36 Buick Fuel Gauge

[Guest KeithElwell]

I realize this topic has been discussed a great deal before and I believe I have reviewed all the old threads. But I have not found definitive information on the ohms resistance values upon which the stock 1936 Buick (Special) fuel gauge should operate. What should be the total ohms range and what are the values for empty and full?

I installed a new universal 0 - 30 ohms sending unit and thought I cut/fit it correctly but apparently not. It registers 3/8 full when actually empty and about 3/8 past full when I manually push the float to full. I am not sure I chose the correct ohms range sending unit but I feel like I have a good ground of the tank and lines.

[FLYER15015]

Kieth,

First make sure you have a very good ground. I even put an aux ground on my '40 from the sender plate to the frame.

Then try a 5-10-15 ohm resister from Radio Shak, in the hot line, till she seems about right.

Just my $ .02...... and worth every penny.......

Mike in Colorado

[Guest KeithElwell]

Mike,

What you advise with the resistors does make sense and I will give it a try. Again, I really think my auxiliary ground is not my problem. I think it could not be a better ground but I will do some more checks to be certain. But the reason I'm thinking the resistors may be the solution is that today I removed the sending unit from the tank and tested it with my ohmeter by itself. Sitting in the float down empty position it shows about 10 ohms resistance but pivoting the arm up to full I measure about 40 ohms - a delta of 30 ohms which is what the mfg'er advertised except it should be 0 ohms on empty. Maybe I start my resistor solution trials with a 10 ohm resistor. Wish me luck.

Thanks Mike

[37_Roadmaster_C]

Keith,

The sender for your '36 is a 0 to 30 ohm unit, empty is 0 ohms and full is 30 ohms. With your sender reading 10 ohms at empty level and 40 ohms at full the guage will read about 1/3 when empty and over full when full, about what you say yours reads. Now for the bad news ... no resistor will fix this problem as you need to get the resistance down 10 ohms across the board. You might be able to adjust the sender rehostat or bend the float arm or..... but what you need is for the resistance to be 0 or VERY close at empty and 30 at full. If you can not get this range I would call the supplier and ask them for an exchange as the one you have is defective. I hope this helps.

Robin

[Guest KeithElwell]

Robin,

I see what you're saying. Adding a 10 ohm resistor to my hot wire will only increase the resistance (duh) when I need to decrease resistance. Instead of making the gauge read closer to empty it would read even more than 3/8 full when actually empty if I added a 10 ohm resistor.

I will contact the mfg'er to see if I boogered something up when I tailored the sending unit to my tank but I feel like I did what the instructions said very carefully.

How did you determine the ohms range and values the OEM gauge is made for? I was thinking about getting a 30 ohm resistor and testing the gauge only until you provided this info.

[Mark Shaw]

Keith,

I am no electrical genius, but maybe the wire needs to be replaced to reduce the resistance?

[jeff]

Check the dash unit, and the wire back to the tank, by disconnecting at the tank and checking it open-circuit and then tied to a good ground. The line from the dash should really not be considered a "hot" feed, rather, it is a shunt to ground, with the tank unit adding resistance as the tank fills. Open-circuit (infinite resistance) will cause all of the current in the gage to also flow through the "full" coil, which will cause the needle to swing past the F. (as you have shown, 40 ohms is as good as infinite). So, disconnecting the wire proves the gage on the dash is ok. Connecting the wire straight to ground means that only the "empty" coil sees the current. This should make the gage read E, which is just a tick above "off".

So, do this first, and then be confident in your dash gage. If finding a good ground doesn't make the gage read empty, check your connections, to be sure you are not adding any resistance along the way.

Then, contact the supplier and see if you can get a 0 - 30 ohm replacement.

Jeff

[Guest Phil Cordery]

Resistance values can be altered thus, 2X10 ohm in series = 20 ohms, 2X10 in parallell =5 ohms, so all you need is the correct value for the parallel resistance, if you have friendly radio man near you, he may have a variable resitance unit you can connect across the tank unit & by simply turning the Knob you will arrive at the correct value which equates to empty & full. then simply read off the scale & replace unit with the correct value resistance ,

[FLYER15015]

Keith,

Robin and Mark have you on the right track, and I had it backwards.

