Fuel gauge calibration point (1 Viewer)

[Spook50]

I'm hoping to catch some more input by putting this in its own thread. Relating to my solid state regulator project, but this could also help with others who don't go with the upgrade once it's available.

What I'm trying to do is figure a good point for baseline calibration of the fuel gauge. Previously I would set it so that with a full tank the needle would hit the upper right tip of the "F" on the gauge, but I'm thinking that may not be the best approach. Given that the OEM fuel tank capacity is 23.8gal (which I had for the longest time, up until just recently, thought was 22gal), my thought is to bottom drain it, and then pour a measured 4 gallons in, then set the gauge so that the needle rests in the upper range of the "E". I'm unsure if this will be the ideal option though so I'm looking for input that anyone else is willing to throw out there. My idea is that by the time the needle reads empty, you'd still have a small reserve to get to the nearest station (or pull-off to throw in a Jerry can).

Potential issues I see are the roundabout accuracy of the temp gauge since that is fed by the same low voltage hot lead as the fuel gauge, and where the fuel gauge will read when full, though those will be easy to figure out after driving the truck for a moderate highway drive to top up the tank.

 

[ChaserFJ60]

I would empty the tank completely on a level surface via the fuel pump. There may be some residual fuel in the tank that's required as a "Filler" before the gas gets pulled in and I would use that as my totally empty baseline, then add the 4gal of reserve from there. I would have the reserve fuel where the needle does not move upwards at all and is completely at the bottom of its mark, Somtimes when people are in a pinch on fuel, if they see the needle move up at all they will assume there is fuel when there is actually none if the needle where to be offset upwards.
Im not sure if the FJ60-62 floats vary, I do know the FJ60 has a mechanical fuel pump and the FJ62 has an electric fuel pump inside the tank.

As for the Full mark, I would try to incorporate as much needle travel as possible, I would go as far as the needle can travel upwards, I think it does have a mechanical stop to prevent it from doing a full revolution. I think that going to the max would help prevent deviations in Full readings to folks with different resistance signals.

 

[Spook50]

I had considered the fuel pump method for just the reason you stated. The problem with that avenue is once you run a fuel pump dry, you're on borrowed time until it goes out. If it were easy and inexpensive to "sacrifice" a fuel pump, that would be the method I use too. As it is, the pump sits inside of a small sump that I would guess (and this is really just spitballing) would have maybe a quarter to a half gallon at the point the pump is unable to draw anymore fuel into the system.

You have a good point about setting it to the bottom-most point of needle movement allowing for that 4gal reserve. Given that, I'll try that first, then see where the needle sits with a full tank when I fill it and then check the radiator top tank's temperature hot to compare it to where the temp gauge needle sits.

As far as the range of movement, with the stock system you can really only easily calibrate the fuel and temp gauges at one point (by setting the voltage for the hot lead to the fuel and temp gauges), compared to an aftermarket electronic fuel gauge being able to be programmed for both full and empty readings. This is what led me to consider making a baseline calibration point for the empty reading on the gauge.

 

[Spook50]

After work today I drained my (fortunately already low) tank and put 4.25 gallons in. With this, my gauge indicated just over the lower line:

So I went about adjusting the pot on my regulator, but unfortunately I'll have to find a different pot to use that can be adjusted to an even lower resistance. As it was, with the pot at its lowest setting, my needle still sat lower than where it started. But it's still a ways above being bottomed out. Given that, I'll run it as it sits now while I investigate and order a different pot. I'm curious to see where the needle sits when I top off my tank, and where the temp gauge will sit with a hot engine (especially during a long highway drive with the AC running. Above 90º Fahrenheit ambient is when my temp needle would creep into the red at highway speeds). Here's where my fuel gauge needle sits now with only 4.25 gallons in the tank:

Just for the sake of easy reference in this thread, this is where the needle sits when bottomed out:

It appears I need to polish the clear plastic of my gauges again....

