"Mike, thanks for your advice.
"Mike, thanks for your advice.
I have had my 1999 Honda BF50A for 5 years and this "drop off of RPM" has occurred once in a while for all the time I have owned it. The Internet searches I have done have all shown such "drop offs" to be fairly common. Poor guys are searching for "demons" as mentioned.
You are right, air leaks in fuel lines, debris in fuel lines or carb jets, loose electrical connections, etc, all could cause this problem. So, over the years, I have rebuilt my carbs twice, and not just a quicky, but taking them apart on a lighted workbench and studying every nook and cranny with magnifying glass and compressed air and cleaner until I understood what every orifice did.
I replaced all gas line fittings and added a gas/water/separator/filter. I checked all electrical connectors, opening them and making sure they were good connections.
The Honda OEM tach has no adjustments on it, just the 5 wires, 2 for the light, two for the circuit board internal power and the one gray one for the signal.
Reading my Honda shop manual very carefully last night, I realized that I do not have a CDI at all but rather, an ICM (ignition control module.) Not that it makes any difference I suppose, since the manual still has about 100 resistances to check for the ICM.
I did do the check you mentioned on the spark plugs but I do not understand your comment about the kill switch. My understanding is that the kill switch will kill the spark by sending a ground to the CDI or ICM, so how can the signals be checked when the kill switch is pulled? What I think is needed is make sure the plugs are grounded when you check for spark. I did not do that when I checked for spark and I fried my voltage regulator. Now I just got to thinking that maybe I fried more than that. But that doesn't quite add up since I have had the RPM drop off problem off and on for 5 years.
But, I am rambling. Anyway, I happen to have a Fluke #83 DVM with a frequency setting. I hooked the meter to the gray and ground wire at the tach by skinning the wires and ran the engine this week with the garden hose on the lower unit. The drop off in RPMs happened to occur at about 3800 rpm,, and at exactly the same time the Fluke showed the same drop in RPMs, so that means the tach is fine and I am left with the original problem and it is somewhere else to be found.
By the way, the frequency on the Fluke had to be multiplied by 20 to get the right reading. I figure that is because the Fluke measures Hz, or cycles per second, so first of all, you have to multiply by 60 to get cycles per minute. Then you have to divide by three due to the three cylinders.
My latest idea is to change props. I got two more props I will check today. I say that because inthe shop manual, page 2-15 (Honda # (61ZV300E3) under the heading "Engine speed does not increase normally" and "engine speed fluctuated" it lists the propeller as being the prime suspect. I have tried everything else, so why not try this propeller test.
If that does not work, I will try the 100 resistance tests on the ICM and if bad, bite the $$$ bullet and replace the ICM. I mean, the ICM, per page 17-3, has the ability to drop the spark to cylinders # 2 and 3 if the ICM thinks the engine has exceeded 6600 rpm. Just like the Prius nowadays, what if my ICM is intermittantly thinking it has exceeded 6600 RPM? Dropping two cylinders would definitely give me a RPM drop and rough moment with the engine, wouldn't it?
Mike, brain gas occuring..... I just realized that your suggestion is accepted.... with a timing light in-line to either # 2 or # 3 plugs, I could see if I loose spark to plug when I am experiencing my RPM rumble. I'll do the RPM test as well as the prop test and get back...."
