Gas gauge calibration unit

-
A micro-controller seems like overkill, but I know some are so cheap they cost less than a precision resistor. However, that only works for high-volume automated assembly, not people handling soldering irons. I bought 2 electronic IVR's for <$30 ea w/ shipping. Search "voltage limiter" on ebay.

I agree with others that what we really care about is "where exactly is empty". I can remember that I filled up for at least a few days, so no big deal whether the needle is exactly at "F". It is weeks later when I need to know "how much left?".

To adjust zero, you can add series resistance if it reads too full, or add parallel resistance if it reads "E" too soon. Both are fairly easy and could be done at the kick panel body connector. To also adjust the full reading, we need "another knob". One way is to booger with the IVR output voltage. I haven't looked close at my electronic ones, but might be possible by adding series resistance or changing the "sensed voltage" feedback.

However, above is beyond most people or not worth their time, so I can see where a store-bought box would sell well if the price is right. It would be fairly easy for a user to install such a box in the kick panel (though might need 12 V power there). The same basic box might work for many classic cars, but would need different connectors. At least that could get the PCB volume high enough to reduce unit cost. I am surprised those clever Chinese haven't done so already. If they can put a computerized voice in a greeting card for $2, they could build an affordable "fuel gage fudger" box.
 
I found I could not load a phone video to youtube greater than 20 mb, so I made a short one. It has to do with email to youtube upload url.

That pot easier for me now than a sender, much less table space.

Now I am working on routine to blink LED to indicate cal modes, and way to enter and exit cal modes. I getting the job done, but my code is fairly crude. I am gaining experience working on a user interface consisting of only two buttons and a LED.

I may need help to find a better way. I will post when I have something that works and perhaps someone will have a better idea.
 
Bill,

Not trying to put you down, just trying to explain my reasoning.

The little 50 cent 8 pin micro is fairly simple, no xtal and other baggage. Now days I use them instead of op-amps. I typically do pressure sender conversion for offset and gain, for less than $1 in parts, no precision components or pots. The circuit is low power and is powered by the 5V sensor supply. The conversion fools PCM to do control operations for special performance applications, involving common rail diesel fuel pressure regulation.

In this circuit there are added parts, a current source to feed sender, IVR, buttons, LED, and MOSFET to drive gauge. It will cost a few dollars, but similar in concept.

This is just an exercise, to see if it works. It is not a plan to make money, or sell them.
If it works I will program chips for $2, to save people the expense of a chip programmer dongle and compiler. I do not plan to build circuits.

The idea of using series and parallel resistors to me seems more difficult to implement. Adding a resistor in series or parallel changes both ends of scale. It can be elusive for some, when only working on a single condition at a time. One would need to characterize the gauge, and sender, then work the numbers. Likely the sender will be diluted by the introduction of resistors, so the span will be decreased. To fix that would require gain increase. If the IVR voltage is increased, the in-tank sender sees more current, that could pose danger. The OEM system already has a current of ~167mA at full tank.

The concept of the calibrator is simple, it merely re-scales sender values of Empty to Full to gauge values of Empty and Full. The sender endpoints, and PWM endpoints are stored in non-volatile memory inside micro in the calibration process, in-car by pressing buttons. Since the sender resistance is taken out of the gauge loop, the PWM can over drive a gauge, if that is required, without raising IVR voltage. In the OEM system the 10 Ohm of sender Full, reduces gauge drive to ~67%. Mathematically the gain is adjustable, that is simple and convenient.
 
Gil, Kudos to you for giving folks another option. I know when I get to that stage in my build I will have to do something, just wasn't sure what it was going to be. One of the issues with the existing mechanical IVR and I don't know about the solid state ones is the lack of meter dampening. It drives me nuts to see the fuel gauge bouncing up and down. As I recall my original 69 Dart Swinger 340 did that when it was new and I just chalked it up to the float bouncing up and down. When I did a ebay search the only device I saw was $52. When it comes time to start on the electrical, $52 dollars for a part the looks like the original, and is plug-n-play might be worth it as well might be the $30 dollar part that Bill refereed to. All of that said every thing I have read so far seems to indicate the device is just a 5 volt regulator. If that is the case, 5 volt regulators are available all day long for just a couple of dollars. That seems to do nothing as far as calibrating the high/low/mid point. I just retired after 30 years with Motorola and a strong technical background and while my hands might not be steady enough to hand solder the micro circuit paths use in today's modern circuits any more, I might well be able to handle a bread board device. I truly appreciate the thought and R&D you are putting in to this as well as giving us another option. I do like the idea of a low fuel warning light, it has become like a crutch to me. I look forward to reading more and have subscribed to this thread.

