Sometimes when I drive the gas gauge moves up...not bouncing around (it does that too), it will just move up a little over a period of a few miles. It also will stay on a spot for a while, then suddenly drop. I expect this on full due to the way the float works, but it will do it at 1/2 as well. It also bounces around some...I thought something was susposed to compensate for gas shifting in the tank? Apparently not. I hate my gas gauge! At least it sort of works (the one in our '86 Ranger didn't...we just used the trip odometer).

 

Uh, yeah, don't take the gas gauges (or 'gages', if you have a late model w/ the idiot light, like mine) all to seriously. The float in the tank itself is one place error is introduced. You may notice that if you park nose up on a steep hill you'll have more gas then if you park nose down, or maybe it's the other way around. Point is the float isn't too resistant to having the truck at an angle.

Second, the line itself is analog. It is 'filtered' w/ what we EE types call a simple RC circuit. Point being is that the reading is not taken right off the float and displayed on the needle. If they did that, then the neddle would go all over the place every time you hit a bump. Instead it is dampened by an electronic ciruit. Now the needle doesn't jump, instead it slowly moves from reading to reading. You may notice this when you fill the truck up and first turn on the key. The gauge starts near empty and slowly dials up to full.

All this is completely normal. The only way they could have made it better is to make it significantly more expensive. Some vehicles out there these days use the computer to average out the levels more intelligently. This would have made your truck more expensive though.

I know what you mean though. If the tank is at the right level, I can get the needle to move nearly 1/4 of a tank simply by leaving a stoplight quickly. Don't trust the gauge, only use it as it was intended: as a rough guideline..

 

mine reads empty when i have about 6 gallons left... whatever at least ill never run out

 

mine doesnt work
the float has a hole in it from what ive gathered talking to the ford mechanics

 

My first Edge had a normal gas gage, but when I got the Edge I have now I found that my gas guage wouldnt move while driving really. I would have to sit and wait a minute and turn it back on and then it would read the correct amount.

 

I reset my trip odometer every time I fill up. My '84 had a dead fuel gauge so I had to. My '91 had a dead fuel gauge so I had to. My '03 has a gauge that "functions" but it doesn't do much for than give you an "rough estimate". I have 4 gallons left when the "CHECK GAGE" light comes on.

 

The sender in the tank goes to the PCM (powertrain control module) and the fuel level is displayed in percentage by the FLI (fuel level indicator) PID (parameter identifier). The IC (insturment cluster) module talks to the PCM and the PCM tells the IC what to display on the fuel "gauge". This varies somewhat by vehicle depending on the type of insturment cluster, but in general the gauges are not connected to anything but modules in the vehicle. It's been this way for a while. You'd be surprised how much of your vehicle is actually controlled by modules. And you thought you were in control!

 

I was told otherwise. Someone in one of these forums looked it up in the shop manual CD-ROM and traced the line directly from the sensor in the tank to the instrument cluster. I'm sure the PID for the the level is there, but it was my understanding that it was not used in the Ranger. The fuel level is one of those things that hadn't quite made it under PCM control yet in the Ranger. I'm sure it's different for other newer-designed vehicles.

This was my information on the Ranger as of the '03 model year. '04 saw a slight redesign of the instruments, so it could be totally different. Is this not the case? Are you sure the PCM gets fuel level info in all late model trucks?

If you are correct, and the PCM does have such info. Can you speculate as to how accurate the PCM's data is. That is if you were to directly poll the PCM, rather than read the fuel gauge, would you have a better estimate of how much is in the tank?

 

I was referencing an LS, which is what I have in my bay at the moment, and thought that they were all working that way by now. The LS, and most other Fords work the way I described but, it turns out that the Ranger is still using the ancient method to display fuel level. So you are correct. I checked on an '04 and found there is no command for the fuel gauge like I was expecting there to be. Ranger still uses an IC module and most others use a HEC module (hybrid electronic cluster).

