Fuel Sender Resistance Scale

[John E3]

I am trying to correlate my fuel usage info from my Engine Gateways with actual fuel consumption (based on sender resistance), and cannot find any details about the actual resistance at specific levels, or any chart. The 240-33 ohm range is non-linear it seems. The best I found so far is only the usual points, 0, 1/4, 1/2, 3/4 and full. The change in difference is apparent between them. This is comparable to an radio volume control which has an "audio taper", but this taper is different (first 10% change = 50% resistance change, low to high). Does anyone have any further insight into sender ohms vs position?

Our last boat had a straight line from cap to tank bottom and my gas sticks were so easy.... Not an option any more!

I built a spreadsheet with 0.1" increments, determined the sender offset from top and bottom, then divided up the 20" travel of the sender (moeller reed type) equally over the range. What I found is that when the tank is low I "use" much more fuel (~50%), then when it's fuller. This became evident after filling the tanks at the midpoint of a trip over the weekend, only the 2nd fill of the season. The usage I was seeing before the fill was scary compared to the reported consumption! Post fill usage seems low compared to gateway reported usage.

My objective is to get a good handle on the reporting vs actual use, to calibrate the usage if needed, and rely on that going forward. Our Axiom MFD includes a Fuel Management tool, getting usage from engine gateways, and manual input of each fill. The gauges are quite inaccurate.

Speaking of gauges, is there such a thing as recalibrating SR gauges? The OP is pretty close, but tachs are significantly off (500rpm at 3800). The synchronizer and gateways agree, +/- 25rpm. The Temp gauges seem pretty close, but port runs about 170 (vs 160) higher when running 3800 rpm, according to the engine gateway, but stays at 160 on the gauge. I believe there are 2 senders, 1 to gauge and 1 to MEFI, so a difference may be related to where they are. Or not. Looks like both are in Thermostat housing, on opposing sides. Maybe just sender difference.

 

[Skybolt]

If it is a wire wound sender then it should be linear, but if it is one of the ones off of Amazon then good luck. Those are not linear in any fashion.

I doubt there are two senders on a single tank.

If you want to be accurate on fuel consumption then change your senders to ultra sonic ones. For highly accurate, switch to Maretron TLM150's but you will also need and NMEA 2000 display to read the levels, like your Axiom.

I use the above and am accurate to the gallon without spread sheets.

EDIT: Fuel management on MFD's usually require fuel flow to be present as a parameter to be able to calculate anything. It's rare that Raymarine or Garmin supports that by fuel level alone. I know My new Garmin's won't do that. And I have tried many times including their useless tech support. Raymarine may have better support for this.

 

[John E3]

If it is a wire wound sender then it should be linear, but if it is one of the ones off of Amazon then good luck. Those are not linear in any fashion.

I doubt there are tow senders on a single tank.

If you want to be accurate on fuel consumption then change your senders to ultra sonic ones. For highly accurate, switch to Maretron TLM150's but you will also need and NMEA 2000 display to read the levels, like your Axiom.

I use the above and am accurate to the gallon without spread sheets.

As noted above, moeller reed type, are what I have. These are not wirewound, or variable resistance, they have incremental step changes. They are new, and changing to ultrasonic is not an option, or interest. As noted, I am only trying to correlate my usage now, so that I can fine tune the Axiom for the future. There are plenty of other tank level options out there.

 

[John E3]

EDIT: Fuel management on MFD's usually require fuel flow to be present as a parameter to be able to calculate anything. It's rare that Raymarine or Garmin supports that by fuel level alone. I know My new Garmin's won't do that. And I have tried many times including their useless tech support. Raymarine may have better support for this.

The Axiom MFD has a Fuel Management screen. A few basic configs, which includes choosing the fuel usage reporting. Mine discovered the 2 MEFI fuel PGNs from the gateways. The screen only shows the estimated fuel remaining while the ignition is on, but does show cumulative usage either way.
I found this image online, for reference.

 

[Skybolt]

The Axiom MFD has a Fuel Management screen. A few basic configs, which includes choosing the fuel usage reporting. Mine discovered the 2 MEFI fuel PGNs from the gateways. The screen only shows the estimated fuel remaining while the ignition is on, but does show cumulative usage either way.
I found this image online, for reference.