Question, Do you still have a 6 volt system ? You know they like BIG wires.

My ground from the sender to the frame is 10 Ga, and I traced the OEM wire from the gage to the sender just to make sure that 70 plus years did not produce any wear spots, and I did find some.

Mike in Colorado

[jeff]

I believe that if you chose a resistance to get closer to 0 ohms when the tank is empty, you will read about the same when it is full. If you chose a resistance to read full when the tank is full (30 ohms) you will still read just below 1/3 tank (9 ohms) when it is empty. Could be wrong, but I don't think you can have it both ways.

(I made a graph, but can't seem to load it!)

First column is your variable resistor. Setting it at or near 0 will make your gage read about empty when the tank is full. Trying to get 30 ohms of equivalent resistance will result in the gage still seeing almost 10 ohms (1/3 tank) when it is empty. Resistance of two resistors in parallel is: (R1XR2) / (R1+R2)

10 40

R E F

0 0.0 0.0

5 3.3 4.4

10 5.0 8.0

15 6.0 10.9

20 6.7 13.3

25 7.1 15.4

30 7.5 17.1

35 7.8 18.7

40 8.0 20.0

45 8.2 21.2

50 8.3 22.2

55 8.5 23.2

60 8.6 24.0

65 8.7 24.8

70 8.8 25.5

75 8.8 26.1

80 8.9 26.7

85 8.9 27.2

90 9.0 27.7

95 9.0 28.1

100 9.1 28.6

[Guest KeithElwell]

Here is what I found when checking the gauge alone.

I'm still running a 6 volt system but my battery is pretty old and not holding a full charge. I placed it on trickle charge and it shows 6 volts.

Sending unit is completely removed from the car, along with the tank.

My wiring is "new" but been in the car quite a while but car hasn't been driven hardly at all. The harness is the cloth covered wire, presumably the same OEM gauge. It also has a bullet/socket connector in the sending unit wire near the tank. Don't know if original had that but it probably doesn't contribute to a good electrical connection.

With power switch on at column the gauge shows empty with sending unit end of wire harness hanging loose not toching ground.

When I press the wire end by hand to a ground screw I drilled into the frame crossmember the gauge goes to 1/4 full. But when I take the wire away from ground the gauge goes way past full.

Turning switch off the gauge goes back to empty.

I can't tell what this diagnostic indicates

One good thing is the sending unit mfg'er is sending me another unit at no charge. Mine doesn't seem to be correct so that's good but looks like I have issues with the car also.

[37_Roadmaster_C]

Keith,

The resistance values I listed were obtained by measuring the sender in my 37 Roadmaster which is the same as yours. Mine actually measures 0.15 when fully down (Empty) and 31.2 when fully up (Full).

Robin

[Guest KeithElwell]

Thanks Robin.

I didn't find this info in my service manual so measuring the gauge response for the values is logical.

[Guest KeithElwell]

Robin,

What does this check indicate?

I decided to make absolutely sure of my ground point I had used for my sending unit flange before. So I cleaned this spot well with a drill motor and wire brush. Then I screwed the end of my wire (that should go to the sending unit) directly to this ground point. This takes the sending unit and tank completely out of the equation and should simulate an empty fuel tank. But I turned on the switch and the gauge showed 1/4 tank again.

I then went directly to the back of the gauge and clipped a jumper between the upper (sending unit) terminal of the gauge to ground, turned on switch and once again shows 1/4 tank.

???

[37_Roadmaster_C]

Keith,

The check you describe and the strange results leave me wondering what you have going on. I would expect that when you ground the sender terminal on the gauge the needle should go to empty not 1/4. Now I will have to do a little research on the operation of the Buick gauge, but I think that any ground resistance may cause the gauge to read high. You might want to try running a temporary ground wire directly from the battery to the gauge sender terminal and see if the gauge reads empty. If it does then you have chassis ground problems, if there is no difference then you most likely have a gauge problem. I currently have my Roadmaster apart so I can do some basic checks and I will report back with the tests and results.

