TOPIC:

Wookie58 wrote:

This seems quite bizarre, still a "win" is a "win" 

 I would suggest a piece of rubber mat between the MAF and the new battery as it appears the damage to the battery has been caused by the MAF rubbing ?


so after I read your post I started to analyze what I found further.  Yes the housing on the maf sensor was shorting to the battery but that was NOT impacting the potentiometer readings as the circuit board is isolated from the ground/metal plate it is mounted on.  So I decided to try and recreate the condition by putting some water on the top of the battery and it did the same thing with having voltage on its case but before I did that I started the bike and it was coughing again…. So my testing was going to be invalid. 

Once I get the bike warm and the engine temp sensor goes from 2.1kohm down to 350ohms, then install fresh non fouled plugs, it runs perfectly.  So what is happening is there is tool much fuel when at room temp, 75degrees here today, once it’s started then the plugs get fouled, some keep running and others stop sparking.  If I hold the throttle at 3000rpm for a few mins and let it cough the engine gets warmed up.  Then I swap out the plugs and hit start and the bike runs well.  The previous owner had 4 plugs under the seat and a spark plug socket wrench too,  he must of had similar experiences….  Now I need to figure out which part of the starting circuit is making the computer put in too much gas?  The recoil spring on the potentiometer door flap has been played with by me, so I know that is out of calibration.  I swapped the engine temp sensor with NOS part and it has the same result and same resistance readings so I don’t thing that part is bad. Started thinking of how to make a circuit at 350 0hms to start with, just to see if the bike starts and runs ok with that reading from the cold start…. Trying to prove that is a factor for to much fuel when cold.   This is still a work in progress.  

Rigging a "test" with a 350 ohm resistor is straight forward. Just disconnect the temp sensor wire and connect a 350 ohm resistor between the cable and ground (you might want to check with an ammeter first what the current flow is so you get the right value resistor) Also you should carry out a resistance check of the sensor from the disconnected ECU plug (as the temp sensor is an NTC (negative temp coefficient) any resistance in the wiring caused by corrosion will fool the ECU into thinking the engine is colder than it is and increase the fuelling accordingly.
This system is very basic relying only on the MAF, throttle position and temp sensor to control the fuelling. If you are happy all these are ok it may be time to get the ECU tested (you can only check injector duty cycle with an oscilloscope) 
Did you check the injectors in the manner I suggested and check their delivery rate ref the spec you found? as the quantity of fuel delivered is controlled simply by adjusting the injector opening time any issues with the injectors will be magnified during cold start when the injectors are open longer

Wookie58 wrote:

Rigging a "test" with a 350 ohm resistor is straight forward. Just disconnect the temp sensor wire and connect a 350 ohm resistor between the cable and ground (you might want to check with an ammeter first what the current flow is so you get the right value resistor) Also you should carry out a resistance check of the sensor from the disconnected ECU plug (as the temp sensor is an NTC (negative temp coefficient) any resistance in the wiring caused by corrosion will fool the ECU into thinking the engine is colder than it is and increase the fuelling accordingly.
This system is very basic relying only on the MAF, throttle position and temp sensor to control the fuelling. If you are happy all these are ok it may be time to get the ECU tested (you can only check injector duty cycle with an oscilloscope) 
Did you check the injectors in the manner I suggested and check their delivery rate ref the spec you found? as the quantity of fuel delivered is controlled simply by adjusting the injector opening time any issues with the injectors will be magnified during cold start when the injectors are open longer



I have not done any injector testing yet.  I did notice that the voltage on the eng temp wire starts at 7.1v then drops as low as 5.3v when the engine is continuing to run, then it starts climbing again to stay around 6.3 v. I was concerned if I used a fixed resistor for a test that the current would fluctuate with the voltage changes and possibly damage the ecu.  I have measured both old and new temp sensors when not plugged in to the ecu wire and they both come in around 2.0 to 2.1k ohms.  Do you know of anyone that checks these old ecu units any longer? I would love to just install it on a running bike and see if it acts the same or not.   I know there are some surviving bikes out there as I see them for sale.