Yeah, I could never get Garmin to do that without Fuel flow. I know RM is a little better at fuel management display of basic system management. I use Maretron's DSM 410 with the fore mentioned senders, works very well.

 

[Bill Curtis]

I am trying to correlate my fuel usage info from my Engine Gateways with actual fuel consumption (based on sender resistance), and cannot find any details about the actual resistance at specific levels, or any chart. The 240-33 ohm range is non-linear it seems. The best I found so far is only the usual points, 0, 1/4, 1/2, 3/4 and full. The change in difference is apparent between them. This is comparable to an radio volume control which has an "audio taper", but this taper is different (first 10% change = 50% resistance change, low to high). Does anyone have any further insight into sender ohms vs position?

Our last boat had a straight line from cap to tank bottom and my gas sticks were so easy.... Not an option any more!

I built a spreadsheet with 0.1" increments, determined the sender offset from top and bottom, then divided up the 20" travel of the sender (moeller reed type) equally over the range. What I found is that when the tank is low I "use" much more fuel (~50%), then when it's fuller. This became evident after filling the tanks at the midpoint of a trip over the weekend, only the 2nd fill of the season. The usage I was seeing before the fill was scary compared to the reported consumption! Post fill usage seems low compared to gateway reported usage.

My objective is to get a good handle on the reporting vs actual use, to calibrate the usage if needed, and rely on that going forward. Our Axiom MFD includes a Fuel Management tool, getting usage from engine gateways, and manual input of each fill. The gauges are quite inaccurate.

Speaking of gauges, is there such a thing as recalibrating SR gauges? The OP is pretty close, but tachs are significantly off (500rpm at 3800). The synchronizer and gateways agree, +/- 25rpm. The Temp gauges seem pretty close, but port runs about 170 (vs 160) higher when running 3800 rpm, according to the engine gateway, but stays at 160 on the gauge. I believe there are 2 senders, 1 to gauge and 1 to MEFI, so a difference may be related to where they are. Or not. Looks like both are in Thermostat housing, on opposing sides. Maybe just sender difference.

From the Telflex paperwork. These are are the standards for the gauges. There is no 1/4-1/2-3/4 standards, just full and empty. Everything in between is just a percentage of the full to empty numbers.

 

[ttmott]

As long as the fuel tanks are rectangular the sender values will be linear.
Fuel level is all any data system can provide based upon the tank level sensors. The rate of level change is so slow and vessel angle/sloshing changes levels that flow/consumption data can't be realistically displayed in a real time fashion.
Modern ECM's provide calculated fuel consumption rate based upon other engine variables like RPM, Throttle Setting, Injector Pulse Width, Manifold Air Pressure, and Intake Air Temperature. Those values (PGN) can be provided on either a Smartcraft or NMEA 2000 data system depending upon your configuration.
Averaging, range, and economy are calculated and provided by the display system (Vessel View for example). This is because your engine's ECM does not know the fuel tank size.

 

[ranger58sb]

I am trying to correlate my fuel usage info from my Engine Gateways with actual fuel consumption (based on sender resistance), and cannot find any details about the actual resistance at specific levels, or any chart. The 240-33 ohm range is non-linear it seems. The best I found so far is only the usual points, 0, 1/4, 1/2, 3/4 and full. The change in difference is apparent between them. This is comparable to an radio volume control which has an "audio taper", but this taper is different (first 10% change = 50% resistance change, low to high). Does anyone have any further insight into sender ohms vs position?

What is the shape of your tank?

-Chris

 

[John E3]

The tanks are perfectly rectangular as best I can tell. They sit outboard of the engines, and all but the outboard side is visible. I can probably see the back side profile if I stick my head in the right place.

Many of the other fuel sender types out there (mostly car related) I found are not linear. The rate of change in resistance is faster as you approach one end of travel. This is very common in other types of electronics as well (referred to as audio or logarithmic taper. I have 50yrs of hobby and some professional electronics background). I am baffled as to why I can't find a technical reference for the "std" sender, 240-33. It's apparently a universal "standard range".
While there may be no 'reference' @ 1/4, 1/2, 3/4, they do correspond with an exact ohm value. I did find a few graphs, but for different ranges. A 220-20 in particular was the closest I could find. These all show the change in resistance, like this.

