Slower then molasses fuel gauge!

[NorthernLite]

So i did Yel-low's faulty fuel sending unit fix. Cleaned everything up real good and shiny. It solved my problem of inaccurate fuel level, but it still takes about 10 minutes of driving before the needle reaches full after i pull away from the pumps. Any clue why it would move so slowly?

[Goliat6288]

mine has always done that

[hawkman]

Originally Posted by Goliat6288

mine has always done that

Yep. I cleaned all the varnish off of my sending unit's contacts and verified the resistances. But it still takes a few minutes to indicate a full tank. I wouldn't say 10 minutes though. It did start maintaining the tank level indication even when the car is off, which is what it is supposed to do but didn't do until I cleaned it.

[yel-low]

The slow movement is attributed to the board inside the cluster itself. Mazda calls it a "print plate". There is a comparator chip and a cap. The cap serves as a transient buffer. It is common for the cap to leak and/or the chip to go bad. I've dissected that thing more than enough times trying get mine back into calibration.

[NorthernLite]

So in other words...just leave it I'm fine with it taking a while, I was just curious if there was a quick DIY fix or not. Thanks for verifying guys, I thought I was the only one with a messed up excuse for a car that I happen to love

[bigb]

This is actually by design so your fuel gauge doesn't bounce up and down when you hit a bump, corner, or drive up and down steep hills. this is also why you have two sending units, one on the left front of the tank, and one on the rear right of the tank. this way if you tip the tank in either axis, the values should roughly cancel each other out and the needle remains sorta constant.

eg. go up hill the tank will fill more to the rear of the tank. so rear sender reads fuller, then front sender reads lower. since the resistance is in serial, it sorta works out, though the senders are of different scale. i think front one is 0-65ohm and rear one is like 0-35 or 40 ohm. it's in the FSM I don't remember off hand. but maybe the stroke is shorter on the rear one. anyways... the print plate comparator averages out the results over a few minutes on top of that to keep the needle steady.

[NorthernLite]

very interesting. so upon learning that, should i be cleaning the other level sender? or just the one Yel-low showed how to clean is the only one cleanable?

[bigb]

you need to drop the tank to clean the other sender, so not recommended. cleaning it will only make the measured value more accurate, however it will not change the speed at which the needle moves.

That being said, if you have to drop the tank for whatever other reason, you may want to consider cleaning both senders as part of the tank service

[yel-low]

Oddly enough I have a 97 Millenia where you DON'T have to drop the tank to get to the second sender. I guess you will just have to pop the rear seat and find out if your second sender can be accessed. My 95 did not have this.
I can take pictures of the two hatch covers if needed.

[NorthernLite]

I did notice a second cover on the drive side of the seat, could that be it then? and removal and cleaning would be the exact same thing right?