If any others have any thoughts on this problem and the test results Keith is seeing please chime in with your thoughts.

Rob in

[jeff]

I think your open circuit, full scale reading indicates a good gauge. Your grounding attempt, giving you 1/4 tank reading, must be a bad ground. I had some flakey operation recently due to a short on the power side, the mounting of the gage to instrument panel plate is very critical - you can get a short to ground very easily. Be careful working under the dash, and be sure your battery is disconnected. By the way, does anyone know what achieves the damping of the gage? The book says there is a mechanical inertia damper as part of the gage, but both of the gages I have tried bounce all over the place. I would love to somehow restore this function. Typical of this timeframe would be a little leather disk, that might re-swell with a careful drop of oil! (but that's just my wild guess, I can't see anything that does this damping function.) Jeff

[Larry Schramm]

If the fuel gauge sending unit has a GOOD ground, then the problem is with the instrument panel ground. Pretty simple circuit. If the rare chance the cluster ground is good, then it is with the cluster gauge or the panel wiring it's self.

[Guest KeithElwell]

Guys,

Thanks for all the input. However the mystery remains.

I ran a jumper between my battery negative post and the sender terminal of my dash gauge, turned on switch, and once again the gauge goes up from empty (switch off) to 1/4 full (switch on). That jumper makes no movement to the gauge needle at all.

By the way, I am embarrased to admit that I forgot to calibrate my ohmeter before I checked the sender unit alone. My meter was off by 9 - 10 ohms until I zeroed it out with the probes held together. I'm now convinced there is nothing wrong with my sender. It measures 0 ohms on empty and 30 on full. But that's out of the car.

I have some extra used fuel gauges but jury rigging them proved they do not operate at all. I am assuming my gauge needs repair so I am now removing it from the panel. Any suggestions where to send it?

[Guest KeithElwell]

Gauge removed from car. Hooked jumper between positive post of battery and hot side terminal of gauge, jumper between negative battery post and sender terminal of gauge, jumped up to 1/4 tank again. Placed sending unit between gauge and battery and got 1/4 tank again, no matter the sender unit arm position gauge shows 1/4. Verified sending unit alone is 0 ohms empty and 30 at full arm position.

I guess gauge is messed up.

[37_Roadmaster_C]

Keith,

I got busy today and was unable to do any checking on the gauge operation. However, after reading the responces after my last post I think you may have a cluster ground problem as mentioned. To try to test this do the following... Hook the gauge up as you described above (out of the cluster) Battery positive to gauge positive, sender terminal to battery negative. Now note the reading, you said before 1/4. Now leaving everything this way hook (touch) the gauge case (cluster ground point) to the battery negative terminal. Did the reading change? If not then bad gauge, if it went to empty then bad cluster ground. I will try to check things out tomorrow and post an update.

Robin

Edited July 15, 2013 by 37_Roadmaster_C
spelling (see edit history)

[Larry Schramm]

If the sending unit is going from 0-30 ohms, then I would say the sending unit is good. The next place to look would be the ground wire from the sending unit to the where it goes. Many times the frame. You can do a voltage drop test by putting a large light bulb like a headlight bulb on the ground wire which goes to the frame. Check the voltage on both sides of the light bulb as it is glowing with the meter lead on a GOOD GROUND. On one side of the bulb you should have battery voltage. On the other side you should have less than .25 volts, but preferably zero. If you have more than about .25 volts, you have a bad ground.

[37_Roadmaster_C]

Keith,

I did a little research today on the fuel gauge operation and found the following: The gauge is a 'balanced coil design' which means that the gauge MUST have a good ground to the ground terminal on the gauge or it will not read properly. I have not had the time to check the operation of mine with the instrument ground disconected. I will try to simulate this and see if I get the 1/4 reading you got.

Please keep us informed on your findings as gauge problems are hard to troubleshoot and repair.

Robin

[Guest KeithElwell]

Guys,

Here is what I found tonight.