I estimated, based on ohms, that we used 68 gallons from Swan Creek to Annapolis. I also estimated that they were about 1/3 full on arrival. We put on 175gal, which would have brought us to almost 275 gallons (ohms agreed). The return trip, based on ohms, used 33 gallons. 2 trips to BBB (worton creek), first 52 gallons (tanks low) and 2nd 38 gallons (over 1/2 full).
The high numbers seem too high, and the low numbers seem too low. However, cumulative for the season the ohms says we used 233 gallons, and the Fuel management on the Axiom estimated 218 used. I totally failed to note the Trip usage for any of the trips.
Based on runtime, and reported fuel consumption, somewhere between these numbers is where I expected to be. All 4 of these trips were around 50g/hr with a little under an hour to BBB and a little over an hour to Annapolis, at that speed. The high usage with a low tank, and lower usage with a fuller tank corresponds with a change similar to that chart. I will try to update my spreadsheet based on this chart (with 13 added to the values to better match 240-33) and see if the number make more sense.

The next course of action would be to buy another sender, and set it up in my shop with an Acme screw to raise the float (10 tpi, 1 turn = 0.1", I have one on hand) and measure the resistance at every turn. Some might consider this overkill, but I want to fine tune the engine gateways reported fuel consumption so I can rely on that. The gauges are simply too unreliable, and stbd goes way off on occasion. The original stbd sender was dead, and when I installed the new one, at 20% full (30gal, measured with a stick) the gauge read way under E. They both read about 1/4 before the 175g fill and almost Full after (before stbd drifted to E, then back up slowly).
I estimate that there is about 15gal below the sender, and maybe 5 more before the float rises, and have considered that in my analysis.

 

[Nater Potater]

Something doesn't jive with the Ohms curve you got (nice graph, by the way) if you believe it to coincide with the Teleflex info page that @Bill Curtis posted back in #6. If you subtract the Empty value (240Ω) from the Full value (33Ω) and divide by 2, you should be seeing 103.5Ω (Teleflex says 103Ω) at half-full, so that "should" denote it as a linear pot. Based on the numbers you're getting, I'd be for replacing your sender.

 

[John E3]

Something doesn't jive with the Ohms curve you got (nice graph, by the way) if you believe it to coincide with the Teleflex info page that @Bill Curtis posted back in #6. If you subtract the Empty value (240Ω) from the Full value (33Ω) and divide by 2, you should be seeing 103.5Ω (Teleflex says 103Ω) at half-full, so that "should" denote it as a linear pot. Based on the numbers you're getting, I'd be for replacing your sender.

The graph was for the 220-20 ohm sender I referenced, not mine, showing the curve dropping faster at the start.
I tried extrapolated those numbers to my chart and it got worse. Both senders are new Moeller reed types.

The reference does suggest linear, but the float type will not be linear over the range. The rate of change (for a given level drop) at the top and bottom is greater than the middle. The longer the float arm, the less of a difference. Maybe the reed type do the same to keep the same exact gauge reading? The floats were around many decades before the rest.

Or I'm just overthinking it. But my numbers suggest not. These are based on ohms, sender working range, offset below and dead space above.

Roundtrip to BBB - 52g (start 3/4 tank)
Roundtrip to Marshy Creek - 42g
To Annapolis - 68g (end 1/4 tank)
Fill, Back from Annapolis - 33g (90 octane, eth free, 90% full at start)
Roundtrip to BBB - 38g (end 2/3 full)

If you add up usage, the tank levels pretty much match (and agree with Axiom total usage). But that 2:1 difference to/from Annapolis, and diff on BBB, makes no sense. Wind was from the south on both BBB runs. Not sure about currents, except return on 2nd BBB was with us, but almost slack on arrival.

I was also using my IR Temp gun and was able to corroborate the approximate tank level a couple times. Bottom up was +/- .1F, then suddenly rises. Probably +/- a couple inches. Only works at ambient temps, morning is better.

