When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
It is not a very good idea to try without removing the connector or cutting the wire to the pin 2. Applying +12V causes too much heat since the sender consumes 5.76 times more power (12/5 X 12/5 = 5.76). You may burn the sending unit doing so. I 'd simply remove the connector or cut the wire and apply the centre pin the DC using a voltage regulator or a potentiometer which can supply 0-4.38V. The way you explained may also cause insufficient voltage when the the resistance of the sender is too small. In this case there goes a big current to get 4.38V.
Agreed on not using 12V. I assumed that was a typo, and you meant 5V.
But before designing the fix, I'd recommend spending some more time on finding the problem. As mentioned before, you can access the connector in the trunk. Then apply voltage carefully, and see if the gauge moves as it should. If it does, that would isolate the problem to the sender, so you can proceed with resistance measurements there.
I might take the step at that point to remove the sender, inspect the float and the resistive surface on it to see if there are any obvious issues to fix. You can also measure resistance with the gauge at those locations shown in the manual.
On the linearity issue, your very first post, if I'm reading it right, suggests an initial nonlinear drop in the first 10 miles. No? You'll be able to measure linearity as much as you want if you can remove the float. You probably have a good data point on "full" since you can check it right after filling up. But especially in a case like this where you don't trust the gauge reading, you don't know when the tank is empty.
Once you have the problem located and quantified, if modifying the circuit can fix (or at least mitigate) it, that should be a pretty simple step, as compared to getting to that point.
I did a quick sketch & wrote down the wrong power supply voltage on my schematic. I meant 5V from the same 5V regulator. My mod from the previous schematic was also un-finished so attached please find the latest:
Your idea would be OK when R1 or R2 is not very small.
Depending on a resistance of R1 or R2, there exists a possibility of excessive current going through R1 or R2, you need to think about. For the safety, the centre pin would be better to isolate from the sender.
Yamae- R1 & R2 are the equivalent of the fuel sending unit's resistance which is 300 Ohms combined. Per design, the floater is calibrated so one of them will swing to 21 Ohms while the other will simultaneously swing to 276 Ohms & vice versa when the fuel level is at full or empty. As stated, since I only need to vary the resistance of either R1 or R2 a little to slightly increase or decrease the fuel gauge voltage, I'll be using a large Rbias potentiometer in parallel with either R1 or R2 & I've calculated to be around 1K Ohms, therefore the total resistance will be slightly below 300 Ohms so the ckt will only draw a little more than 16.7 mA from the 5V regulator. It will be fine. My concern here is will it be linear so the needle on the analog fuel gauge will display accurately throughout the entire range. I'll find out when I get a chance to play w/ it.
I see what you mean that R1 and R2 are not the resistance of the sending unit itself but external resistors for the experiment. And yes, your idea is just fine.
One thing I'd better to add. I don't remember exactly but the input pin of the AD converter shows rather high impedance. So without using R1 and R2, you can use a 10k ohms potentiometer or so alone to supply the DC to the AD converter connecting the top side to the 5V DC supply circuit and the bottom side to the ground and the centre to the AD converter = meter. The sender's resistance curve is almost the B curve which can be described as a linear curve. As you already know that a 1/2 tank should send 2.43V and this is almost the half of the supply voltage 5V. This simply means that the curve is almost linear.
Of course it would be better and more accurate to use a 300 ohms potentiometer but the high impedanced input does not require the big current and the improvement of the accuracy is just a little. You almost can neglect it. I still have some difficulties to explain complicated affairs in English but I hope you understand my points and you find the root cause of the problem soon.
Per Oldskewel's info obtained from the service manual, R1 & R2 are the resistance of the sending unit which is 300 Ohms combined (I'll verify the resistance once I disconnected the connector just to be sure). The reason I've stated they're the equivalent resistance of the sending unit's is because I want to show it as 2 discrete resistors in my schematic instead of just a potentiometer. Sorry for the confusion. Btw, your English is great & you didn't have a problem explaining it. Anyway, per the schematic & the troubleshooting from the manual that you & Oldskewel have provided, pin 2 (B12) is the V fuel gauge & I can pull it up to 5V (pin 3 or B13) or shorting it to ground (pin 1 or B5) to simulate the fuller than full or emptier than empty conditions, respectively. Based upon my observation from my fuel gauge's needle, I suspect the voltage at pin 2 will be slightly less than 4.38V when the tank is full & also less than .48V when the tank is emptied so the entire voltage curve is off. By adding a resistor, most likely in parallel w/ R1 (between pin 2 & 3) as shown in Fig 2 as I'll need to re-adjust/increase the voltage at pin 2 back to 4.38V (tank full), all I'm doing is shifting the curve up. The equivalent resistance of Rbias & R1 ((Rbias*R1) / (Rbias+R1)) will just be slightly less than the value of R1 so based on my calculation, Rbias will be in the range of 950 to 980 Ohm. Having a large Rbias like that, the circuit will only draw a couple mA more so it'll be alright. I'll keep both of you posted when I can get a chance to play w/ it. Thanks.
Oh and to answer Oldskewel's question about the fuel gauge's needle which I stated at the beginning of my post that within 10 miles after filled up, the needle dropped a quarter, I've exaggerated the condition a little bit. It actually only dropped about 1/8 & I think the reason is the fuel gauge is an analog gauge so its response time is slow & somewhat inaccurate. It showed full at filled up but once the level of the fuel & the float are settle down, the fuel gauge's voltage is slightly lower than the intended set voltage which is at 4.38V, therefore the needle dropped 1/8. Right now, it's all speculation so I won't know for sure until I play w/ it.
I still think you should spend more effort diagnosing things before getting too deep into the electrical fix.