For the Ranger, one side of the fuel guage is connected to VREF which is supplied by the PCM. The other side goes through the float and whatever fuel level is there supplies whatever pull down voltage it does at that level. This brings the 5 volt VREF signal down to say 0 to 5 volts, 5 being full 0 being empty (these aren't the acutal values). For the diagnostics end of things the PCM supplies the voltage gained through this process to the OBD tool whic does an A/D conversion to display a percentage.

Sometimes I forget how old Rangers are in the inside.

 

To answer you other question... If you took the cluster apart you could mark percentages on the gauge face, or voltages I suppose. The tool only allows 10% increments. You could do the same for coolant and other gauges. Or you could just buy the tool for $10,000 and monitor whatever PIDS you want while you drive.

 

So the gauge is analog, and the fuel level indicator gauge basically boils down to a voltage meter, correct? This is probably why it is so innacurate. But you say the line IS ALSO sampled by the PCM. I buy that.

Yes, the Ranger is stone age in many respects. The oil level indicators which are actually idiot lights in gauge format strikes me as one example. I'm sure that LS has a nice little computer that computes estimated gas mileage and such, just like your TT. Such features are undoubtably a long time off in the poor cousin Ranger..

Or slap a $100 scan tool on your laptop and write a tool to scan the PIDs for you for far less. Better yet: mate a scan tool device to an innexpensive embedded processor and get it to do the same thing. I have parts for such a project sitting on my work bench at home. I had thoughs of using a Z80 or PIC based microcontroller to query the PCM via the OBD-II interface and display the results on a small LCD, which I'd embed in the dash. John Griggs ('N3ELZ') and I discussed this project quite enthusiastically nearly a year ago. As I say, I bought some of the needed parts, but never quite got around to implementing it. Instead they collect dust on my workbench..

 

The fuel level gauge is a voltmeter, yes. So are the rest of them for the most part. They just look nice in a cluster even though they are totally useless in reality because they tell you nothing usefull at all.

You could do what you suggest but you would have to build it to be vehicle specific. There are many different protocols out there, even within the same vehicle. Databases change all the time. There are probably hundreds of people involved in the diagnostic tools testing and development. That's what it takes to manage the whole fleet. One vehicle one module, which is what I think you meant, would be much easier. It would still be a bit of work unless you could get a head start by using what someone else has put together for PCM reflash reasons.

The LS is an '05, that's all I'll say about that.

 

I think the oil gauge is worse: it is a simple logic indicator. It goes to 0 when no voltage is present, and to a pre-defined 'neutral' point when voltage is applied. It's my understanding that both of the needles on the right side of the pannel are essentially idiot lights. They might as well be LEDs!

I really don't care about the LS. I'm not even sure what it is. I assume it's one of the luxury Lincoln cars that I'd prolly never even be interested in. But thanks..

Yes, I meant a unit to work in one specific model, the Ford Ranger. Better yet, one specific unit, mine! I bought a low-buck scan tool that interfaces via an RS-232 port to a PC. I poped the cover and tooled around w/ a continuity checker. Didn't take long to figure out how it worked, they don't exactly hide it. See www.scantool.net, they even supply schematics and links to the suplier of the chip they use here (link). I've since determined that the 'ELM320' is actually a $0.35 8-bit PIC microprocessor flashed w/ their code and w/ an adhesive sticker on it! John one upped me, he bought the proc and cords seperately and peiced one together using spare parts he had lying around. My parts bin isn't as deep, so I just ordered it pre-built.

As I see it I have two options. I can reverse engineer the ELM320 and figure out how to incorporate the PWM conversion in my own micro, along w/ the pooling, statistics, and LCD interface code. Or I can yank the ELM320 off and interface it to my micro using an RS-232 interface. This is far easier and probably what I'm going to do. I have no need to program the PCM, I just want to read from it, so my task is somewhat easier.