 

[Travis M]

I'm a little late to this discussion. New to the boat. Didn't know how much fuel was in the tanks when I bought the boat. Fueled once but didn't top off. Still, I felt my fuel gauges were reading low and started to troubleshoot.

Of course, I have concluded that the senders are right and not linear. Neither is the movement of the gauges. The result is a fairly linear gauge position as an interpretation of the non-linear sender resistance as a representation of the depth of fuel in the tank.

The sender is acting linearly given the sweep of the arc of the float. Just not linear given the vertical depth of fuel. The gauge compensates.

 

[John E3]

Hi Travis,
Swing arm floats do have a non-linear change from bottom to middle, then it reverses from middle to top, but the resistance material in the sensor could be compensating for this.
I have reed type sensors, so the physical rise is linear from bottom to top. The boat is out of the water now and I plan to pull a sensor, stick the tank, and measure it's resistance from bottom to top, possibly tonight. No chance of discrepancy with this check.
I was amazed to find that both senders measure the exact same value prior to the haul. Both fills we did this year were not equal in an attempt to get them to the same levels. I will not be pulling the stbd sender as it's under the carpet and I don't want to damage the carpet any more than I did to get it swapped out initially.

 

[hughespat57]

There are a lot of variables involved in a boat fuel tank.
Even if perfectly rectangular, does the tank sit perfectly level when fueling?
Is the sender in the front, center or aft portion?
When you top off can you be sure you are stopping at exactly the same point?
And unless you have actual flow meters the burn is calculated buy the engines based on injector pulse timing, fairly accurate but a calculated value based on injector flow rates.

A man with one watch knows what time it is.
A man with two watches is never sure.

 

[John E3]

There are a lot of variables involved in a boat fuel tank.
Even if perfectly rectangular, does the tank sit perfectly level when fueling?
Is the sender in the front, center or aft portion?
When you top off can you be sure you are stopping at exactly the same point?
And unless you have actual flow meters the burn is calculated buy the engines based on injector pulse timing, fairly accurate but a calculated value based on injector flow rates.

A man with one watch knows what time it is.
A man with two watches is never sure.

Yep, and pretty much everything you noted was already discussed, except topping off. Unless you have a straight pipe down, pumping out the overflow is the only way to know when it's full. I miss those fill pipes that our wooden cruiser had. 10 seconds with a dowel and you know the exact level. The 1928 yacht I grew up on, and identical one that my nephew just bought is that way as well. No need for senders and gauges!
Once I can 'calibrate' my readings I can compare that to the engine gateway reported usage.

 

[John E3]

I pulled the sender and measured it last night. It is absolutely nonlinear. 25 steps at 3/4" increments. First step from bottom was a 19 ohm change, last step at top is 5.3. I'll plot and post it later.
This confirms what my checks suggested.

 

[Travis M]

Are you saying 3/4" increments of moving the float up. As if riding the level of fuel?

Makes sense, right? If 103 ohms is half, that's 70 ohms to full but 136 to empty for a 33 to 240 sender.

I measured my gauge. The needle movement is also not linear and compensate for the non linear sender.

So, correlation from resistance change to fuel burn would have to use a chart. But, gauge should show a mostly reliable sweep.

Assuming everything is working correctly.

 

[John E3]

Yes, as noted I have reed type senders. As the float rises it makes a jump to the next value every 3/4" for a total of 25 steps.
The difference between steps from mid point to the top (full) is near zero (at 6ohms/step) but from mid point down, the change increases quite a bit at each step (difference of 6 midpoint to 19 at the bottom step)
I'm assuming this is to try and approximate the same change you would get with a float lever arm, although the pivot point would impact this as well. If at the bottom the change is greater there, than as it rises. If the pivot is nearer the top, the change would be greater at the top, instead.
My OEM ones had the pivot point at about the middle and that would create more of a 'bell' type of curve. Greater change at the top and bottom and smaller change in the middle.

The numbers are pbly more telling:

 

[John E3]

My calibrated chart. The whole point of this exercise is to use my Axiom MFD Fuel Management data to track fuel use and not rely solely on the gauges, especially at refueling time. Now that I have calibrated it I can validate the Axiom numbers, next season.