Fuel Gauge Modification

The M series fuel gauge functions by utilizing two bimetallic strips heated by Ni-chrome wire, one to push the needle up and one to push the needle down. The two strips are controlled by a center tapped resistor in the tank mounted fuel level sensor. The center tap position of this resistor is set by the float of the fuel level sensor. The two bimetallic strips Ni-chrome heating elements are temperature limited by a set of contacts. These contacts, over time pit and eventually stop functioning.

To eliminate this pitting, a modification can be accomplished to reduce the contact current from 600 ma to 600 ma. The solution is performed by adding a MOSFET transistors to carry the current. The procedure is as follows:

1. Break the heater connections
    a. Modify the contact stack
    b. Gold plate the contacts
2. Build and install circuit board
    a. Cut board and cut the traces
    b. Solder on MOSFETs and resistor
    c. Epoxy board to meter
3. Wire the circuit board
4. Test and adjust the meter

Reference pictures

Break heater connections

To make the break in the circuit between the contacts and the heating elements the following is done:

 Drill out the rivet holding the contact in place

Contact & heater assembly removed from gauge

Contact & heater assembly disassembled

Build and install circuit board

Wire circuit board

I decided to use Litz wire for connections "A" and "B" because it is a very flexible wire

Test the meter

Testing consists of hooking up the meter and sending unit on the bench with a 6 volt battery to power the meter/sender. The two meter arms are iteratively adjusted to get the meter needle to land on the full mark when the sending unit arm is in the full position and the empty mark with the arm in the empty position.

The big problem is to determine the sending unit's full and empty positions. My current plan is to attach a weight to the float so it drops to the bottom of the tank and measure the sender resistance. To measure full I can just fill the tank and measure the sender resistance. I then can adjust the sender's arm for each position to get the resistance for testing the meter.

When testing it is a good idea to connect a ammeter meter in series with the power to the meter. This is a good way to make sure everything is connected correctly and lets you a way to monitor the contact open/close activity. When everything is connected and power is applied the gauge will take some times to settle. The meter should be drawing 200-250 ma. Initially the gauge is settled when the contacts open and the ammeter meter shows zero current for a moment. The contacts should open/close at a rate of once every 5 seconds.

 Test setup

Note ammeter shows ~300 ma.

Clock Mod/Improvement

I recently installed a car dash into my M15 with all the bells and whistles including a clock. I had rebuilt this clock myself by completely disassembling it, ultrasonically cleaning the parts, and reassembling it. I used the approved clock oil and had it on the bench running flawlessly for several months. I felt it was good to go as it kept time within 5 minutes a month. After installing within a month, kaput. I suspect this is due to the mainspring winding mechanism with those easily burned contacts.

So, what does one do? Taking the clock out for repair is such a hassle I figure I need to come up with a permanent solution to the contact issue. All the engineering solutions I came up with required significant drain current that one would worry about the battery running down. So I decided to use the contacts to my advantage. Instead of the contacts carrying the rewind solenoid coil current, I used the contacts to trigger a MOSFET that carries the the solenoid current. Generally, MOSFETs have a very low off current and one that I had in my stash is very low, low enough where I could not measure any current.

Pros:
1 Contacts last a long time.

Cons:
1. Not stock.
2. Increased complexity.

Here is the before and after circuit:

Modification is as follows: