Fuel gauge troubleshoot


PeteK

GT Owner
Mark II Lifetime
Apr 18, 2014
2,470
Kalama, Free part of WA State
What’s an “oscilloscope”?

(Pause)

Just joking—I happen to have an olde Heathkit o-scope! Can’t remember the last time I used it though.
 

nota4re

GT Owner
Mark IV Lifetime
Le Mans 2010 Supporter
Feb 15, 2006
4,281
olde Heathkit o-scope!

Thanks a lot, Pete. I actually remember those. Now I'm feeling old!
 

HighHP

GT Owner
Jun 3, 2019
469
Spokane, WA
olde Heathkit o-scope!

That baby should work great. Hook 'er up and let us know what you see.

A lot of those for sale on eBay, the word "vintage" is used in the description on a few. Vacuum tube? Vacuum tubes were hard to find 20-40 years ago. I understand they now are plentiful due to the resurgence in demand for the true sound of music that only vacuum tubes can amplify. Huummm....maybe similar to nota4re's comment on another post regarding speaker cables.
 

PeteK

GT Owner
Mark II Lifetime
Apr 18, 2014
2,470
Kalama, Free part of WA State
Hey, everyone knows speaker cables sound different!

Re the Heathkit scope, it’s buried in a box somewhere. I haven't seen it since I moved from VA to WA late last year.
 
  • Like
Reactions: HighHP

2006GT

GT Owner
May 13, 2006
13
Thank you very much for the very helpful additional responses. I'm starting to see the way forward in solving the gauge issue, but it is different than I had first imagined. For those who may be interested, here are some recent details of my saga. I have characterized the gauge signal from the gauge control module. It is a pulse-width modulated 5 volt, 1 kHz square wave. Decreasing fuel level correlates to greater off time in the duty cycle. The gauge has a zener diode that limits the voltage to the air core motor. The coils in the air core motor are driven directly by a microprocessor on the gauge circuit board. The driver for one of the coils has failed -- a repair would entail a replacement of the microprocessor. This may be a design deficiency if the failure mode involves overheating of the chip because the coil current is near the maximum capacity of the microprocessor. The microprocessor is readily available from other sources off the shelf, but the code is most certainly proprietary to AutoMeter. I've attempted many times with no success to contact AutoMeter to see if I could obtain only the programmed microprocessor from them. It's a very long shot -- I would be stunned if I were able to buy it from AutoMeter. I could replace the microprocessor, reverse-engineer the code, and perhaps add external current drivers to the microprocessor outputs. Technically feasible, but very far from cost effective. At this point, for me it is enough that I know that the solution exists to repair the original gauge.

I am going to explore several possibilities going forward. One is the feasibility of transplanting a stepper motor into my original gauge housing -- there may be some packaging issues with this idea. Another is the purchase of the AutoMeter replacement if they remain available. A third is adapting the original gauge faceplate to the Speedhut replacement stepper motor gauge and reducing the Speedhut 270 deg. needle sweep to the approx. 120 deg. sweep of the original gauge. A fourth is a full replacement of all the gauges with the Speedhut versions. Maybe I haven't yet discovered the approach I will finally take.

In pursuing this project, I have become more impressed with Speedhut and have become aware of other vendors and suppliers, like Redline Gauge Works in CA, that also could be great resources in solving problems with gauges in the FGT. Any ideas or thoughts?
 
  • Like
Reactions: HighHP

twobjshelbys

GT Owner
Jul 26, 2010
6,189
Las Vegas, NV
Any ideas or thoughts?
I'd just buy the autometer and move on to enjoy driving the car.

Long term pursuing a solution that allows the larger gauges - speedo and tach - to either be repairable or immune would be a welcome contribution!
 
Last edited:

HighHP

GT Owner
Jun 3, 2019
469
Spokane, WA
Great analysis and report. Very interesting and informative. Thanks much.
The generally accepted assumption is that this failure is due to some type of rapid voltage fluctuation applied to the gauge. Given your review, do you think this could have caused the failure? Sounds like you think that the driver for the coils is simply undersized and high operating temperature eventually fails the driver.
If this is the case there should be a noticably higher failure rate in the very hotter climates????
Please keep us posted on your progress. Thanks.
 

