Is it possible the float arm is hitting a baffle? Is it possible the float arm is too long and is only half way through it's travel when the tank is full?

The sending unit with two corks is the '30 to early '31 one. It has a larger top diameter and therefore hole in the tank then the later one. It like most of the sending units of that time had 0-30 ohm range. If the modern replacement does not have the same range it will not read correctly

I wasn't aware the tank had any baffles in it?  Also, as I indicated, I can freely move the arm up and down through it's arc and the gauge registers exactly as  I would expect.  It just doesn't work properly when I fill the tank.  Also when you mention 2 corks, mine are side by side together as if it was one cork.   I have been tinkering with a universal unit without much success. I have the original unit but when I checked with the information at the filling station, my measurements don't seem to agree with either the early or the late unit. It's probably because I am not measuring correctly.

If the guage seems to operate out of tank when you fit it back put a extra wire from 1 of the mounting screws to a good clean ground as the tank is insulated though surface rust or dirt.

Tony

Posted by: @1931-roadster

↑

I wasn't aware the tank had any baffles in it?  Also, as I indicated, I can freely move the arm up and down through it's arc and the gauge registers exactly as  I would expect.  It just doesn't work properly when I fill the tank.  Also when you mention 2 corks, mine are side by side together as if it was one cork.   I have been tinkering with a universal unit without much success. I have the original unit but when I checked with the information at the filling station, my measurements don't seem to agree with either the early or the late unit. It's probably because I am not measuring correctly.

I don't know if the 31 tank had baffles but I assumed they do. If not then that can't be the issue. Nor do I know the float arm swing of your sender vs your tank. I'm pointing out issues I've run across with other tanks. I had to shorten and slightly bend the float arm on the universal sender I bought for my 38 to give it enough arc to read full. Before I did that the float would touch the top of my tank before the sender got to the "full" end of it's travel. Since it reads when in the car I don't think it's a ground issue. If it was a bad ground it wouldn't read at all. I would guess there's something keeping it from making it's full arc since it reads properly on the bench but not in the tank.

When I bench test the unit, I have it installed in the tank with leads going to the connections on the car.  Through one of the screw holes, I insert a hook and latch onto the arm and it freely moves up and down without any resistance.  And, the gauge on the car records correctly what I am expecting which would be empty, half and full.  That is the part that confuses the heck out of me, because when installed on the car and I add gas, then the reading is there but completely wrong.

Hello Paul Rozon,

Your statement- "When I bench test the unit, I have it installed in the tank with leads going to the connections on the car." Could you verify to us that your fuel tank is on the bench or your fuel tank is in the car? The reason for asking is that there might be a difference in orientation of level between bench and car that could change the actual versus anticipated fuel level. OR, could your floats not be as buoyant as you think they are and not be riding on the top of the gasoline? 

When testing the tank, it is off the car on the garage floor with leads going to the connections on the car.  Now, the float on the sending unit is black plastic of some sort and it does not sound or appear to have gas inside it.  I assume that it is floating but I don't know for sure.  What I do know is the arm can move freely through out its arc when installed in the tank.  I test the arm movement by putting a hook through one of the screw holes and move it up and down without any effort.  While doing this my wife watches the gauge in the car and it matches where the float is in the tank.

Is the float touching the bottom of the tank when showing empty on the gauge or is it above the bottom? My thinking is it's possible the float is simply not at the top of it's travel when the tank is full of gas. If you can bend the float arm to where the float is lower in the tank, it'll raise the rheostat position higher when the tank is full. If the float is already touching the bottom of the tank and you bend it down you won't get an accurate empty reading. Posting pictures of your tank and sending unit and how you're testing can help us better understand what's going on. As it is we're just guessing.

Information from Filling Station website

Item ID: 364485
Make: Chevrolet

1931 OWNERS NOTE:
In 1931, there was an early and a late version of the sending unit.
This sender will only work with the late version which has a single float system with a 1 1/4 DIAMETER bolt pattern - Bolt Centers: 1.425", 1.215", 1.250", 1.250", 1.250". Early senders are a two float system and a larger diameter. This sender cannot be used for earlier models.

Item ID: 362855
Make: Chevrolet

This Fuel Tank Sending Unit is for 1930-Early 1931 Chevrolet cars & trucks. It replaces early senders with dual cork floats. Comes with gasket & screws. Dimenions: ID 1.865", OD 2.94". 1 1/2 BOLT PATTERN Replaces GM# 862080, 362855

When I read the information for 364485 it states a bolt pattern of 1.25 but all the bolt centers are different. So, how do you get a bolt pattern of 1.25?

When I read the information for 362855 it states a bolt pattern of 1.5 but all the bolt centers are different, according to Steve. So, how do you get a bolt pattern of 1.5?

I tried to get information regarding the bolt patterns and bolt centers and Steve was unable to explain and generally speaking was very unhelpful.  Apparently there is different information in the catalogue vs the website, which appeared to be his major concern.  

I have two corks on the original sending unit, but the hole centers are different.  I have 1.25, 1.25, 1.18, 1.32, 1.18

Very confused when looking at bolt patterns and sizes.

I get accurate readings when I am testing.  When testing and float is at bottom I show empty, when I pull up on the float to the top I show full and in between the two I show half. I then install the tank on the car and fill it up and I show 1/4 to 1/2 on the gauge after a short while (not immediate).  So things work perfectly when testing off the car.  The travel of the float appears to be smooth and very easy to move up and down.  Put tank on car and fill with gas and there is the problem.  I am currently working with a universal unit that has not working for me (I have made many tests) and now I want to get a unit specific for 1931, but there appears to be 2 different units in 1931.  I don't want a universal that I have to adjust to make fit, but one which I simply bolt on without adjustment that should work right out of the box.

Hello Paul Rozon,

To figure out the correct sending unit you need, you could try measuring your tank opening/gasket/screw fastener spacing. Make a tracing to send to Filling Station for advice. You could try to locate part numbers on the sending unit you have and also tank part number if possible to compare with information found online at: chevy.oldcarmanualproject.com. Look for: 1929-1932 Chevrolet Master Parts Price List for Six Cylinder Models. Look for: Sheet Metal Page 89. Look for: Gas Tank Page 95&96. I=29, U=30, IND=31, C=32. If interested: Electrical Group: Gasoline Gauge-Dash Unit Page 109 (view whole page to see illustration). Your cars serial number might reveal early/late model year maybe.