Repairing fuel level sender in minivan

Ted, was this the van that you did the transmission rebuild on? If so, how did it go putting it back together?

-Kirk Matheson

Hi...

Thanks very much for sharing this. Mine shows empty if there's less than 3/4 left, and the dealer wants to change the whole pump/sender unit... expensive.

I have the luxury of getting at it through the trunk, so guess I'll have a go at it myself :)

Thanks again, and take care.

Ken

What normally causes this problem is the center of the rheostat wears out and bending the arm will not help if this occurs. I don't understand, if you are going to go thru all that trouble to remove the in tank pump to repair the sending unit why not replace the sending unit, I thinks its about 80 bucks.

Thats alot of work to put back together to find out it didn't fix it.

Glenn Beasley Chrysler Tech

Ken, depending on the mileage of your vehicle the dealer may have suggested replacing the intire pump because the next thing to go will be the pump. If I get a problem like that in the shop with a vehicle with 100k or more on it. I will give the cust 2 options, the sending unit and the pump, same labor. If they plan on keeping the vehicle I would advise the pump and most owners that plan on keeping the vehilce for awhile will take the complete pump as the repair

Glenn Beasley Chrysler Tech

Yup. On this one, that didn't happen. The center of the rheostat was still nice and shiny, plenty of material there. I've seen the fuel level sender on a non-AWD vehicle and the rheostat on it is a lot smaller, and in my uneducated view, looks a lot flimsier.

Well because exactly what you said about the fuel pump. I priced the parts before even tearing into the job, and the sender itself is not available from the aftermarket, and the cheapest aftermarket fuel pump/sender I could find was in the $180 range. I couldn't even imagine what a new fuel pump/sender from the dealer would cost, I'm sure it's a lot more than $180 if the sender itself is $80.

My thought when going into it was that if I was going to start buying any parts I would buy the complete fuel pump/sender assembly. It seems dumb to buy just the sender then have the fuel pump kick the bucket 6 months later and then if you buy an aftermarket replacement you get the sender again, so your $80 you spent on the first sender is now wasted. But before buying anything I wanted to get a look at it - with this van considering how the prior owner treated it, the problem could easily have been elsewhere, and not in the sender at all.

If I had found the center of the rheostat worn out then I would have made the decision whether to just replace the entire pump/sender assembly, or to do nothing, put it back together then just drive it until the fuel pump dies, without a level guage. Based on the difficulty getting the tank out I probably would have swore a blue streak then just bought the pump/sender assembly.

It was a calculated risk. After the bending procedure I tested it and even putting a fair about of lateral force on the float arm and moving it, the wiper did not pull away from the rheostat. And it tested out good on the ohmmeter while moving the arm (I used an analog ohmmeter of course) no dropouts.

Of course, the wiper could simply have lost spring tension and it will just wander away from the rheostat in a few months and I'll be back with no guage. But at least now I know proof positive that the dash wiring to the sender does in fact work.

Ted

Yes. I did the R&R, Transmission Exchange

did therebuild.

I added to the pictorial some of the showstoppers. The biggest one is that the van's frame is bent. I knew that it had been involved in a collision before I bought it but I did not know how bad it had been bent.

The worst bending is in the front where the front engine/transmission mount bolts to, right under the radiator. The frame there is bent up, which is a common problem with front end collisions on these vans. Obviously the front trans/engine mount is thus higher than it's supposed to be. The result of this is that the engine is rocked backwards somewhat more than it's supposed to be.

With a regular non-AWD this would not matter. However with an AWD it does because it causes the power transfer unit and rear driveline output shaft to basically dip down. In my case it is at least an inch lower than it's supposed to be. Thus when the engine revs under load, such as if you power brake it, the engine shifts in the mounts (like it's supposed to do) and the rear driveline bangs down on the crossmember, with much loud scraping and banging. I'm sure it's not good for the driveline output shaft. Worse, because the power takeoff from the transmission is not parallel to the rear carrier, it causes massive driveline vibration. When I first test drove it, when I got out on the highway it sounded like I was driving on a steel grating bridge.

Obviously the correct way to fix this is to pay someone with a laser frame alignment machine $800 to straighten it out. And I may eventually do this. But in the meantime I have shimmed the engine mounts. I shimmed the rear trans mount with 4 washers, (the LH side mount) and the front engine/trans mount with as many washers as it could take and still have threads left to hold the nuts on. And I also passed a bolt through the holes in the front engine/trans mount rubber to help give that mount some extra stiffness.

This helps tremendously. The PTU power output shaft is still a few degrees off parallel with the rear carrier. But not as badly as before and so it does not vibrate on the highway anymore. There's a few oddball resonances in it, for example at 30Mph there's a resonant point in the driveline and it will moan, but that stops when I hit

33 or more Mph. And more importantly I can start off from a stoplight with normal throttle and not have the driveline bang against the crossmember.

Ted

I'll bite. Why was an analog ohmmeter necessary for testing the rheostat?

Is there something unusual about the rheostat?

Granted, it's easier with an analog to see a smooth, clean reading sweep so probably almost having to graph the data.

If the rheostat is a small area, I see your point about the analog meter.

Generally I reserve the analog for testing capacitors. My analog uses so much current for its testing that it can actually polarize delicate electrodes if I'm not careful. But it's a cheap analog meter though.

Has nothing to do with that. A DVM, even a graphing one, is not anywhere near sensitive enough for a changing resistance like testing a potentiometer (rheostat)

The problem with potentiometers (pots, in the trade) is as Glenn said areas of the resistance material can wear out. This can be seen by when you move the wiper arm of the pot (turn the handle, slide the level, whatever) and you see the needle of the analog VOM move, if the wiper hits any areas where the resistance material is missing, even minute, tiny areas, there will be a slight almost unnoticable stumble or drop of the needle as it is sweeping up the scale of the VOM. You may not see this unless you wipe the arm back and forth several times. Of course, it's arguable if for a fuel level guage you need that kind of sensitivity, but for an audio pot you certainly do.

A non-graphing DVM will of course be nothing more than a jumble of numbers, and even a graphing one probably would not show that kind of drop.. And a cheap, small pocket analog VOM without a sensitive needle would also probably not show it either. An oscilloscope could show it, of course.

I have several DVMs also, my analog meter is an older Radio Shack model back when RS actually sold decent analog VOMs, with a large meter in it. Besides testing pots, it is also useful for testing diodes, as DVM's generally don't supply enough power to a diode to get it to work.

Unfortunately, the big problem with analog meters is that they are only as good as the meter used, and it is expensive to make a large, sensitive meter. Back in the olden days, even though a good multimeter was expensive, an oscilloscope was far more expensive, so people were willing to pay for an expensive multimeter. Today, scopes are cheap and the few testing tasks that you can't use a DVM for and need an analog meter for, a scope can do. So, people buy scopes and DVMs and can do everything that an analog meter can do, so there's no market for the good ones anymore.

Ted

Your good explanation reminded me that I have what used to be a very expensive HP (got it surplus from a FAA radar station that was still using wire-wrapped boards in its electronics - I'm thinking, wire wrapped and this is what tracks airplanes, uh oh...) analog meter that did millivolts. It weighs what, 10 or 20 pounds and runs on AC. I guess I could send some current down the rheostat/pot and use that? I guess I got this because I needed to do microvolt, =B5V, readings at that time and was tired of always fussing to get the voltage up enough for an oscilloscope to read. I really needed a 6? digit DVM but they are still the same price, about $1000 for a good Fluke.

With the 'scope, would you not need to send current down the pot to see a trace? Or do scopes now take resistance readings? I have the very last of the old but fast Tektronix analog scopes - but have not used it in a long time so can't remember if the scope can generate current - it has a test current, probably a square wave at 1000 Hz and 1 volt p-p if I'm too lazy to breadboard the thing properly.

Well this is theoretical since we all know that most mechanics and dealerships are parts-changers and probably none of them would test a pot. But what I would do if I was going to use a scope is to run some temp jumpers from the sender to the connector on the underside of the vehicle that plugs into the fuel pump, turn the vehicle on and use the vehicle 12v power.

Usually in audio circuits what happens is that the cheap circuits run DC voltage through the pots along with the AC signal. That lets them reduce the amount of circuitry in use. Over time the DC power burns out points in the pot. You can hear it if you have ever turned the volume control on a cheap radio that has a few years on it and you hear a lot of scratching and static when the volume knob is turned. So with a scope you would just jack into the circuit itself right at the pot.

Ted

Except you would quite possibly burn out the pot, as it is designed to work on 5 volts, more or less.

Do the cars drop to 5 volts in a typical 12 volt car circuit with regulators or dividers when there are electronics? I guess they'd have to.

I have seen 5 volts for chips, that is, logic chips and custom chips and integrated chips, but not for pots. That pot is usually a physical thing, hmmm, it was a physical thing. They do have digital resistors which are controlled by logic. But I thought this was a physical analog resistor thingee with a wiper so how could it burn out? They make them that cheaply now? If so, one more thing to worry about.

Would not the 5 volts be for the logic stuff over the analog pots? Don't mean to second guess you here but I'm just surprised that analog pots would be so vulnerable to only 12 volts. I can't recall my specs but usually most stuff I would buy would be much higher in specs, to

24?? volts, but it's been a long time since I had to talk to Digi-Key.

I was suprised to hear Ted's talk of the DC carried with the AC in cheap audio circuits and burning out the pots. I gather these are rather small and delicate pots? Must be fine wires to burn up. Or maybe carbon instead. But the audio circuit would have the DC and AC much higher than 5 or 12 volts? Or maybe high amperage? Stuff never gets simpler.

Posted too quickly. He could not burn out the pot since he is not doing anything but scoping an existing circuit. Ted is not introducing anything and most oscilloscopes are so loaded up, I forget the tech term but millions of ohms of resistance, they don't add to any circuits.

All he is doing is looking at the electricity that already exists in the vehicle as it flows between the sender and fuel pump. If there are any burps, he knows something is not working correctly.

This is safe. What I once did, 'scoping an extremely high voltage strobe light was not safe. It burned up the metal of my 'scopes probes. Vaporized the metal.

I guess your point is if he hooked up 12 volts to a pot that is outside the vehicle? As I said in my previous post, I did not know pots could be vulnerable to such a small difference from 5 to 12 volts if there were simple analog pots, that is, just physical metal or carbon with no logic to them.

They can't. What 'burns up' pots is current, not voltage. You could, possibly, put a Van De Graph generator on a vehicle fuel level sensor and using a high voltage probe, see some voltage difference with 100,000 volts on it or more. This is the same generator you see in the science museums that make people's hair stand on end.

If you put a 12 volt battery that can source 100's of amps across the terminal of a level sensor you would not have a level sensor any longer.

On the vehicle in question the same plug the fuel pump uses also carries an additional pin for the level sender, I assumed this was obvious since I mentioned that the aftermarket replaces these sensors only as a complete fuel pump assembly. And you are correct the idea was to run jumpers from the existing fuel level sensor pins in that plug, not from the fuel pump power source pins in that plug. (actually, the fuel pump and level sensor share 1 pin in the plug, as a matter of fact)

Ted