So...anyone got a failed cluster laying around? If I were there....I'd just buy one out of a truck in a junkyard and send it to you Pat.
Keep in mind...I've done the 0411 swap and it didn't change the 'hummingbird' effect.

The only data on the motor itself is as follows,as listed on a yellow tab.

>Sie
07 58

Let me know if you want me to drop it in the mail.

playtoy_18 wrote:

The only data on the motor itself is as follows,as listed on a yellow tab.

>Sie
07 58

Let me know if you want me to drop it in the mail.

Is yours doing the hummingbird dance for sure?

Soooo... did we ever come to a definitive conclusion as to how to fix this? My mom's 99 Tahoe just started doing this at 117,000 miles. The fuel pump module was replaced about 2 years ago with a Delphi unit, and this fluttering gauge issue just started this week. I'm pretty sure this is not an issue with the fuel pump module/sending unit in the tank, at least in this case. Any info on how to get replacement stepper motors? This is defintely the "hummingbird" syndrome, as the needle reads correctly, but flutters so fast it actually looks blurry.

Edit:

I did come across this a few minutes ago:

Customer Concern: The fuel gauge is vibrating very fast. It looks like it is reading accurately. The vehicle has a single fuel tank.

Tests/Procedures: 1. Has installed a gauge tester or substitute resistors in place of the fuel tank sending unit. 40 ohms for empty, 250 ohms for full. (The sender is linear - 145 ohms for 1/2 tank.)

2. If the gauge is stable in all positions tested, replace the sender in the fuel tank.

3. If the fuel gauge on the cluster still has the rapid movement. Disconnect the Powertrain Control Module (PCM) Blue connector. With the key on, connect a gauge tester from pin 9, Purple wire, to ground.

4. Set the tester to 45 ohms, then 90 ohms. Check for a steady needle at a 1/2 tank and then with a full tank. If the fuel gauge still vibrates, repair the Instrument Panel Cluster (IPC).

5. If the fuel gauge is steady, verify a good connection at the PCM. If OK, replace the PCM.

Just did some testing on this issue. I have a 99 Cadillac Escalade cluster in my 96 pickup which I bought new in 2002 and swapped in. It was a plug and play swap, everything worked perfectly, and it still does to this day. So, I took this Escalade cluster out of my truck and swapped it into my mom's 99 Tahoe. For the first couple of miles of the test drive tonight, it worked fine, and I thought I had found the culprit. But, then my known good Escalade cluster started doing the fluttering too, exactly as the original Tahoe cluster does. So, I think I can safely rule out a bad cluster or "stepper" motor for the fuel gauge, as the truck does this with both clusters.

I'm curious... there are about 3 different Delphi replacement fuel pumps listed over on RockAuto for this truck. Would the RFI filter (or lack of one) be a possible cause of this fuel gauge issue? This fuel pump was replaced, as mentioned before, and it is supposedly a Delphi or Delco unit, not aftermarket. But, if the 3-digit code off of the old one was not noted and the new one not specified as the same 3-digit code, could that be the culprit? I'm almost to the point of dropping the tank and doing some probing with the multimeter.

You can just buy the sending unit from Oreilly. I replaced my pump with a delphi a few months before and I had a similar problem and they told me they could only get it with the pump, but I looked online and found just the sending unit for 100$.

The sending unit isn't the problem. I've replaced many....with no change.

Some video of the diagnostic process so far.....

This is with everything connected normally to the stock gauge cluster:

http://www.youtube.com/watch?v=sgthlmt6E6Q

Then, I swapped in my known good 99 Escalade cluster out of my 96 pickup. The gauge behaved the same:

http://www.youtube.com/watch?v=KfVFzggZVyY

Pin #9 on the blue PCM connector is the PCM signal to the gauge. I inserted a 10-Ohm resistor into the blue connector and grounded the other end of the resistor. This essentially takes the place of the PCM's signal sent to the gauge and provides a set value going to the gauge. The range seems to be 0-100 Ohms, with the low end resulting in a reading of "Empty" and the higher resistance resulting in a "Full" reading on the gauge. With various resistors in place, the gauges on both the stock Tahoe cluster and the 99 Escalade cluster read perfectly, with no vibration of the needle.