I made up a 12 ga wire with crimped terminal ends and firmly attached this to the sender terminal of the gauge and to the sender. I then used the aligator clip on a jumper to hold together the case of the gauge and the flange of the sender and the other end of this jumper aligator clipped to the clamp bolt on the negative battery post. And finally another aligator jumper clipped to the hot terminal of the gauge and the other end touched to the positive battery post. The gauge jumped up to about 1/4 full with sender arm down and I was able to get the gauge to respond when I raised the sender arm. The gauge pegged out before the sender arm was fully up. At times during this check the gauge would jump up to full before I lifted the sender arm - it seemed erratic because some of my connections are poor (alligator clips, touching wire to terminal, etc.).

I'm thinking I have nothing to lose if I reassemble and reinstall everything very carefully, cleanly, and firmly - making sure I make good connections. And see what happens. Still gotta think the gauge will start at 1/4 full but we shall see.

[Guest KeithElwell]

Larry,

Once I reinstall everything that will be a good time to do your voltage drop test.

Robin,

Should I install a ground wire between my instrument cluster and the dash structure? Was this done at the factory?

[37_Roadmaster_C]

Keith,

There was no ground wire on my dash. I do not know if there was one from the factory or not, it would not hurt anything to add a wire. Based on your last test I am willing to go out on a limb and say you most likely have a bad gauge. I will try to get some more info and readings from mine as I need to confirm proper operation. Another thing you might want to do before you reassemble the instrument cluster is make the following resistance measurements: 1) measure the resistance from the sender terminal and case of the gauge, 2) measure the resistance from the battery terminal and the case of the gauge, 3) measure the resistance from the battery terminal and the sender terminal of the gauge. The last reading should be the sum of the first two measurements, if not then there is a problem inside the gauge. More to come .......

I just found this article about fuel gauge operation. It is about a 62 Corvette, but the only difference in the gauge system is the sender resistance (old Buick 0-30 ohms 60's GM 0-90 ohms) all else is the same. This article is very good about explaining how things work.

http://www.vetteclub.org/warehouse/tech/To%20be%20Filed/C1%20tech%20papers%20from%20the%20Restorer/58-62%20Fuel%20Gauge%20Explained.pdf

Robin

Edited July 17, 2013 by 37_Roadmaster_C
add info (see edit history)

[danhar1960]

These may be of some use or help.

Danny

[Guest KeithElwell]

Thanks much Danny. I look forward to downloading that info tonight and reading thru it.

[danhar1960]

Keith,

If they don't download, let me know and I'll email them to you.

Danny

[Guest KeithElwell]

Danny,

I'm not sure how to download them. May have to join the website, etc. Please email it to me at elwellk_d@yahoo.com

Thanks

[danhar1960]

No problemo. Will do tonight.

Danny

[Guest KeithElwell]

Just a short post here to give my status.

Received a new sending unit and tailored it to fit my tank. And it tests correctly as does the old one, now that I was more careful with my test. The new sender looks like a better design, which includes a dedicated ground stud on the flange.

Now doing more checks of gauge and wiring using the troubleshoot manual Danny kindly provided and the advice from Robin and Jeff.

If gauge and wiring check ok I will carefully reinstall everything and hope for the best.

[Larry Schramm]

I know that 0-90 ohms resistance for fuel tank sending units was standard in the early 70's when I was working on the line at AC Spark Plug making them.

[Guest KeithElwell]

Been putting in a lot of hour at work this week so haven't worked my gauge problem. But I did reinstall one of my two gauges in the car and hooked up to my new sending unit. Made really sure I had good electrical contact. But once again the gauge came to 1/4 - 3/8 full when power on, and way past full when lifting sending unit float to full.

Next I will do the same check with my second gauge. If it reacts exactly the same I can't believe it is a gauge problem. But if the second gauge reacts differently I will send both to a gauge repairman.

[danhar1960]

Keith (and others),

I have emailed a full PDF copy of the complete manual to those that requested it.

It should come in 2 emails and are zipped into a self extracting file.

Hopefully it works. If you didn't get them, let me know.

I will get around to doing the "AC How to service Speedometers" manual as well at some point soon.

It too may be helpful.

Danny

[Guest KeithElwell]

Sure appreciate it Danny. I've used the manual already.