Yanmar 2GM20F - temp sensor keep buzzing after engine being rebuilt

[svsputnik]

Hey guys

Last year, I got a headgasket that leaked and got some water sitting in the combustion chamber, causing a low compression on one of the 2 piston. I got the engine back home and pretty much rebuilt it over the course of the winter (changed the piston rings, piston bearings, gaskets, hoses, got the head remachined, piston cleaned, injectors rebuilt, etc etc etc). You know, once you get the beast downstair, you realize just how much you don't want to get it down again and start changing everything that should be changed anyway.

I got the engine back in the boat last weekend. Re-plugged everything, bled the diesel line and, holding my breath, tried to start the beast. Well, it did. Hooray for me; i went from not knowing where was the sparkplug on my lawmower to overhaul a diesel engine. (I know, the 2GM20 is a really easy engine to work on, but let me dream a bit here).

But joy was short-lived. The buzzer that always buzz when you turn the key to indicate that there's no charge coming from the alternator didn't stop. I looked at the control panel and saw that it was the water temp that was still on...


It never stop, so it can be the engine overheating in a millisecond. (I mean, I'm in Quebec, Canada. It was still snowing when I did the test so I cannot believe the engine was already hot). So I'm guessing it's an electrical problem. Which is a bit weird, since I have moved the control panel. I just unplugged the wire when I got the engine out and re-plugged it the past weekend. Nothing fancy here. I did ID all the wires before, so I'm 99% sure the wire is at the good place.

To make things more confusing, the previous owner also installed a second sensor with an after market gauge, going from 120 to 240. I've never seen that gauge work before (because I'm guessing water temp should never reach 120 at the first place). Well, after stopping the engine, I realized that the gauge was peaked out in the 240 zone. (Took me a while to see this, I'm use to just ignoring it. My first thought was actually "wow, that thing actually works" and just after a few second did I think "uh oh, there's a problem there") . Even after killing the DC power, the gauge remained maxed-out. So I'm guessing it's dead now (?).

SO: Any idea? If not, anybody know how the temp sensor works? How can I multi-meter that thing? Is the sensor producing current only when there's a problem for instance? Or it's the inverse?

Any hint, clue or question would be nice, I'm not too sure how to tackle that problem.

Thanks
sam

[sailorchic34]

Several thoughts. Is the temperature gauge mechanical or electrical. That is are there wires going to it or a capillary tube. Mechanical would keep reading a high temp with power off.

at idle it would take a minute or two for the engine to come up on temperature even without raw water flow. Was raw water exiting the exhaust.

Did the head gasket go back on the correct way. The holes are not symmetrical in the head gasket. (well on the 3gm, I assume the 2 gm is similar)

I would check raw water flow first as that normally is what does it for temp alarm.

[Wotname]

Originally Posted by svsputnik

Hey guys

Last year, I got a headgasket that leaked and got some water sitting in the combustion chamber, causing a low compression on one of the 2 piston. I got the engine back home and pretty much rebuilt it over the course of the winter (changed the piston rings, piston bearings, gaskets, hoses, got the head remachined, piston cleaned, injectors rebuilt, etc etc etc). You know, once you get the beast downstair, you realize just how much you don't want to get it down again and start changing everything that should be changed anyway.

I got the engine back in the boat last weekend. Re-plugged everything, bled the diesel line and, holding my breath, tried to start the beast. Well, it did. Hooray for me; i went from not knowing where was the sparkplug on my lawmower to overhaul a diesel engine. (I know, the 2GM20 is a really easy engine to work on, but let me dream a bit here).

But joy was short-lived. The buzzer that always buzz when you turn the key to indicate that there's no charge coming from the alternator didn't stop. I looked at the control panel and saw that it was the water temp that was still on...


It never stop, so it can be the engine overheating in a millisecond. (I mean, I'm in Quebec, Canada. It was still snowing when I did the test so I cannot believe the engine was already hot). So I'm guessing it's an electrical problem. Which is a bit weird, since I have moved the control panel. I just unplugged the wire when I got the engine out and re-plugged it the past weekend. Nothing fancy here. I did ID all the wires before, so I'm 99% sure the wire is at the good place.

To make things more confusing, the previous owner also installed a second sensor with an after market gauge, going from 120 to 240. I've never seen that gauge work before (because I'm guessing water temp should never reach 120 at the first place). Well, after stopping the engine, I realized that the gauge was peaked out in the 240 zone. (Took me a while to see this, I'm use to just ignoring it. My first thought was actually "wow, that thing actually works" and just after a few second did I think "uh oh, there's a problem there") . Even after killing the DC power, the gauge remained maxed-out. So I'm guessing it's dead now (?).

SO: Any idea? If not, anybody know how the temp sensor works? How can I multi-meter that thing? Is the sensor producing current only when there's a problem for instance? Or it's the inverse?