 

[ChaserFJ60]

Does it have any reserve fuel stock? Would adding fuel change the float output to needle reading curve?
I'm sure you've already seen this figure but ill post it here anyways

I went ahead and plotted The resistances in excel using the book data, This is probably not 100% correct but its what I had to use

Based off of my limited data, you want input signal to needle position plots to be near parallel to these lines, or within the two extremes.
This chart is assuming that 0 = Empty, 1 = Half Tank, 2 = Full.

Now say you to adjust where the needle points for full and empty. In my made up case we would want the needle above E in the white line more so we would assume a new empty position of 0.25 = Empty, The Half full region would be untouched so 1 = Half Tank still, But we also dont want the needle going past the F mark so we lower its position to 1.75 = Full.

Any offset will change how the needle sweeps, I have aftermarket speed hut gauges and the fuel gauge is super annoying. For the first 200 miles the needle only falls to the halfway point, Then the other half of the gauge sweeps quickly for the last 60 miles, its unintuitive and annoying. I guess I could curve fit it and get fancy, but I really dont feel like it.


If we add a reserve amount of fuel and reassign the resistance values without changing the original needle travel position we still get a new curve

Some new curve fitting would have to be applied in your program to match the OEM curve, Maybe it doesnt deviate too much and is unnoticeable? Idk. just throwing this info out there. Hopefully its not all nonsence and im not way off topic.
I bet what Toyota did was plot the actual fuel level (gallons to % full) and curve matched that data with the float to the sender. I'm not sure if the data I'm throwing out is on par with what your doing or not. A controls or electrical engineer would be more helpful than I could ever be on this topic.

 

[tmxmotorsports]

All tech right here lol when I drove cross country in my 62 at 200 miles on the odometer in a panic looking for gas stations.

 

[Spook50]

Does it have any reserve fuel stock? Would adding fuel change the float output to needle reading curve?
I'm sure you've already seen this figure but ill post it here anyways
View attachment 3396957
I went ahead and plotted The resistances in excel using the book data, This is probably not 100% correct but its what I had to use
View attachment 3396986
Based off of my limited data, you want input signal to needle position plots to be near parallel to these lines, or within the two extremes.
This chart is assuming that 0 = Empty, 1 = Half Tank, 2 = Full.

Now say you to adjust where the needle points for full and empty. In my made up case we would want the needle above E in the white line more so we would assume a new empty position of 0.25 = Empty, The Half full region would be untouched so 1 = Half Tank still, But we also dont want the needle going past the F mark so we lower its position to 1.75 = Full.
View attachment 3396992
Any offset will change how the needle sweeps, I have aftermarket speed hut gauges and the fuel gauge is super annoying. For the first 200 miles the needle only falls to the halfway point, Then the other half of the gauge sweeps quickly for the last 60 miles, its unintuitive and annoying. I guess I could curve fit it and get fancy, but I really dont feel like it.


If we add a reserve amount of fuel and reassign the resistance values without changing the original needle travel position we still get a new curve
View attachment 3396995

Some new curve fitting would have to be applied in your program to match the OEM curve, Maybe it doesnt deviate too much and is unnoticeable? Idk. just throwing this info out there. Hopefully its not all nonsence and im not way off topic.
I bet what Toyota did was plot the actual fuel level (gallons to % full) and curve matched that data with the float to the sender. I'm not sure if the data I'm throwing out is on par with what your doing or not. A controls or electrical engineer would be more helpful than I could ever be on this topic.

What's funny is I perfectly understood what you were saying, and that resistance curve is actually an issue I had in mind with trying to find an "ideal" calibration point for these regulators. Unfortunately, as you illustrated (and I am horrible at explaining myself so glad you were able to put a visual up on here), the variation between full, 1/2 and empty is a legitimate curve and not a straight linear path. The issue you describe with your Speedhut gauge is very similar to how my fuel gauge behaved before I started experimenting with this conversion. It does seem to be improved, though not perfect, with a solid state regulator. I wonder if a floatless sender (Holley now makes one that uses LiDAR so a completely solid state sender) would have a more linear path than the OEM float style sender to allow for a much more accurate indication of fuel level, but that would be something to consider for future possible modifications for fuel level indication. My goal still is to make mine as reliable as possible for an otherwise completely stock application, which I think due to the curve we're talking about, will take some trial and error to find a "perfect" calibration point of known fuel quantity associated with needle position. Something else I had in mind which prompted me to make the regulator adjustable was the age of wiring and connectors in different vehicles. This will cause slight variations in resistance in the electrical path from battery to fuse panel to gauge to sender to ground, so I wanted to allow compensation for that.