Keep up the good work.
 
Bobby,

Thank you for the kind words. I have been called many names, and it does not matter a bit. My real name is David, but the Kit is from working on Kitcars.

Back to the board. Designs have shifted to surface mount. Board costs are closely related to holes drilled, surface mounting eliminates holes. The surface mount parts take less materials and the manufacturing process is automated. The resistors and capacitors are very inexpensive in quantity, just cents. Through hole parts such as axial leaded resistors and radial leaded capacitors are now expensive, and hard to source.

I will try to use 1206 (0.12" x 0.06") and 0805 (0.08" x 0.05") parts. Even the 1206 are not easy to source, since they are considered old technology. Typically the parts are handled with tweezers. A small dot of solder is used to plate one side of foil, then the part is moved in-place and soldered on one end, then the other end soldered. I like to de-stress by quickly heating both ends, by moving back and forth. The part floats on surface, attracted by the surface tension of the solder, and centers on the pads.

My prototype is ugly and uses a an inexpensive phenolic single side foil from RS. It is what I had in my junk box. If we get board made they will be FR4, have solder mask, and silkscreen reference designators for parts locations. Assembly will be fairly easy.
 
I have written some code for the calibration process.

The calibration process, with the help of the user, grabs the end points of sender resistance, and pwm drive, for empty and full tank. The points are stored in eeprom for the re-scale function. The eeprom stores the values, and does not require power.

The user interface consists of two buttons, one marked E, the other F, and a LED. The calibration will likely happen once, at empty tank for empty points, and again when filled.

The button pushes are as follows:
Press both buttons, about one second, the LED will blink, release. This is entry to cal mode. The E, F button clicks are used to change gauge reading to desired value. E moves towards E, F moves towards F. The resolution is about 0.5% per click, so it might take a few to notice a change. This is done slowly to give the gauge time to react. Bi-metal gauges are slow. If at empty, make gauge read empty, then press and hold E button to exit. Similar for full, do at full tank, except F button used for exit. In both cases, the final click is removed from adjustments. The LED stops blinking, and the gauge function follows sender. Re-calibration is only necessary again due to gauge drift, sender change, or an EMP that clears the eeprom and renders all modern cars dead on road.

I think this way is fairly good, but perhaps a more creative person can visualize a better way. Please help with comments.
 
Geez Dave, sounds to me like you have this deal handled. If you can lay out the "drill" so people can easily handle it, this might be a goin deal

Do people like FARR still print boards?
 
Del,

I do not know about FARR still print, what is that?

There are places to submit PCB artwork, they collect a batch, then they send to China. The cost is reasonable. I am sure the artwork becomes public domain.

I can layout the PCB for through hole parts and SMT on same board, perhaps that would be the best, people would have the choice. The SMT is so small they will easily fit under the TH parts.

The other option is, someone like you that can build a nice circuit from schematic, using a standard protoboard, RS parts, and document so others can do the same.

But first, we need to test, and make sure it does what we want.

I want to do more testing. It works for me doing the right things, but need to test for doing wrong things. I have been thinking about doing a 2 minute vid for the cal and show how it works.

I would like to send it off to you for testing on your gauges. Play with cal ....

I changed the LED blink in cal mode to ON, so the blink mode is used for low tank. I have that set for 20% now.

I did not have a logic level mosfet for the gauge drive, so just using a NPN power transistor. Either will work, pin outs the same.
 
Turns out is is "Far," LOL

http://www.farcircuits.net/

He / they make circuit boards for some of the amateur and electronics projects

I'd be happy to. PM me if you need my address again
 
Dave, I was half joking in one of the earlier but the older I get themore crutches I need. How much harder to add a low warning indicator at say 10% full. I have become so accustom to them that a lot of the time I don't even notice the gague until the light pops on. Funny, when I was young I didn't want idiot lights now hmmm.

Bobby
 
Bobby,

KitCarlson said:
I changed the LED blink in cal mode to ON, so the blink mode is used for low tank. I have that set for 20% now.
.
Click to expand...

The low fuel is set at 20%, I have it that way in code. To set with buttons, more complex. Typically with in-tank electric, or tanks with some crud, it is best to keep them above 1/4 tank. The number can be changed, perhaps by a user poll.
 
No 20 % is great. The 10% was just a arbitrary number. I must have missed the part where you said it was going to have a low warning sensor. You can stick me in line for a board if you do one, even though it will be awhile before I need one.
 
