# fuel gauge project

Discussion in 'Electronic Projects Design/Ideas/Reviews' started by scopeman, Nov 23, 2009.

1. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
fuel Gauge

Hello guys, Is there still a possibility of getting this puppy to work?

2. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
Try the test I posted. Hook +12V to +; Hook 0V to the instrument case; Hook the wiper of the 1K pot to (); Hook the ends of the pot to +12V and 0V, respectively. Slowly move the pot up from the ground end while recording the voltage between () and 0V. Make a table of gauge reading vs the voltage input.

3. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
Hi Mike .. Sorry taken a while to get back! I did what you asked.. the meter movement seems to mover better for some reason. I didn't have a1k pot on hand but sued a 10K if that's ok, I can run and pick one up if needed!

here is what I come up with:

EMPTY = 12.0v
1/8 = 11.51v
- = 11.51v
1/4 = 11.03v
- = 10.53v
1/2 = 10.0v
- = 9.64v
3/4 = 9.15v
- = 8.73v
FULL = 8.23V

Kendall

Joined:
Jan 12, 1997
Messages:
-
Likes:
0

5. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334

Hi Kendall,

Ok, just to confirm, the Power supply was set to 12V. right?

The connections to the gauge were like this:
Power supply +12V to Gauge Plus terminal
Power supply 0V to Gauge Can
Middle terminal of 10K Pot to C
One end of 10K Pot to +12V
Other end of 10K Pot to 0V?

Confirm these connections

Make the following new measurements with your Ohmmeter:

1. Resistance from Gauge Plus terminal to Gauge Can =
2. Resistance from Gauge Plus terminal to Gauge C terminal =
3. Resistance from Gauge C terminal to Gauge Can =

After you tell the resistance measurements, I might have you make another...

6. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
HI Mike, Yes used 12.02v and connected that way listed above.

+ to can = Open
+ to C = 825KΩ
C to can = Open

no problem if I need to make other test!!

Kendall

7. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
The Final Test

Kick me for being stupid. I went back and re-read the entire thread and I see that I confused the entire issue by thinking that the gauge had an internal connection to the Can, which is how car and aircraft usually work. Your Ohmmeter test indicates that there is not a connection to the can in your gauge!

However, there is obviously something else inside the can beside just a simple D'Arsovol meter movement. This is indicated by the very high resistance (825K) from + to C.

Without opening the can to see what is inside, here is my last ditch suggestion: Set the DC supply to 14.25V, which will be close to the Jeep's system voltage with the engine running above idle. Connect +14.25V to Gauge +. Using your 10K Pot as rheostat , connect wiper to C and one end of the pot to 0V.

Turn Pot until needle just moves off the empty stop (so the needle is noticeably above empty). Connect your DMM in DC voltage mode and read the voltage from C to 0V (i.e. across the pot). Record the voltage. Without disturbing the Pot wiper position, momentary disconnect the C connection from the pot, and read the pot Resistance.

Repeat this for 1/4, 1/2, 3/4, and just barely short of Full Scale.

This will give a table like this:

Code (text):

NeedlePos   Pot Volts    Pot Ohms
E
1/4
1/2
3/4
F

If I have these readings, I will integrate the gauge into the opamp circuit I worked out before.

Last edited: Feb 13, 2010
8. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
Hi Mike.. not sure if you got my last message but using that 10k pot I can only adjust from full to about 1/4 mark on the gauge! any ideas?

9. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
Just to confirm. With only two connections to the pot, and no connection to the gauge can (case), the Full reading is achieved when the pot is set to Low resistance. As you increase the resistance, the Gauge reads progressively less full as you increase the resistance, but when you reach 10K (the max), the gauge still reads 1/4?

If that is the case, just keep adding more fixed resistors in series with the pot. Find the total resistance required to make the gauge read a needle width above Empty.

10. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
Mike, yes that's correct! no connections to the case! two connections at the pot. I cant seem to get any good configuration with adding more resistors! I do have one 240k resistor in place of the pot and get a reading of 14.20v from a 14.25v power supply. the gauge reads just a hair above Empty.
if that helps..

Kendall

11. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
Ok, so fill in the meter reading vs external resistance table.

12. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
Hi Mike,

This may be a dumb question.. are you wanting me to determine with a fixed resistor(s) per increment on the gauge? and record a voltage for those locations also?

Kendall

13. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
Reread my post titled "THE FINAL TEST"

14. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
Hi Mike..

I think I have got what you need!

From the top:
I have a +14.25V from source to the gauge (+) and from the (C) on the gauge to resistor(s) to (-) 0v at supply source. I presented various resistors in place until I matched up E, 1/4, 1/2, 3/4 and Full note, E values are for the meter needle just above the E!!

What I came up with:

E= 277.5kΩ and 14.18V*
1/4 = 11.05k and 13.17V*
1/2 = 5.22k and 12.27V*
3/4 = 3.36k and 11.33V*
F = 2.26k and 10.30V*

*Voltage where read across the resistors

Let me know if there is something else you need

Thanks,
Kendall

15. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
Perfect. It'll take me a couple of days to finish the design...

Looks like it is a fairly linear voltmeter, ~870Ω/V, requiring 4.5mA through it or 3.95V across it for full-scale,

Last edited: Feb 18, 2010
16. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
Deleted

Post deleted.

Last edited: Feb 21, 2010
17. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
The Final Circuit

Ok, here is my circuit. It uses the sender model developed earlier (see the previous postings where EXCEL is used to develop the Sensor Resistance vs Gallons equation).

A 78L05 voltage makes a reference voltage of ~100mV, which appears across the Sender. As the resistance of the Sender changes, the Sender current is a fairly linear representation of the float position. The Sender current also flows in the feedback path around U1 LM358 opamp, where some of that current flows through the gauge. Resistor R6 shunts the gauge to calibrate the FULL reading. I simulated the gauge reading with the input voltage at 12V and 14.5V, and the 78L05 regulates so that there is no discernible effect on the gauge reading. The input network should protect against transients found in automotive electrical systems.

Since the resistance of the Sender is not zero when the Tank is EMPTY, it is necessary to inject some current into the non-inverting input of U1 to zero the gauge when the tank is EMPTY. Resistor R1 sets the EMPTY reading. Bad news is that R1 and R6 interact, so if one is changed, the other will need to be changed, too. I suggest that you buy 1% resistors as shown in the schematic. That way, no adjustment of the FULL and EMPTY readings should be necessary.

Since the LM358 comes two per package, I used the second one as a comparator to light a Red LED when the fuel level drops below about 2 gallons. Look at the plots. The Red trace is the voltage across the gauge as a function of the number of gallons (Xaxis). The Blue trace is the current through the LOW FUEL LED. I would use an ultra-bright Red LED for good daylight visibility.

This has been a fun project. I would appreciate some feedback on how it works out.

#### Attached Files:

• ###### JeepFuelGauge.png
File size:
51.2 KB
Views:
251
Last edited: Feb 21, 2010
18. ### ericgibbsWell-Known MemberMost Helpful Member

Joined:
Jan 4, 2007
Messages:
21,350
Likes:
649
Location:
Ex Yorks' Hants UK
hi Mike,
The 'final' image is been reported as an 'invalid png file'

19. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
Hi Mike, hows things going with any possible solutions with the fuel gauge?

Kendall

20. ### scopemanNew Member

Joined:
Nov 23, 2009
Messages:
40
Likes:
0
HI Mike, sorry I see the wonderful work here you posted! I will move forward on this circuit and keep you posted to how it works out!

Again, many thanks for you help!

Kendall☺

21. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,263
Likes:
574
Location:
AZ 86334
I'm viewing it with Firefox, and it looks fine. Is anybody else having trouble?