Either way, you're right, it's a lot of work to check the fuel level. But that wasn't really what I was after. I want the unit to scan several 'PIDs' and display real-time data on engine RPM, speed, etc. I had hoped to figure out a way to compute estimated gas mileage and range in real-time as well. Like that fancy display you have in your TT..

 

The funny thing about the gauges is that even if they were highly calibrated insturments, what would they tell you? My Cobra used to run a coolant temperature of "M" when it was really hot out. but it was usually between "N" and "O" the rest of the time.

You are correct about the oil pressure gauges, they went that way somewhere in the early 90's. If the pressure is above a certain point the gauge is always in the same place. It only has two positions, the place it normally is in the middle and the bottom. They got rid of that for the '04 Ranger though, no "gauge" at all.

While I would like to know all the things that are going on with my truck in real time, it really doesn't seem worth the effort to me. Once you got the tool up and working your truck would get stolen or totalled and then you would find Ford switched from SCP to CAN and now your tool doesn't work anymore. If it is fun for you though, then I say go for it.

The LS is a Licncoln, which is really a Jaguar, which is also the Thunderbird and even the new Mustang.

 

Well, obviously it isn't all that worth my time because the project has been sitting on my bench for nearly a year! Still, it seemed like something I could pull off fairly easy. I was able to piggyback an order at work for the parts, and I had some leeds on the info I'd need to make it happen, so it didn't cost me much.

And I don't have vehicles stolen or totalled all that often, so I wasn't worried about loosing my investment of time or effort. I had planned to keep this truck for at least 8-10 years. That was before I had so many problems w/ it. Now I'm unsure if I'll keep it once the warranty runs out.. What I can afford to do is another issue of course.

But SCP? CAN? It's my understanding that ALL domestically produced Ford vehicles use the SAE J1850 PWM OBD-II standard. GM and Chrystler use SAE J1850 VPW and the rest of the plannet uses ISO 9141-2 or ISO 14230-4. Even if Ford switched standards, that just requires a change in interface chip, which I can get for < $20 or hack myself. Hell, I found a post from a kid on the web not too long ago who had used some parts from an old TV remote to build an interface to his early 90's VW!

As mentioned, I never got that far along on the project. But I was able to use peice of software written by a hobiest like me to poll and log a few key 'PIDs'. Somewhere I've got a neat little excel graph of my engine's RPM, speed, and a few other stats on a quick ride around town.

 

You're right, I was thinking of module programming which you aren't trying to do. In that way protocol isn't important. To read the data you only need to be compliant with the OBDII port.

 

Now the catch is that it seems my OBD-II scan tool can have an address. Mine is pre-programmed w/ the 'standard' scan-tool address, and knows the 'standard' address that the PCM supplies all the standard OBD-II info at. The question is, can I program my embedded device to originate requests from a different address? Would the PCM still play ball? The objective would be to be able to leave my device hardwired permanantly into the vehicle and still have the OBD-II interface be fully functional. This way if I were to have the truck serviced by Ford or any other tech that would use a scan-tool on the truck, they wouldn't even know or care the device is there.

That probably wouldn't fly simply because I'd have two devices (mine and the tech's scan-tool) both trying to communicate on the OBD-II interface at the same time. Even if they were using different logical addresses, the comms would still collide on the physical interface, which I'd be perverting into a full-on bus. That is unless I built some sort of bridge or switch capable of dealing w/ collisions and multiplexing requests from several devices. But then that would be beyond the scope of my project. Disabling my device before having the vehicle serviced would probably be a simpler, more cost effective solution..

..This and several other challenges are things I never quite got around to solving!

 

I should be able to help you more with this type of thing at years end, at least answer your questions.

 

How's that? I know you work w/ this stuff already.

 

It's likely that I will be in a position that will give me more intimate knowledge of module programming soon, as well as the fact that I have some Ford specific training coming up.

 

yeah my gas gauge always seems to go down.

 

holy old thread batman!