Bobby,
I had to look for the LED for low fuel, it was in post #2. It took awhile.

I do most of my work when I sleep, and have a difficulty separating real work. Whlile the design is nearly done, a few nights of sleep may identify problems and improvements. When I was young, I could think on my feet.

While I get things done, my behavior would not go well in the work place. Now days meetings are where problems are solved.... Not sure that works well with electronics.
 
Pawned said:
I know I am sounding argumentative, but my question is where do you think your switch mode regulator or for that matter any electronic parts you plan to use are made. Grand Rapids, Minnesota???

There is no gray market involved. I ordered two boards, I screwed up the first board because my arthritic hands do not solder like they use to.
The board is so nice I ordered 5 more. In case someone local needs one. I will probably give them a board. For $1.15 I can be a spender
Click to expand...


It took about 40 days to get these 5 boards for a total cost of $5.75. That is why I ordered so many. It is very common for me to purchase more than I need if it is not available locally.
So far, my fuel gauge is working to specs.

I am still waiting to hear how much Mr. Kit's assembly is going to cost.
I would also like proof that it works

Ed
 
Pawned said:
It took about 40 days to get these 5 boards for a total cost of $5.75. That is why I ordered so many. It is very common for me to purchase more than I need if it is not available locally.
So far, my fuel gauge is working to specs.

I am still waiting to hear how much Mr. Kit's assembly is going to cost.
I would also like proof that it works

Ed
Click to expand...

Stop trying to turn this into an argument over ego, which is exactly what you sound like. This is about different approaches to a problem, and in your case, I see no way to allow for differences in gauge unit and sender unit differences from car to car.
 
67Dart273 said:
Stop trying to turn this into an argument over ego, which is exactly what you sound like. This is about different approaches to a problem, and in your case, I see no way to allow for differences in gauge unit and sender unit differences from car to car.
Click to expand...


No ego is involved, I want to see his fix in action. When will there be a working model?? If it is better than mine, I will be the first to pull mine out and use his.
I still want to know how much it will cost.

Whether or not it works, may have been a slight dig, I had mentioned that I was working on the problem and refused a number of requests to tell anyone my approach and or costs, until I got a working model. I also gave my work to others for free, only wanting a Thank You as payment. I did not say I wanted a royalty for the work I put in.

I found my first working model needed some improvement so I improved it and actually found it dropped my cost. Less than $1, but far be it from me to harp on that.
 
Your last comments only reaffirm my opinion. You seem to me to have an attitude about his. Dave has not asked for any money, and his schedule certainly does not mark to the beat of either your or my drums. What I'm asking here, is that you tone this down. You seem to me on edge, here, about this whole deal.

This is NOT an "either or" deal. This is two different guys working on the same problem from different approaches.
 
I have been busy on an engine rebuild project. I am getting closer, I got carried away working on heads for about a week. I forgot how long it takes to do valve job, and balance combustion chambers. I hope to install in a couple days when the weather warms up.

The unit was completed 2/6/14. I did a few calibrations for various sender resistance ranges and gauge current requirements.

My next step is to box, and send out for Del to try. I need to make some documentation for testing/calibration procedure. It has been long enough it will be all new to me again. I will test one more time, to check the documentation.
 
The weather here turned cold, so I found time to get back to the project. I sent it off to Del 67Dart273 for testing with real gauges.

I am off to build a simple injector flow bench, with hopes of restoring few old injectors. I will start a new thread for my progress of "DYI injector cleaning and flow testing". It will be posted in the "Fuel and Air Systems" section of forum.
 
Kit,
I will look for your injector tester. I built a pop-tester (2000 psig) for my diesel injectors w/ an electronic sensor and data recording. I hope to build a rig for my gas car injectors and will be interested to see your design. Most I have seen use a simple on-off switch, sometimes off a timer, and an array of graduated cylinders, though a single one should suffice. I expect your design will involve a microprocessor and perhaps an electonic level detector (pressure sensor).
 
Here is an updated schematic. I moved a couple pins around for future flexibility, in controlling the LED. The LM7805 regulator could be replaced my a switch mode regulator set at 5V like Pawned used if one desires, other wise the LM7805 will need a heat sink. It is also used to power the micro, so the voltage should not exceed 5V. The gauge will still drive higher if necessary, since there is no series resistance (10 Ohm minimum sender) in the circuit.
 

Attachments

  • gauge_cal_7805.pdf
    16.8 KB · Views: 245
You must log in or register to reply here.
-

Back
Top