As far as a factory-designed fuel reserve, I honestly have no clue. I think the only way to find out for sure at this point, given I've found zero documentation on any fuel reserve, would have been on a nearly new 60 or 62 with the gauge reading empty, and then finding out how much comes out when the tank is bottom drained. On the other hand, hopefully finding an ideal calibration point will help to answer that question or at least give us a close idea.

Edit: It just occurred to me that any factory designed reserve can be verified by measuring the resistance of the (preferably new) OEM sender and determining how much fuel in the tank is necessary to reach the nominal 120Ω

 

[ChaserFJ60]

for the float idea you had there was some talk on this old thread of mine of floatless senders, Mind you this is specific for the speed-hut gauges but maybe it will provide more direction?

Speedhut gauge criticism

Just got these new guages by speedhut.com for my 1984 fj60 with a 2fe. I got the quad left and just a GPS speedo. I got them in the green color to try to match the original aesthetic but it dosent show well in the dark at all. Infact trying to see these guages with the backlight turned up all...

forum.ih8mud.com

 

[Spook50]

I do remember that thread. Had to skim over it again for a refresher though. The Tanks Inc sender was what I was originally thinking I'd like to replace my OEM sender with, but seeing this new Holley just today, I'm very curious about that one. Especially since it could easily be swapped over to an extended range tank. From the few comments I've seen it does look like a much more linear path from empty to full resistance than the OEM sender would be.

I think just by nature of its design my regulator will work well with an aftermarket sender as well as an OEM sender, but I suspect most would be sticking with an OEM sender until they're NLA. So that puts us back to square one of finding the ideal setting point for a known fuel quantity to coincide with the appropriate position of the gauge. Next time I bottom drain my tank I'll add fuel a small amount at a time (maybe half gallon I'm thinking) while measuring the resistance of the OEM sender and that should tell me what, if any, reserve Toyota designed into the system.

 

[CruiserTrash]

@Spook50 listening to all these design constraints and thinking … you should put a trim pot on your board to adjust for OEM or aftermarket senders, or variances in gauge characteristics. I would pre-set them, match-mark them, and in the instructions tell the end user or installer NOT to touch the trim pot under ANY circumstances. But it’ll be there if needed.

In a former life I designed and made guitar effects pedals. One of them was three FET gain stages in series. The FETs operating window was a little temperamental, especially as you increased the signal gain at each stage. It required three trim pots and a lot of “tuning” to get right. If not tuned properly it would sound gated or squawky, just really bad. I’d send them out tuned perfectly and after about a year the customer would ping me that it sounded like s*** so I’d have to retune it. I was very thankful for building trim pots into those things.

 

[Spook50]

@Spook50 listening to all these design constraints and thinking … you should put a trim pot on your board to adjust for OEM or aftermarket senders, or variances in gauge characteristics. I would pre-set them, match-mark them, and in the instructions tell the end user or installer NOT to touch the trim pot under ANY circumstances. But it’ll be there if needed.

In a former life I designed and made guitar effects pedals. One of them was three FET gain stages in series. The FETs operating window was a little temperamental, especially as you increased the signal gain at each stage. It required three trim pots and a lot of “tuning” to get right. If not tuned properly it would sound gated or squawky, just really bad. I’d send them out tuned perfectly and after about a year the customer would ping me that it sounded like s*** so I’d have to retune it. I was very thankful for building trim pots into those things.

