This is the diagram of the gauges... Maybe that would help?
Butch

 

When your engine is cold, does the temp sensor indicate that it’s cold, or is it always showing the hot temp? How are you testing the temp sensor on another car? If we are talking about the gauge that goes to the dash, it works off the the yellow/white wire, according to resistance/. I don’t understand how you can test your installed temp sensor on another Samurai?? Regarding the fuel gauge, if you disconnect the fuel gauge connector at the rear of the does the gauge drop down to empty??

 

It shows cold and then slowly moves to the top of the gauge.

I tested the sensor on another car as in I pulled the sensor out and put in another samurai and it worked fine. 3 different sensors, so the sensor is not the problem.

 

Attached are the relevant pages from the Samurai service manual, there is enough detail in there for you to verify that your gauge senders have the correct resistance, if they do the problem will be in the gauge cluster, which, I'm 95% certain it is, and specifically the voltage regulator internal to the fuel gauge.

 

Is there any way to replace the voltage regulator or repair it? These clusters seem to be pretty rare/expensive.

 

I've never taken one apart, so I can't give you a definitive answer - my temptation is to say yes - but be aware that this will require some amount of "butchery" of the cluster, as well as some degree of experimentation by someone with a fair degree of electrical/electronic experience.

I want to make it very clear that what I am about to suggest is based on working with instrument clusters that were not from a Samurai, and it is simply how I would approach it.

I'm guessing the regulator is short circuited and applying the full battery voltage (approximately 12V) to the gauges - the first thing I would do is get an idea of the current the regulator needs to supply, I would do this by inserting a milliampere meter between the temperature gauge and the gauge sender, measure the current with the engine warm, next, insert the milliampere meter between the fuel gauge and the gauge sender, measure the current with the tank full, add these two measurements together and then add 25% of the total - this will be the current that the regulator will need to supply.

The next step would be to build a variable voltage regulator to power the gauges - I would probably use something like an LM317 IC regulator (if the current is less than 1A) - rather than building, these can be sourced out of China fairly cheaply...

Wiring it up - if I'm right, the back of the gauge cluster has some sort of printed circuit board - the trace powering the gauges will have to be cut to isolate the gauge supply, the original supply side can be wired to the input of the regulator, the output goes to the gauge and I would set the voltage at around 8V or so, and then test and tweak the voltage, it's going to be a time consuming process, the gauges are designed to have a slow response time, you don't want the fuel level to be constantly varying as the fuel in the tank sloshes over every bump & pot hole.

Drive from the full tank to an estimated half tank - or run it until empty and then add half a tank of fuel and adjust the voltage so the needle shows half a tank - if the voltage is too high the gauge will read high - you're going to have to adjust a little, wait a few minutes and see how the gauge responds.

One last thing - I don't guarantee that this will work, but, the chances of damaging anything is very small - essentially the cutting of the trace on the circuit board.

5pcs LM317 DC-DC Converter Adjustable Linear Regulator Step Down Circuit Board | eBay

This compact and easy-to-use module is perfect for use in a wide range of applications, including solar panels, wind turbines, and other renewable energy systems. Current output : 1.5A. >Input mode : VIN input positive level.

www.ebay.com

 

If it’s anything like the Toyota mechanical voltage regulators which used a heat source to turn 12v to 6v then hopefully I can just clean the contacts. If it was supplying the full 12v then I feel like the fuel gauge would peg.

 

I suspect it is a mechanical regulator, but it's internal to the fuel gauge and you're risking damage when you open that - how much of a risk I cannot say from a distance, but you sound like you have some experience, so, go brave.

 

Neither the schematic nor the pages from the service manual indicate that there is any sort of voltage regulator involved--what's the source of that idea that there is one?

2 different problems: one is with the IP temperature circuit, the other is the fuel level sensor.

If anything, the schematic is showing a heating element that would open the contacts if the current in the sensors was too high. Maybe dirty or welded contacts.

 

One thing I just noticed is that if I turn the key on in the ignition, but don’t start it. The fuel needle seems to show an accurate reading of fuel, as the tank is pretty much empty.

 

I'm really enjoying this post... I'm a retired electrician but have almost no knowledge about electronics.. My '86 Samurai has a fuel gauge that is not accurate, full or empty... I pretty sure it's a grounding problem, but haven't gotten around to checking it out... Sorry to detract from the original post, but I just wanted to mention my fondness for these discussions..
Butch

 

A bad ground on the fuel gauge WILL cause it to read lower than it should, this one is reading a 1/4 tank when empty, which is higher than it should, so the reverse of your situation.

I don't know to what extent an electrician will work with resistance, but, the gauges are very simple resistive circuits, have a look at the pdf I upload earlier, and if you need further explanation, open a new discussion and we can explore there.

 

Will do Fordem… I didn’t want to derail this post… Thanks
Butch

 

From the other thread with this same issue, there appears to be a "7V" connection on both gauges. Maybe that is the "regulated" voltage that originates within the fuel meter?

The fuel meter has 4 connections and the temperature only 2. TU routes to the temperature sending unit, and FU is for the fuel level sending unit.

If the 7V was actually reading higher than 7 volts, then the gauges would misbehave as you are seeing.

 

From the other thread with this same issue, there appears to be a "7V" connection on both gauges. Maybe that is the "regulated" voltage that originates within the fuel meter?

The fuel meter has 4 connections and the temperature only 2. TU routes to the temperature sending unit, and FU is for the fuel level sending unit.

If the 7V was actually reading higher than 7 volts, then the gauges would misbehave as you are seeing.

View attachment 119509

That is my assumption aswell. Wouldn't be that difficult to rig up a solid state regulator with a couple capacitors and one of those micro voltage regulators for a couple bucks. Just have to not break more things on the way lol.

 

It looks like there may be some resistors soldered to the underside of the green flex board--check to see if they have blown.

 

Hey there, my 89 Sidekick has issues with the entire cluster, no odometer, no speedo, temp gauge unreliable, and gas gauge always showing 1/4 tank or completely empty, my 4WD light comes on and glows glows brighter depending on my RPMs even when I'm not in 4 wheel, and only mRPM gauge works correctly. I was told: These vehicles are some of the most HAULED in America, and in order to reflect true mileage on the motor, the cluster gauge is made to be disconnected while towing, also the gauge gears are plastic and wear out after the years. Some dont get reconnected, some get reconnected incorrectly, and some are broken or worn out during the practice. I do not have a "fix it" answer, but I do know our odo/speedo's are exempt from safety inspection. Good luck

 

chalecelauren - first up, Sidekicks & Samurais are different vehicles with different clusters, and moving on, don't believe everything you're told - the gauge cluster is no different to any other gauge cluster when it comes to being disconnected & reconnected - no one in their right mind would even try to pull the cluster to disconnect the cable at that end when it's so much easier to do at the other.

If you want to prevent odometer from recording the "towed distance, the easiest way is to get under and pop the speedometer cable loose from the transfer case, there used to be electrical disconnects to make this easier - the speedometer is purely mechanical, disconnecting it has ZERO impact on the electrical side of the cluster - the tach and the fuel & temp are NOT affected.