A "mechanical guy," if faced with this problem would probably address it by first using his multimeter to prop open his toolbox. Then taking the sender out, checking for mechanical binding and free movement, inspecting and cleaning off any residue buildup on the resistive element, and if calibration were required, that would be achieved by carefully bending the float arm.
Not that those steps would necessarily be correct or better. But they might be. And you maybe don't know what the problem really is yet.
Also, on adding a resistor, as you can tell from Yamae's general responses, there really is an important safety concern here. This is the gas tank we're talking about. A safer option to consider for adding a resistor to adjust things would be to do it in series between R2 and ground. For example, a 30 Ohm Resistor there would shift the entire curve up by about 10%, and reduce the gain by a little less than that. Maintaining linearity, if it existed before (which the parallel resistors do not, I think). Also by adding a resistor _outside_ the sender circuit, like this, you would only be reducing any current flow through the sender in the tank, and if there were any failure such as a short circuit, it would be safe as well.
In term of adding a resistor, having a 30 Ohms in series w/ R2 will also shift the curve up by 10% & the ckt will draw less as I = 5/330 = 15.2 mA instead of 5/300 = 16.7 mA thus less power consumption, however adding a parallel resistor in this case would be much easier. Anyway, before I do anything, I'll take a measurement at pin 2 when the tank is full then make an assessment & decide the next step. Per your suggestion, I may open the tank & take a look at the sending unit & the float. If there is any signs of degradation, I may go ahead & replace the pump & sending unit if I decided to go that route.
Btw, here are a few pics of my recent acquired 2000 LS400. Even though, she's 17 years old, however she's in pristine & like new condition w/ only 46K miles. The passenger & backseat appear to be brand new. She's a keeper for sure.
Btw, here are a few pics of my recent acquired 2000 LS400. Even though, she's 17 years old, however she's in pristine & like new condition w/ only 46K miles. The passenger & backseat appear to be brand new. She's a keeper for sure.
Hello. I just purchased a 1998 LS400 in April. Engine and transmission are flawless. My fuel gauge was working perfect until recently. I filled the tank and it went to its usual full capacity. I noticed after 50 miles it didn't go down. After driving it more and toward emptying the tank it has dropped and now sits at half mark. I noticed the low fuel light came on but the gauge sits at half. The trip computer does work with range and miles since refuel. I took the car to a Lexus dealer to have it looked at and they claimed it was a fuel sender issue. Dealer wanted $850 to replace the fuel sender. I purchased an OEM part myself for $150 (including shipping) and I'm going to attempt to replace it myself but I am having doubts. If it is the fuel sender then how would it still be able to set off the low fuel light? Is it possible the instrumentation is stuck? I should mention my temp gauge reads above the middle mark. And the RPM and speedometer get stuck on occasion but are freed by tapping the top of the dash. Any advice would be a great help.
Last edited by WaveRunner; 09-21-17 at 11:32 PM.
Reason: adding detail
I'm not convince your fuel sending unit is bad. Does the needle moves to empty when you shut off the car? If it doesn't then it's probably stuck. I'm sure you have tried tapping the top of the dash but you may want to pound on it a bit harder. Hopefully, it will free itself. If not, you'll have to remove the cluster & take a look at the fuel gauge. If you have some basic electronics, you can use an external power supply & vary it between 0 to 5V. If the needle swings between empty & full then you know the gauge is working. By then, you can proceed to replace the fuel sending unit. You'll have to remove the backseat to get to it but make sure the tank isn't fully full. It's a pretty easy diy job. Good luck & let us know if you need additional help.
Yamae & Oldskewel- I'm still working on a few hot projects at work so have no time to play w/ the fuel gauge but noticed it is off over the entire range so here's my plan to fix it when I get a chance. Assuming it's linear which is a very good assumption since the fuel gauge is off over the entire range, I can add another potentiometer (Rbias) in parallel w/ the sending unit's 300-Ohm resistor then also connect its center tap to pin B12 (option 1), however by doing so, I'll have a total of 4 resistors (2 from each pots) to play w/ so rather than that, I'll just add the Rbias potentiometer or even a fixed resistor from B12 to ground (option 2). This way, instead of having 4, I'll only have 3 resistors to play w/ so it'll be a bit easier to calculate the value of Rbias & set the bias voltage (either option will work). Since I can't set the fuel level in the tank to be exactly half & also don't want to take a chance of running the tank all the way down to determine the empty level, what I'll do is to fill up the tank at a gas station. Once the fuel pump at the gas station clicked a couple times, I'll considered the tank is full. I'll then check & re-bias the voltage at B12 to 4.38V (I suspect it'll be less than 4.38V prior to the adjustment as it's off). Assuming it's linear, the resistance of the sending unit will increase to 145.8 & 276 Ohms & the fuel gauge voltage will decrease to 2.43V (half full) & .48V (empty), respectively based on the position of the float as the fuel level continues to drop. What do you think?
Dear friend ,you have really did a very good job ,its scientific,recently I bought a 09 IS300 (same body of IS250 - IS350) V6-3000 cc, I had a problem that the gauge range is approximately 50% to full tank so I cant know when my fuel will finished , I have checked the wiring diagram and find out I have two fuel level senders not one ( one in the left side under the rear seat with fuel pump - the other in the right side).I have measured the resistance of both sides senders ,the left approx. 100 ohms , the right side is about 175 ohms both of them are connected in series and connected to the driver gauges.I ve just made some maintenance and cleaning to the connection and I wish my problem is solved .Unfortunately I cant drive because of COVID19 . two senders connected in series
I wish this may help you or other friends
wish me luck Firas
05-12-17, 05:59 PM