Any hint, clue or question would be nice, I'm not too sure how to tackle that problem.

Thanks
sam

Sam,

You have given us a lot of information here which is good. Unfortunately we need a little more and we need to help you understand how the audio alarm, visual alarms (lights) and gauges work.

I realise you are pretty sure you have connected up the wiring correctly but either you have a faulty alarm sender (sensor) or you have got the wiring wrong.

First let's disregard the after market temperature gauge (and its sensor) as we are not sure whether it worked previously or not (but we will come back to it later ).

The audio alarm (buzzer) is heard whenever there is low oil pressure or an engine overheat problem. There is a corresponding light (idiot light) for each alarm condition. Each alarm condition has it's corresponding sender i.e. one for overheat and one for low oil pressure. The senders electrically work the same way. They are simply a switch which "turns on" when the alarm condition exists and connects the case of the sender to terminal on the sender. As the engine block is the negative side (ground) of the circuit, the sender thus provides a ground to the light (and buzzer) when activated.

The first test to do is check what is currently happening when you first turn the the key on (before starting the engine).

This what should happen:
Key on.
Buzzer sounds.
Oil pressure light on.
Temperature light off.
Charging light is on.

Note, the buzzer is sounding because there is no oil pressure before staring the engine.

If this is right, you can move on but if not, you have to correct whatever is wrong before proceeding.

Now start the engine and check the following.
Buzzer stops.
Oil pressure light is off.
Temperature light is off.
Charging light is off.

Clearly in your current situation, the buzzer is staying on but before solving the problem, we need to know if the above sequence of lights is occurring or otherwise.

Note the oil pressure alarm is self checking (both audio and visual) as the alarm condition obviously exists before starting but the temperature alarm is clearly not self checking.

Here is how to fault find the two alarm circuits:

Oil - self checking as above and to remove the alarm, disconnect the wire from the sender. The light should extinguish and the buzzer stop.

Temperature - initially no light or buzzer, remove the wire from the sender and touch it to the engine block. The light should illuminate and the buzzer should sound. This checks the light, buzzer and wiring but not the sender itself. The only way to do this properly is to remove the sender and heat it up - preferably in a pot of hot (boiling) water and test the "switch action" of the sender with a meter (ohm meter etc). You can of course heat it up slowly and measure the temperature of the water when the switch closes if you want to be pedantic (like me ).

The above is enough for you to solve your buzzer problem but please ask if you want more information.

If you are still reading, we can now discuss the after market water temperature gauge.

We will assume it is an electrical gauge with an analogue scale; if so, then almost for sure the sender is a temperature sensitive resistor i.e. it's resistance changes as it is heated. They are usually have a negative temperature coefficient (resistance decreases when heated). So its sensor is different in operation to that of the temperature alarm. To recap, the alarm sender is a switch type operation and the gauge sensor is a variable resistance style. If you get the wires swapped, neither will work.

To do a quick check of the gauge, turn the key on and if the engine is cold, the pointer should just move off it's low side stop. Now pull the wire off the sensor and the gauge should go to back to it's low side stop i.e. cold. Now touch the wire to the engine block and the gauge should go to full scale i.e. hot.

[svsputnik]

Wotname; Thanks you so much for all the help provided. I'm gonna go back to the boat tomorrow night and validate everything. I realize that I haven't paid attention if the water temp light was on before starting the engine.

Knowing that I can just stick the wire to the engine block to trigger the alarm makes the whole thing much easier to test / validate that I haven't inversed the wires.

I'll keep you posted tomorrow with the result of the tests!

Thanks
sam

[Wotname]

Originally Posted by svsputnik

.......

I'll keep you posted tomorrow with the result of the tests!

Thanks
sam

Looking forward to it

After the rebuild and now this, you are going to be full bottle on the 2GM series

[boeing1]

did you stay the water temp light and buzzer stays on after your start the engine, if that is the case, I would suspect a wiring problem or bad temp sensor, after starting a cold engine it should take a while before getting hot enough to trigger the temp sensor.

[Blue Crab]

Awesome post Wotname. Thanks. Same engine here.

[svsputnik]

Quick update. I've just disconnected all the wires coming from the sensors and the alternator. Weirdly enough, the light for the oil and charging went off but the light for the water temp is still on with the buzzing. Now I'm suspecting a short somewhere...

[Wotname]

Originally Posted by svsputnik

Quick update. I've just disconnected all the wires coming from the sensors and the alternator. Weirdly enough, the light for the oil and charging went off but the light for the water temp is still on with the buzzing. Now I'm suspecting a short somewhere...

Yep, you are onto the problem, the wire going to the water temp alarm sender is shorted to the negative side of the electrical system, probably to the engine block somewhere or there is corrosion in an connector housing.

The big multi contact housing between the engine loom and the panel loom often gives problems.