http://www.youtube.com/watch?v=ZBe8ys50kls

Here is a 100-Ohm resistor installed, yielding a full tank reading on the fuel gauge:

http://www.youtube.com/watch?v=KMgoHMl5hng

Finally, I attached a 47-Ohm resistor to get a half-tank reading on the gauge. Again, it read steadily with no vibration or any other odd behaviors. In this video, I also have my Snap-On Scanner attached to the ALDL and I'm reading the voltage coming from the fuel level sender in the tank and going into the PCM. You'll notice the reading on the scanner fluctuates. The gauge does not because it's getting a steady signal from the 47-Ohm resistor.

** EDIT to above paragraph: the fluctuating voltage seen on the Snap-On scanner was the result of me getting in and out of the truck and the resultant fuel moving around in the tank. I figured this out a few days later, when this rather obvious phenomenon occurred to me. The sending unit is NOT sending out an erratic signal in this case, it is steady once the fuel movement in the tank settles down, and this can be clearly seen on the Snap-On scanner. **

http://www.youtube.com/watch?v=OIHXEqodNT0

According to Lextech's 411 conversion wiring file, the PCM receives the fuel level sender signal on Pin 13 of the white/clear PCM connector. This is what the Snap-On scanner SHOULD be looking at, but I want to probe that lead with a multimeter myself and see if that's correct. If the wire going into the PCM from the fuel level sender is showing this fluctuation, then my assumption is that I need to start looking at the pump module & sending unit along with the associated wiring. The above tests would indicate to me the problem is not in the gauge cluster or with the PCM, but in the signal being sent to the PCM from the sending unit.

We will see.



Last edited by JR96CK on Fri Oct 22, 2010 10:50 am; edited 1 time in total

Tagging along

Wow! That's pretty interesting stuff JR. Good work.
If you crack this....you'll be one of the most popular guys in the GMT400 world.

Thanks, Howard. The truck is an hour away from here at my mom's, so I only get to do a little bit here and there. I WILL get it figured out though, eventually. I'm going to try to get up there again today and do a little more testing.

So I managed to get up there and do some more diagnosing today... sorry, not as many videos this time, so I'll have to resort to a written explanation.

Using Lextech's wiring diagram to convert the 98-2000 blackbox PCM over to the 411, I found that the white/clear connector on the blackbox PCM contains the two wires we need to look at in terms of what signal is getting to the PCM from the in-tank sending unit. Pin #13 is a purple & white wire which is the signal coming from the sending unit and feeding into the PCM. Pin #19 (also on the clear/white connector) is the ground wire (black) for the sending unit. I un-pinned these two wires from the clear/white PCM connector and measured the resistance between the purple/white and black wires. With a little over half a tank of gas in the truck, the reading was around 156 Ohm. My reference material states the range of values coming off the sending unit should be 40 Ohm for empty and 250 Ohm for full, with half a tank being 145 Ohm. This reading of 156 Ohm therefore looks correct for the amount of fuel in the truck, and the reading on my multimeter was very steady. This first step in today's diagnostic session supported Howard's assertion that the sending unit is not the culprit. This truck has a relatively new Delphi pump module anyway, so everything back there should be good. Moving on....

I then proceeded to leave the aforementioned purple/white sending unit wire and its associated black ground wire unpinned (the connector was re-connected to the PCM, however), and turned the key forward. At this point the gauge was still connected to the PCM, but the PCM was not connected to the sending unit, since those two wires were un-pinned. All other connectors were back in place on the PCM, so essentially the PCM was seeing an open circuit across the sending unit, as it was completely disconnected. I turned the key forward and witnessed the same erratic gauge behavior and vibration as before, and the needle was past full because of the high resistance value of the open sender circuit. Again, this points to the fact that the sending unit and its associated wiring are NOT causing the vibrating gauge, at least not on this truck. Next step....

