I have a 2 stroke fuel injected dirt bike that runs pretty poorly ar low engine speeds and low throttle openings. According to the internet these models can often have issues with the crankcase pressure sensor (really just a map sensor) going bad. Fortunately they actually use two identical sensors - the other one is just used as a reference to ambient pressure. The typical diy diagnostic for these is to swap the two sensors and see if the condition improves. I have done that but haven't had a chance to test ride. While I was removing them I wanted to try to do some more in depth testing so I hooked up my handheld vacuum pump and multimeter across the various terminals. This didn't yield much data at all. While some terminals did have resistance values of a few hundred ohms between them, there was no change in resistance with varying amounts of vacuum on any pair of terminals. Interstingly, the two sensors' resistances were different from each other.
All of the map sensor diagnostics I see through google involve using scan tool data (which I don't have an obd adaptor for currently) or backprobing the wiring. I was hoping to avoid that and see if I get a smooth resistance curve in response to vacuum chage in a controlled environment. I did learn that there are digital type map sensors that alter the timing of a square wave and basically require an oscilloscope to diagnose. I wonder if my sensors are of this type?
For more info on my symptoms of running poorly; it starts well and idles just fine, above maybe a 1/3-1/2 throttle and low/mid rpm range (it's hard to tell which has more influence) runs really strong and smooth. Between just off idle and the low/mid speed rpm range it is really not smooth, it's kind of surging/hunting and sputtery. It makes it nearly impossible to ride smoothly at slower speeds. It definitely seems like it could be a sensor problem, map or tps.
Not likely to work unless you can get data for your specific revision of the sensor. The sensor is designed to supply voltage, and sensor manufacturers can and do revise the internal circuitry in such a way that the output voltage curve is the same but resistance measurements don't line up. I remember one time when Delphi shifted production of their 3 bar MAP sensors from the US to China; they completely redesigned the circuit at the same time. There was one company that made a product that relied on resistance instead of the voltage output and used this sensor - that change left them in serious trouble.
Now if you can apply 5 volts to the sensor and measure its output voltage, that would be a good test.
In reply to MadScientistMatt :
Interesting, I guess my assumptions about how map sensors work were entirely incorrect! I took it for a test ride with the sensors swapped and I think it actually did run better so that's an improvement. I would really like to know how these sensors die so fast. My bike only has about 20 hours on it I think it's felt bad like that pretty much since new.
Main things that can kill MAP sensors are excessive heat, oil getting in them, and poor initial quality.
In reply to MadScientistMatt :
That's funny that they would be vulnerable to oil since it's literally connected to the crankcase by a short tube! It would by design get constant exposure to fuel/oil vapor.
In reply to freetors :
On the Bosch MAFs on my '80s Saabs, the (platinum??) sensor wire is heated to red hot for few seconds when the engine is shut down, to burn off whatever might've glommed onto it while running. I assume others function similarly??
In reply to procainestart :
MAP sensors are just a pressure transducer, no hot wire funny business.
On a 2 stroke, I assume that they see some heavy duty pressure pulses when attached to the crankcase. That will probably age that sensor out a lot faster than the one used for baro correction.
In reply to Pete. (l33t FS) :
Well, a "P" looks like an "F" until I get new glasses next week. Or learn to pay closer attention... 😁
Most sensors are 3 wire. Power ,ground and signal. You need to have power to the sensor then you can read a voltage on the signal wire. Most sensors operate at about 5vdc max so you are checking to the hundredth of a volt.
Pete. (l33t FS) said:
In reply to procainestart :
MAP sensors are just a pressure transducer, no hot wire funny business.
On a 2 stroke, I assume that they see some heavy duty pressure pulses when attached to the crankcase. That will probably age that sensor out a lot faster than the one used for baro correction.
Yeah I guess that'll be the plan going forward; keep a spare or two around and rotate the baro sensor to the crankcase when they start acting up.
I do wonder if the ECU is running off an averaged map voltage or if it's calculating from the individual pulses?
In reply to freetors :
Depending on what it uses for a crank signal, it could be taking a MAP sample at a very specific engine position.
Or it could just be heavily damped (electronically or pneumatically) and reading the average, which probably works just as well and is less complicated.
On a Megasquirt install, I had a very spiky MAP signal that I damped with a vacuum coupling in the hose, that I glued a piece of carb cleaner straw into. When I updated to a newer unit that could do software based damping, I played with that, then ended up just going back to the damping orifice.
freetors said:
In reply to MadScientistMatt :
That's funny that they would be vulnerable to oil since it's literally connected to the crankcase by a short tube! It would by design get constant exposure to fuel/oil vapor.
Generally it's only a problem if the sensor isn't positioned to allow liquid oil to drain back.
