On the subject of gas tank gauges reading correctly......

[Sir Sam]

I know you guys have some issues with the new/old gas tank float having different resistances.

 

 

I am going to have these same issues in my CRD project since I will use my original float with a 02+ TJ gauge cluster.

 

 

I am wondering if anyone else is interested in contributing to my solution.

 

I was going to make a programmable converter to make the gas gauge read correctly.

 

I was going to use a 16 pin PIC (16F88 probably). Connect two pins to the float connection.

(there might be other ways to do this that would work better, or are simpler, or provide infinite resistance)

 

The other 14 pins can be used to provide 14 different resistances. That is, 14 different levels of fuel.

(we might find some digitally adjusted POT too, or a different pic might even have that function, more research would be good)

That is 100% full.

93% full

86% full

78% full

71% full

 

etc etc etc

 

What that means on your fuel guage is that your needle will jump between 14 different settings, if more were desired we could probably arrange that.

 

In terms of cost, the pic I am thinking of right now costs about $4 each, 5V regulator, couple of adjustable POTs, couple of capacitors, solder wire etc...so it adds up, but for the basic components we could probably put it together for about $15-20.

 

 

Anyone one interested in buying in on some parts?

 

Any EE's have anything to contribute?

[Eagle]

One thing to keep in mind is, you need to verify if the meter scale is linear or logarithmic.

 

I don't know, but I have always believed that the fuel gauge, temperature gauge, and oil pressure gauge are the same instrument with different markings. Yet, if you look at the temperature gauge, the left side goes from 100 to 210 (delta = 110), while the right side goes from 210 to 280 (?) (delta = 90), in the same angular sweep. The voltage meter is also asymmetrical. Since the other three gauges are, essentially, volt meters, this leads me to believe that the sender out put is not linear so you may have some experimenting to do before you get your conversion box calibrated.

[ogdawg]

I saw this mentioned in the other thread, sounds like a good idea. I have mostly worked with the Cypress PSoC so I think I will try something similar with that.

 

Here is what I am thinking. Since the gauge is basically a volt meter use the PSoC to read a voltage coming from the sending unit with an ADC. Take that value and use either an equation or look-up table in software to convert the value so it is correct for the gauge. Then output the voltage with a DAC.

 

Once I get my truck running I will try it out.

[Sir Sam]

I saw this mentioned in the other thread, sounds like a good idea. I have mostly worked with the Cypress PSoC so I think I will try something similar with that.

 

Here is what I am thinking. Since the gauge is basically a volt meter use the PSoC to read a voltage coming from the sending unit with an ADC. Take that value and use either an equation or look-up table in software to convert the value so it is correct for the gauge. Then output the voltage with a DAC.

 

Once I get my truck running I will try it out.

 

Yes, that would be a way to mimic the voltage returned, and would work the same way. I was mostly working under the idea of mimicking the resistence of the sender, and letting the gauge do its normal thing.

 

There are probably some other simple ways to go about this too, I just really haven't thought of many alternatives yet, I mostly just want something easy to implement and get a working gauge.