[svsputnik]

Originally Posted by Wotname

Yep, you are onto the problem, the wire going to the water temp alarm sender is shorted to the negative side of the electrical system, probably to the engine block somewhere or there is corrosion in an connector housing.

The big multi contact housing between the engine loom and the panel loom often gives problems.

My boat has been repowered by the previous owners who didn't think necessary to remove the old wires before putting the new in. It's such a mess of ducktape from the engine to the control panel. Oh well, let's dig right into it I guess. #alwayssomething

[Wotname]

Originally Posted by svsputnik

My boat has been repowered by the previous owners who didn't think necessary to remove the old wires before putting the new in. It's such a mess of ducktape from the engine to the control panel. Oh well, let's dig right into it I guess. #alwayssomething

I feel your pain
And somehow I know there will be some bad surprises once you start removing the duct tape so I wish you the best of luck and you will feel better once the bad stuff is exposed and fixed .

[charliehows]

yeah, i removed 3 full large rubbish bags of old wiring from my boat and that wasn't all of it, some i just left in place cos it wasnt in my way.

[svsputnik]

Alright. I spent the last hour removing electrical tape and old wires. Even though my water sensor goes through 4 different quick connect, there's no short that I can see of. The only suspect would be the switch that allow me to test the buzzer and lights. But it's connected to all the lights, not just the water temp. So I don't see by which magical means it would only light up the water. That project is definitely going through some serious periods of high hope than kill me please ! :#

[Wotname]

Originally Posted by svsputnik

Alright. I spent the last hour removing electrical tape and old wires. Even though my water sensor goes through 4 different quick connect, there's no short that I can see of. The only suspect would be the switch that allow me to test the buzzer and lights. But it's connected to all the lights, not just the water temp. So I don't see by which magical means it would only light up the water. That project is definitely going through some serious periods of high hope than kill me please ! :#

Okay, this getting serious .

Do you have the stock Yamnar 2GM20F panel or something else or a partly modified stock panel or ???

The stock Yanmar panels I am familiar with do not have a switch that allows one to test the buzzer and lights. Maybe yours is different???

It has two switches, one to illuminate the tacho etc and one to silence the buzzer.

If you do have a switch that tests all the lights and the buzzer all at once, then there will some diodes connecting the lights before going to the switch OR the switch is a multipole switch and each light will go to a separate pole of this switch. If so, maybe the one pole of the switch is faulty (i.e. the one going temp. light).

Might be time to break out the multimeter and I am going to assume you only have a basic understanding of the electrical aspects so I apologise if I am going too simply here.

Here is a basic check:

Remove the wire from the temp. sender.

Select DC volts on the meter.

Connect the meter between the wire and the engine block. If its a digital meter, it really doesn't matter which lead goes where but good practice says the red lead goes to the wire and the black lead goes to the engine block.

With the key off, meter should show no reading.

Turn the key on and the meter should show about 12V.

If it shows 0V or very low volts (say less than 5V), then there is something wrong with the wire or with some other wiring that has been added to the temp. alarm wiring by a PO

Standing by for more info

EDIT: try removing the oil pressure alarm wire from it's sender along with the wire to temperature alarm sender and the wire to the temperature gauge sensor.

Does the temp. light still come on??
If not, reconnect each wire one at time until the light comes on again.

This might help to narrow down where to look next.

[svsputnik]

Hehehe. Interesting it is! I have a modified version of the stock panel I guess. I have the B-panel with the tachometer. Added to that was an extension for the after market gauge that the previous owner installed. The one thing I'm not sure that is original on my panel though is the light switch. On mine, it's a off/on/momentary. Where the off is the default, the on lights up the light of the tach & after market gauge and momentary light on the sensor warning light and make the buzzer buzz. So I don't have a second switch to silence the buzzer, but 1 switch that can force it if for some reason I would want that. See attached img to for what it looks like (apology, we don't see the switch).

For the test, you're suggesting, I'm not sure I understand exactly. Here's what I was thinking doing:

- Disconnect the harness from the console and connect a 12v source to the temp wire at the console harness. That would send 12V along the line. Then, on the engine side of the wire, go with the multimeter from the water sensor wire to the sensor. That should read 0V (correct?). Then connect to the ground and then 1 should read 12V. If I read 12V between the harness and the sensor, it means I'm getting a ground somewhere the the problem is coming from the wire itself. If the wire is OK, then it means it's in the console itself. Am I ok in my assumptions?

I'm guessing we're both saying the same thing right, except your way is much easier?

So the next step will be the console. The switch would be the first one to be tested (I believe it is multipole, since the different light have their own pole). So having the switch in off position, I would be able to put the black lead on the negative of the switch the the red lead of the multi-meter on all the different poles of the sensors. I should read 0 on everytning if the switch is good, right??

Thanks so much for your guidance!

I did disconnect all the sensor wires from the engine and the water temp was still buzzing. Which in itself is a good news since it greatly reduced my stress that maybe the engine wasn't fixed properly.