresses the diaphragm, the flapper closes and fuel is pushed to thesquirters. In the case of the so-called "pullover," vibration is openingand closing the flapper, thereby creating an "automatic" pump.
At one point, I thought pulses in the intake tract were causing astand-off situation in the system. I tried very small squirters. I eventried several antipullover squirters. These are shrouded to preventairflow from pulling fuel out of the squirters. I made a 21/2-inch tallspacer in an attempt to add plenum area. I tried using a steel ball witha ballpoint pen spring in the well under the squirter. This was verycritical and not too consistent. Many of these things worked to adegree, but all were little more than a crutch.
I installed a squirter that had brass tubes for nozzles. I attachedpieces of small latex tubing 12 inches in length to them and cranked theengine. I held the tubing up and pointed it away from me. At 5,500 rpm Iwas treated to the sight of fuel shooting out over the front of the car.No way was this pullover.
Next, I disengaged the throttle arm that actuates the accelerator pump.The pump should not move at all now. The same thing happened--rpm plussquirters equaled copious amounts of fuel shooting out. This was to thepoint that the engine just didn't want to run at high rpm, and even at alower rpm it would go rich to the point of limiting power.
This washer can solve the problem. This is a small Honda valve coverhold-down. It is 0.825 inch in diameter and has a 1/4-inch hole in thecenter. It fits nicely inside the spring and over the rivet in thediaphragm. You should file a tiny groove across the rubber face of thisspacer. This prevents a vacuum from pulling the flapper out.
Another experiment in controlling the rogue fuel supply was to plug thepower valve (I used epoxy to shut an old one). I then used a small No.14 squirter. The idea here was to use the squirter to supply fuel forthe top end. The jets were then dropped from 71s to 54s for themid-range. This did show an improvement. With some tweaking, it made thesystem manageable but didn't give the fuel mixture control I wanted forgood performance. Needless to say, the tiny squirters couldn't provide athrottle response without a stumble. Patting the throttle several timeswould do the job, but that was awkward in traffic.
During all of this, I used a number of different manifolds and managedto cross-pollinate several carburetors. I sent the carb back to BobOliver at Competition Carburetion and asked if he would straighten outall my misdeeds. He did. I now had a carb I was sure of. It had 71 jetsand a 2.5 power valve, along with a No. 21 squirter.
Now for the cure. After all the foregoing, the culprit could not be anintake pulse. Neither could it be venturi velocity. Nor was it possiblefor the float bowl to be pressurized in some form. The problem had to bein the accelerator pump itself.