I'm actually already there with the trim pot. The design is a fairly basic (some minor customization for the 6x series) switching voltage regulator with 8-36 VDC input and an output adjusted via a trim pot installed in place of a resistor. As it sits now, and this is if I use a reserve level of fuel as a reference point for setting the empty indication of the gauge, I will likely have to use a lower value resistor that's in series with the existing trim pot. I won't jump in yet with modifying the regulator components at least until I top off my tank tomorrow and see where the needle sits, and maybe not until the end of next week when I run the tank low enough again to bottom drain it and determine how much fuel has the sender sitting at the resistance specified in the FSM (thanks to @ChaserFJ60 for reminding me that that's in there before I got too far ahead of myself). Step one will be topping off the tank and seeing where the gauge sits, and if it's still looking good step two will be determining what reserve amount (if any) puts the sender at the indicated measurement. I'll go off that measurement and set my regulator to have the gauge rest at E and then see if I'll have to change out any components on my regulator design.

The downside is that unless I went way more in depth with designing a digital control, there's no way to perfectly compensate for the resistance curve that the OEM sender has. This is a limitation of the factory mechanical voltage regulator too unfortunately.

 

[CruiserTrash]

Right, you can see the factory sender/gauge combo sit about halfway for far too long, then plummet during the last half of the sweep. You're just doing things the Toyota way!

 

[Spook50]

LOL you're not wrong. I think I may have opened up a can of worms in this stage of the development. I'm a big proponent of the "KISS" method of design, since that limits potential failure points. I'm still hoping that with enough trial and error, and finding an ideal set point for the regulator that purchasers can use to calibrate their gauges, I can at least alleviate that curve a small amount and get some more reliable readings with the factory gauges and sender.

A possibility that I'm considering is once I determine any reserve quantity of fuel, fill the tank halfway between that point and full (so say it's 3.8gal reserve, add 10 gallons. This is just an example pulled out of thin air) and then set the regulator to have the needle rest at the 1/2 mark, then see how it and the temp gauge behave.

A big caveat I'm having to keep in mind is the temp gauge, which is also powered by the fuel gauge's regulator. So any adjustment I make that'll lower or raise the needle of the fuel gauge will have the exact same effect on the temp gauge.

 

[cps432]

I'm hoping to catch some more input by putting this in its own thread. Relating to my solid state regulator project, but this could also help with others who don't go with the upgrade once it's available.

What I'm trying to do is figure a good point for baseline calibration of the fuel gauge. Previously I would set it so that with a full tank the needle would hit the upper right tip of the "F" on the gauge, but I'm thinking that may not be the best approach. Given that the OEM fuel tank capacity is 23.8gal (which I had for the longest time, up until just recently, thought was 22gal), my thought is to bottom drain it, and then pour a measured 4 gallons in, then set the gauge so that the needle rests in the upper range of the "E". I'm unsure if this will be the ideal option though so I'm looking for input that anyone else is willing to throw out there. My idea is that by the time the needle reads empty, you'd still have a small reserve to get to the nearest station (or pull-off to throw in a Jerry can).

Potential issues I see are the roundabout accuracy of the temp gauge since that is fed by the same low voltage hot lead as the fuel gauge, and where the fuel gauge will read when full, though those will be easy to figure out after driving the truck for a moderate highway drive to top up the tank.

This is how I gave gauged how much I have left in the tank before I’m actually out. When the needle hits the top corner of the E, I know I can get to and from work two more times before I absolutely have to fill it up. Completely empty happens when the needle is touching the top of the bottom line on the E.

I never knew the capacity at that point. I just knew how long I could push my luck on a known route. Three or four gallons sounds about right. If the actual volume of the tank is 23.8 then marketing it as a 21 makes sense because you don’t want to completely top off the tank.

Knowing I would have 4 gallons at the top of the E would give me a bit of peace of mind. Especially out here in Montana where it can get tough to find gas way out in the sticks. Having an idea of how many gallons you actually have left (like a modern car) would be nice.

 

[ChaserFJ60]

I think there is some gap between the bottom of the tank that already acts as a reserve, Not sure how low the level can be before no gas can be pulled up

 

[Spook50]

I think there is some gap between the bottom of the tank that already acts as a reserve, Not sure how low the level can be before no gas can be pulled up

I think so too. I wish when I had the tank off to replace my fuel pump a couple years ago that I would've had the foresight to measure the length of my fuel pump intake from the flange bottom compared to the tank depth and done a "fill test" to measure the quantity that would be left before the pump would start sucking air. 4 gallons of reserve before reaching that point would definitely be a good safe margin I think. It's something else I'll have to check during my trial & error process.

