Inaccurate fuel gauge?

[nola mike]

Replaced the sender a few years back. Went through great pains to get it sort of accurate. This year noticing that it's always full. Until it isn't. Realized that it's probably because I cleaned up some connections in the helm, and the gauge is now seeing lower resistance. I don't particularly feel like messing with the sender again, since I made some adjustments to account for the attitude of the tank when the boat is in the water. Seems like this would be a weakness on any sender reading from the stem of the tank. I'm tempted to just put a rheostat in line with the gauge, so I could calibrate it on the fly. Has anyone done this? Better solution? I generally have a good idea how much fuel I have, but I like to know when I'm under ½ tank

 

[440roadrunner]

Some random thoughts. Not all gauges are a simple current/ resistance situation. Some gauges are a "bridge" and generally can be identified by having two gauge terminals plus REQUIRING a ground, either the gauge housing or a 3rd terminal.

Some gauges you can do research and discover the resistance values for empty, full or 1/2 for testing

Also there "was" a product called "metermatch" which is an electronic device that inputs a sender "to the device" then scales the output to the gauge to make it accurate. You program it, and there are (don't remember) several settings up and down the range

It has evidently been bought by "Tanks Inc"

MeterMatch Fuel Gauge Interface Module

Tanks, Inc. manufactures fuel system components and polyethylene, steel and stainless steel gas tanks for street rods and special interest vehicles.

www.tanksinc.com

I would think there is more to your problem than improving grounding, as usually poor wiring/ connections/ grounding is not stable and would have given you randomly differing readings over time.

 

[dingbat]

What type of sender do you using?

A perfect ground path (0 ohm) isn’t going to improve the situation if the output of the sender isn’t in spec.

Have you put a meter on the output of the sender?

Have used a pair of KUS-WEMA senders for 10+years in combination with a Garmin GFS10 fuel flow meter.

The senders are very accurate and stable. No bouncing around in rough water. Gauge readings match usage pretty closely. Overall very high with the setup.

SSS/SSL Sender - SAE 5 Hole Pattern Mount Level Sender | KUS USA

This is our SSS/SSL electric liquid level sender. It mounts with a standard SAE 5-hole pattern. It works great with our industrial and marine applications.

kus-usa.com

 

[Scott Danforth]

what is the shape of your tank? is it fairly shallow and under the floor? bow mounted? stern mounted?

what angle does your boat sit at while static? on-plane?

I know of many shallow, below deck tanks that dont read worth a carp because they are long, and shallow and the boat sits at enough of an angle that the gauge doesnt start moving until the tank is well below 20%. some only read well when on-plane when the boat levels out.

some of the ultrasonic gauge senders work well, however most need to be installed on the forward end of the tank

 

[Wave34]

Nola Mike;
You say: ' to account for the attitude of the tank'.

It may be just that, the angle of the tank creating a bad reading at the level sensor.
On my boat, when stopped, the fuel tank is not level, in leans down at the front, and my gauge is at the back, so I read always lower than reality, and at one point I read 'E' and there is still half a tank. Frustrating. But it is the boat that is made like that, it is the same on other boats like mine.

The solution would be a ultrasonic fuel level sender that measures the volume of the fuel in the tank regardless of the angle and distribution.

Adding a Garmin GFS10 like dingbat did would be a good solution also.
dingbat, is this flow sensor accurate? I see in the specs that it doesn't read at low flow, like when moving at idle in channels.

 

[dingbat]

dingbat, is this flow sensor accurate? I see in the specs that it doesn't read at low flow, like when moving at idle in channels.

It is accurate but not precise below 1000 rpm.

At that point I’m burning roughly a gallon, gallon and a half an hour. Readings are cyclic (wonder if I’m not seeing the carb floats) but total reading at the end of the day falls in-line with expectation and gauge readings

The key to reliable gauge readings is resolution (minimal) and a means to average or trim the output from the sensor or the input to the gauge.

I get little if any bounce using 21mm resolution in a 14” tall, 94 gallon tank

 

[nola mike]

what is the shape of your tank? is it fairly shallow and under the floor? bow mounted? stern mounted?

what angle does your boat sit at while static? on-plane?

I know of many shallow, below deck tanks that dont read worth a carp because they are long, and shallow and the boat sits at enough of an angle that the gauge doesnt start moving until the tank is well below 20%. some only read well when on-plane when the boat levels out.

some of the ultrasonic gauge senders work well, however most need to be installed on the forward end of the tank

Yes, this is the main problem. It's a shallow, below deck tank at the stern. It's pretty much impossible to set it up to read anything less than full until it's less than 3/4, and usually then only when off plane. Guessing the extra stern weight of the new engine isn't helping.

What type of sender do you using?

A perfect ground path (0 ohm) isn’t going to improve the situation if the output of the sender isn’t in spec.

Have you put a meter on the output of the sender?

Have used a pair of KUS-WEMA senders for 10+years in combination with a Garmin GFS10 fuel flow meter.

The senders are very accurate and stable. No bouncing around in rough water. Gauge readings match usage pretty closely. Overall very high with the setup.

SSS/SSL Sender - SAE 5 Hole Pattern Mount Level Sender | KUS USA

This is our SSS/SSL electric liquid level sender. It mounts with a standard SAE 5-hole pattern. It works great with our industrial and marine applications.

kus-usa.com

Using a regular arm type sender. The bouncing actually helps me. Once it starts bouncing, I know that there's enough empty space in the tank for the gas to slosh around. Good point about measuring the sender. I did that when I replaced it a few years ago, haven't done it recently.

I would think there is more to your problem than improving grounding, as usually poor wiring/ connections/ grounding is not stable and would have given you randomly differing readings over time.

Not really. I had some corrosion at the main ignition fuse that was causing a pretty steady voltage drop. That was worse (and gauge readings changed) with increase current draw, particularly when the blower was on. I still have some drop across the ignition switch, but don't know when I'll get around to replacing it. Temp jumps up 7' with the blower on, 2' when the radio's on...

Basically I'm not looking for any real accuracy, and I'm not looking to devote a bunch of time or money to it. More like an idiot like for fuel. Like I said, I usually have a pretty good idea of how much gas I have based on the distance I've traveled. I'm figuring I can pick a point when I'm at 1/2 tank or so, and then install the potentiometer and adjust so that the gauge reads 1/2 tank when I'm sitting level in the water. Any reason that wouldn't work?

 

[Wave34]

I get little if any bounce using 21mm resolution in a 14” tall, 94 gallon tank

Does that mean that you use the analog output of the sensor to your gauge, and not the NMEA2000 ?

I have an older Garmin GPS, and Garmin say that I need to upgrade the firmware to be able to accept the fuel flow sensor on NMEA2000.

 

[dingbat]

Does that mean that you use the analog output of the sensor to your gauge, and not the NMEA2000 ?

I have an older Garmin GPS, and Garmin say that I need to upgrade the firmware to be able to accept the fuel flow sensor on NMEA2000.

Using analog sensor.
I have two tanks (Main and Aux) on one gauge. Flip a switch to select which tank to read and I wanted to keep that functionality.

The GFS10 gave me the option to tie the main tank analog sensor into NMEA network which I did.