2006GT

GT Owner
May 13, 2006
13
Four months later, I have an update. Based on some bench top troubleshooting, it seems like stiction may be a main issue. The gauge would indicate the proper level if it was helped with a flick of a finger. Over time, the damping fluid in the gauge movement thickens and the microcontroller is unable to provide the current to fully move the air core motor to achieve an accurate reading. In our testing, the microcontroller began to draw a large current and failed. We're working on some promising external drivers to operate the gauge. Does anyone have a non-working fuel level gauge that I can obtain? Thank you very much in advance.
 
  • Like
Reactions: FlagstaffGT

PeteK

GT Owner
Mark II Lifetime
Apr 18, 2014
2,470
Kalama, Free part of WA State
Interesting, and thanks for the follow up. My temp gauge was starting to get sticky around 150-180 degrees before I pulled the whole panel and replaced with Speedhut. Could be for the same reason as the fuel gauge.
 

HighHP

GT Owner
Jun 3, 2019
469
Spokane, WA
Thanks for the progress update. Much appreciated.
 

NorthwoodGT

GT Owner
Jun 12, 2009
1,217
Michigan
Four months later, I have an update. Based on some bench top troubleshooting, it seems like stiction may be a main issue. The gauge would indicate the proper level if it was helped with a flick of a finger. Over time, the damping fluid in the gauge movement thickens and the microcontroller is unable to provide the current to fully move the air core motor to achieve an accurate reading. In our testing, the microcontroller began to draw a large current and failed. We're working on some promising external drivers to operate the gauge. Does anyone have a non-working fuel level gauge that I can obtain? Thank you very much in advance.
I have some other dead gauges I'd be more than willing to ship to you for free but nothing in a fuel gauge right now (at least this week) lol!
 

Howard

GT Owner
Mark IV Lifetime
Apr 26, 2007
1,149
Florida/North Jersey
Aren't all the small gauges identical on the inside, getting their signal from the controller and only differing by the gauge face?
 

fjpikul

GT Owner
Mark IV Lifetime
Le Mans 2010 Supporter
Jan 4, 2006
11,680
Belleville, IL
2006GT, I suggest you send a PM to Analogdesigner and get his data on the gauge problem.
 

2006GT

GT Owner
May 13, 2006
13
Just another request for a non-working fuel level gauge for a 2006 FGT. Anyone have one that I can obtain?

Thank you very much.
 

SkeletonGT

Well-known member
Mar 21, 2019
48
Novi, Michigan
I have spent some time driving the factory gauges and I can tell you that the range of operation is 0 to 47% PWM for full scale. I drove each of the gauges using the same 5V PWM signal sweeping up and down and they all work the same regardless of their purpose. It was very interesting to watch the gauge movement as each gauge was swept. On the cluster that I was testing, one needle actually stuck and jumped as it swept others were perfectly smooth, and I even had one where the zero position would shift around like the needle was loose.
 
  • Like
Reactions: analogdesigner

JOEA2

GT Owner
Mark II Lifetime
Aug 16, 2007
355
STATEN ISLAND,NY/SEA GIRT,NJ
I gave up and replaced all my gauges with Speedhut gauges. I couldn't take it anymore!

Joe
 

PeteK

GT Owner
Mark II Lifetime
Apr 18, 2014
2,470
Kalama, Free part of WA State
SkeletonGT, thanks for doing those measurements.
 

SkeletonGT

Well-known member
Mar 21, 2019
48
Novi, Michigan
Correction. Full scale IS 100% duty cycle using 1khz PWM. Earlier I was using 500 Hz which was why full scale was only 47%. Using 1khz is the way to go. If anyone is trying to use their GT cluster with a transplant, like a coyote, let me know. I can help you make those work together.