Cullen wrote:

Wookie58 wrote:

I have not done any injector testing yet.  I did notice that the voltage on the eng temp wire starts at 7.1v then drops as low as 5.3v when the engine is continuing to run, then it starts climbing again to stay around 6.3 v -  I don't understand why you  are measuring voltage on the sensor wire unless you have found specs for this, there is no point in measuring something if you don't know what you are supposed to have (it simply causes confusion)
 . I was concerned if I used a fixed resistor for a test that the current would fluctuate with the voltage changes and possibly damage the ecu. 350 ohms is a known resistance value for running at normal temperature, therefore the ECU is designed for the current draw of 350 ohms for prolonged periods of time (remember current is determined by resistance and voltage) before you fit a resistor connect an ammeter in series with the sensor at normal operating temp then multiply that current reading by the voltage (12v to be on the safe side) to get the "wattage" as resistors are rated in both ohms and watts.
 I have measured both old and new temp sensors when not plugged in to the ecu wire and they both come in around 2.0 to 2.1k ohms. I would recommend rechecking that resistance reading from the ECU plug (with it disconnected) as any additional resistance in the wiring will cause over fuelling.
  Do you know of anyone that checks these old ecu units any longer?  Not sure where you are in the world, there are a few company's in the UK that do for around $80
I would love to just install it on a running bike and see if it acts the same or not.   I know there are some surviving bikes out there as I see them for sale. According to CMNSL this is a one year only part for the "G" model so they are going to be rare !! If  you come to the final conclusion the ECU is faulty (after everything else has been tested in line with the manual) it might be worth considering an aftermarket replacement. Things have come a long way in the last 40 years and something like a megasquirt ECU which is "mappable"  would be far superior for both performance and gas mileage

Wookie58 wrote:

Wookie58 wrote:

I have not done any injector testing yet.  I did notice that the voltage on the eng temp wire starts at 7.1v then drops as low as 5.3v when the engine is continuing to run, then it starts climbing again to stay around 6.3 v -  I don't understand why you  are measuring voltage on the sensor wire unless you have found specs for this, there is no point in measuring something if you don't know what you are supposed to have (it simply causes confusion)
 . I was concerned if I used a fixed resistor for a test that the current would fluctuate with the voltage changes and possibly damage the ecu. 350 ohms is a known resistance value for running at normal temperature, therefore the ECU is designed for the current draw of 350 ohms for prolonged periods of time (remember current is determined by resistance and voltage) before you fit a resistor connect an ammeter in series with the sensor at normal operating temp then multiply that current reading by the voltage (12v to be on the safe side) to get the "wattage" as resistors are rated in both ohms and watts.
 I have measured both old and new temp sensors when not plugged in to the ecu wire and they both come in around 2.0 to 2.1k ohms. I would recommend rechecking that resistance reading from the ECU plug (with it disconnected) as any additional resistance in the wiring will cause over fuelling.
  Do you know of anyone that checks these old ecu units any longer?  Not sure where you are in the world, there are a few company's in the UK that do for around $80
I would love to just install it on a running bike and see if it acts the same or not.   I know there are some surviving bikes out there as I see them for sale. According to CMNSL this is a one year only part for the "G" model so they are going to be rare !! If  you come to the final conclusion the ECU is faulty (after everything else has been tested in line with the manual) it might be worth considering an aftermarket replacement. Things have come a long way in the last 40 years and something like a megasquirt ECU which is "mappable"  would be far superior for both performance and gas mileage




 

i measured the resistance at the ecu for the engine temp. It was 2.575kohm.  At the joint near the engine it was 2.572kohms. A 0.003kohm difference so I don’t think there is any wire corrosion or excessive resistance in the wiring for that.  Ecu connections have been cleaned up too.   

 

i was measuring various voltages just to collect more data, whether it was useful or not begs the question.  

 

next will be the circuit current draw to determine the resistor sizing so I can run my test.  I need to get the bike out of the garage to do that but it is currently blocked in by a zx-11 being worked on by my son. Give it a few more days…

Or you could put in a variable resistor in.
Dial it up till the engine runs well.
Pull it out and measure the resistance and solder in a resistor with that value.
Or just leave the variable in place and use it to control your mixture.

Or skip the above steps--- run with hotter plugs---and accept your gas mileage may be a touch lower.
You may want to get your oil analyzed for presence of gasoline before you do this though.
You could have so much fuel spraying that your oil is getting diluted.
 

F64 wrote:

Or you could put in a variable resistor in.
Dial it up till the engine runs well.
Pull it out and measure the resistance and solder in a resistor with that value.
Or just leave the variable in place and use it to control your mixture.

Or skip the above steps--- run with hotter plugs---and accept your gas mileage may be a touch lower.
You may want to get your oil analyzed for presence of gasoline before you do this though.
You could have so much fuel spraying that your oil is getting diluted.
 

I would love to use a variable resistor, if anyone has a suggestion on which kind that would be great.   I have changed my oil because it had thinned out quite a bit and you could smell the gas in it.   Thx

Cullen wrote:

F64 wrote:

Or you could put in a variable resistor in.
Dial it up till the engine runs well.
Pull it out and measure the resistance and solder in a resistor with that value.
Or just leave the variable in place and use it to control your mixture.

Or skip the above steps--- run with hotter plugs---and accept your gas mileage may be a touch lower.
You may want to get your oil analyzed for presence of gasoline before you do this though.
You could have so much fuel spraying that your oil is getting diluted.


 

I would love to use a variable resistor, if anyone has a suggestion on which kind that would be great.   I have changed my oil because it had thinned out quite a bit and you could smell the gas in it.   Thx

Resistors (and variable resistors) are rated by "wattage" as well as ohms, you need to know the current flow to buy the right wattage resistor or it will just burn out. Also less than 350ohms will increase the current flow through a 40 year old ECU above is designed operating range

Cullen wrote:

F64 wrote:

Or you could put in a variable resistor in.
Dial it up till the engine runs well.
Pull it out and measure the resistance and solder in a resistor with that value.
Or just leave the variable in place and use it to control your mixture.

Or skip the above steps--- run with hotter plugs---and accept your gas mileage may be a touch lower.
You may want to get your oil analyzed for presence of gasoline before you do this though.
You could have so much fuel spraying that your oil is getting diluted.




 

I would love to use a variable resistor, if anyone has a suggestion on which kind that would be great.   I have changed my oil because it had thinned out quite a bit and you could smell the gas in it.   Thx

Resistors (and variable resistors) are rated by "wattage" as well as ohms, you need to know the current flow to buy the right wattage resistor or it will just burn out. Also less than 350ohms will increase the current flow through a 40 year old ECU above is designed operating range


i checked the current while bike was running and it was 7mA.  I am going to start another post with what I did today.

Tried a science experiment today.  Bike was at 68degF.  Engine temp sensor read 2.64kohm on its own.  Plugged it into the circuit and measured total resistance at 890ohms, then started the bike.  It coughed and sputtered as expected. Shut bike off and placed a heat gun to blow near the engine temp sensor.   Disconnected the temp wire and watched the readings dropping , as it got around 1.06Kohm on its own, I plugged it back in to the ecu circuit, the total resistance reading was 0.546Kohm.  As I was watching that number slowly drop it appeared to jump from 0.540Kohm to 525.9ohm on the meter which was due to it auto scaling, so no big deal.    I unplugged and checked the sensor on its own and it was at 1.000Kohm. Plugged it back in and  I turned the key on and hit start.  The bike purred like a kitten. No bog or choking or smoke.  So it seems to like the 550 ohm total resistance value which is achieved when the eng temp sensor gets down to 1.06Kohm and below.   

my thoughts are if I could make a parallel circuit that would be switchable, I could run it at 1Kohm for cold starting and then switch back to the original circuit via a selector switch and let the bike run as designed.  This way I don’t foul the plugs on a cold start.    So I know it draws 7mA, 12v when key is on but drops to 5.83v when engine is running and the current stays at 7mA. 

comments/suggestions are welcome!

Only take note of the resistance readings of the component When it is disconnected. Measuring resistance with the circuit connected and the meter will read the lowest value (path of least resistance - electricity is lazy) the 526 ohms reading will be through the ECU not the sensor. Wiring a switchable solution is straightforward with a 5 pin relay (when I get chance I will do you a diagram) It's a cost effective solution to what looks like a probable ECU fault (especially when the ECU is unique to the "G" model according to CMNSL)

Wookie58 wrote:

Only take note of the resistance readings of the component When it is disconnected. Measuring resistance with the circuit connected and the meter will read the lowest value (path of least resistance - electricity is lazy) the 526 ohms reading will be through the ECU not the sensor. Wiring a switchable solution is straightforward with a 5 pin relay (when I get chance I will do you a diagram) It's a cost effective solution to what looks like a probable ECU fault (especially when the ECU is unique to the "G" model according to CMNSL)
 

ok sounds good . Thanks