Interesting that your gauge needle can go down to the bottom line of the E compared to mine, which I snapped the "bottomed out" picture of with power off. I forget, is your rig a 60 or a 62? Mine's a 62, but I have a 60 quad gauge in my garage that I'm curious to look at now and see where the fuel gauge is sitting without power. If there's a variance between the 60 and the 62, I might have to use the "half of capacity minus reserve" point to set the 1/2 indication that I mentioned earlier to keep it fairly uniform between the two.

 

[ChaserFJ60]

I think so too. I wish when I had the tank off to replace my fuel pump a couple years ago that I would've had the foresight to measure the length of my fuel pump intake from the flange bottom compared to the tank depth and done a "fill test" to measure the quantity that would be left before the pump would start sucking air. 4 gallons of reserve before reaching that point would definitely be a good safe margin I think. It's something else I'll have to check during my trial & error process.

Interesting that your gauge needle can go down to the bottom line of the E compared to mine, which I snapped the "bottomed out" picture of with power off. I forget, is your rig a 60 or a 62? Mine's a 62, but I have a 60 quad gauge in my garage that I'm curious to look at now and see where the fuel gauge is sitting without power. If there's a variance between the 60 and the 62, I might have to use the "half of capacity minus reserve" point to set the 1/2 indication that I mentioned earlier to keep it fairly uniform between the two.

Its an FJ60, There may have been a large allowable tolerance in the needle travel as the gauge is vague in its precision, I wouldn't be surprised if 10 cruisers had 10 different positions all in the same relative area for E and F.

 

[Spook50]

Its an FJ60, There may have been a large allowable tolerance in the needle travel as the gauge is vague in its precision, I wouldn't be surprised if 10 cruisers had 10 different positions all in the same relative area for E and F.

Good point! Definitely another good reason to have adjustability in my design.

 

[CruiserTrash]

I have a 60 pickup tube in my stash I could measure. That would necessitate measuring the depth of the inside of the tank to have any meaning though.

Also, here’s a PO horror story: the send the return fuel lines were swapped where the hardlines stub out at the engine bay on my first 60. Until I figured it out I would run out of gas when there was still plenty of gas in the tank.

 

[Spook50]

I have a 60 pickup tube in my stash I could measure. That would necessitate measuring the depth of the inside of the tank to have any meaning though.

Also, here’s a PO horror story: the send the return fuel lines were swapped where the hardlines stub out at the engine bay on my first 60. Until I figured it out I would run out of gas when there was still plenty of gas in the tank.

Oh yeah I remember reading through that thread. At least it was a fairly simple fix for that one once you found it.

So it would appear that the voltage regulator is now too low, as my second gauge picture in post #4 is where it sat (after adjustment) with 4.25 gallons of fuel in the tank. here's where it sat after topping off the tank this evening:

Also for reference, here's where my temp gauge was after a hard 20-mile run on the freeway with the AC going in 85° temps. Granted this is just slightly under what I would consider normal, but I know my engine is running warmer than it should be due to I suspect just overdue cooling system maintenance (though I suspect now it's really not as close to overheat temps in 90+° temps like I had considered before). To guesstimate, I expect it should have been sitting right at the third line up from C or perhaps right at the upper edge of that line.

I'm suspecting that there's very little reserve designed into this system and when the gauge is on E, you're sucking air. Seems odd that Toyota in the late 80s would do that, given a lot of other automakers designed a small reserve into their systems. Then again I could still be wrong. I just won't know until I start fiddling with actual resistance readings of the sender and compare those readings to what my gauge shows.

So my next plan of action is a bottom drain, then add fuel a half gallon at a time with a multimeter connected to the sender lead at the gauge connector in the dash and to a reliable body ground. Once I reach 120Ω and see how much fuel I've put in, I'll plug in my gauge and see where it's sitting. If it's bottomed out, I'll dial up the regulator until the needle moves and then slowly back until the needle just barely rests on the lower limit. Then another highway run and fill-up and see where my gauges are at that point. This will also tell us if there's really any reserve designed into the system once the gauge is on E.