I took a blue wire I had laying around from a 411 connector and pinned that into the #13 location of the white/clear PCM connector, and I took a black wire from that same 411 connector and pinned it into the pin #19 location on the white/clear PCM connector, then I plugged the connector back in. This allowed me to connect various resistors between the two new temporary wires to supply differing "fake" fuel level signals to the PCM in place of the in-tank sending unit. Again, the range of values here needed to be 40 Ohm for low readings on the gauge, 250 Ohm for full readings, and 145 Ohm to yield half tank readings. I put a 100 Ohm and a 47 Ohm resistor together in series (147 Ohm) to get close to the half tank 145 Ohm reading. With the key turned on, the gauge read around half a tank as expected, but it still had the vibrating needle. Again, even with a steady and known value being fed into the PCM in place of the sending unit and its wiring, the gauge shows the same behavior. It's looking to be the PCM, at this point, as yesterday's tests clearly showed the gauge was steady when resistors were plugged in to feed the gauge instead of the PCM. We're not done yet....

I kept the resistors in place between the temporarily-pinned ground wire and the temporary blue sending unit signal wire. I moved over to the blue PCM connector and un-pinned the solid purple wire at the pin #9 location, which is what sends a signal from the PCM to the gauge, as we saw yesterday. In its place, I pinned a green wire (again, from a donor 411 connector) into the pin #9 spot on the blue connector and plugged it back into the PCM. At this point, I had the PCM reading the 147 Ohm resistor setup mentioned earlier as the input value in place of the sending unit, and now I had the output from the PCM going not to the gas gauge, but connected to my multimeter's positive lead. The negative lead was attached to a good ground on the radiator support and it was set to read continuity. Normally, this will sound a steady tone when reading DC input. I turned the key forward and the multimeter sounded a buzzing sound, as it does when connected to an AC signal. Clearly then, with the PCM cut off from the sending unit and receiving a known steady resistance value for input, and with the PCM disconnected from the gauge and sending a signal to the multimeter, it seems the PCM is generating and sending out this pulsed signal which is causing the gauge to vibrate. Which leads me back to something Patrick said earlier....

I don't think so, but is this output from the PCM to the gauge supposed to be a PWM signal which the gauge then smooths out? If that were the case, then the gauge is not doing its job in this instance and it needs to be replaced. However, if the PCM is just supposed to send out a steady value to the gauge, then clearly the PCM is the problematic element here. I don't see any reason why the fuel gauge would need to be PWM regulated on the GMT400 trucks, so I'm doubtful that's the case. The 99 Escalade cluster was a direct plug-in on my 96 truck so I would think if the 99 Escalade cluster (which also showed problems when I swapped it into this Tahoe in the earlier videos), were intended to be connected to a pulsed signal, it would not have worked correctly on my 96, which uses a signal straight from the sending unit and has no input from the PCM.

Bottom line is, I think I'm looking at a problem within the PCM here, not the gauge and certainly not the in-tank sending unit or any associated wiring & connectors. It seems from all I can tell the PCM is getting a clean signal from the tank but is generating and sending a dirty signal to the gauge, and the gauge manifests this in its buzzing movement in response to the dirty input.

One other thing, I connected the Snap-On scanner again and did a functional test which allows me to take control of the gas gauge, essentially telling the PCM what value to display. This of course was done with all the wiring re-connected normally at the PCM and no resistors or temporary connectors in place. I could command any value to be sent to the fuel gauge, and the gauge responded accurately... but it vibrated the whole time. This was just another way of doing what I did earlier when I plugged resistors in place of the sending unit between the purple/white and black wires going into the PCM, but I thought I would play with the scanner a bit and see what it could reveal.

Here are a couple of videos showing what I just described above, although the quality of the videos is less than ideal (bad lighting). You probably cannot see the vibration of the gas gauge in these clips, but it was vibrating in all of them.

Buzzing output from the PCM:

http://www.youtube.com/watch?v=zQpk4pEdA5o

A video of the Snap-On scanner override:

http://www.youtube.com/watch?v=PX82k2Y6FyI

I recall reading above that Howard did a 411 swap and the hummingbird effect remained. Interesting. Hmmm....

I wonder if the clusters are specific to certain years. We know there are suttle differences in the wiring for 96-97 and 98+. I have two Escalade clusters and on the PCB on back it says 97 C/K Truck Tach. Then on my Police cluster it says 98 C/K Police. My truck, a '98 -mfd 8/98- has had the vibrating gas gauge since day one with both the original and one of the Escalade clusters. My fuel pump went out in July 2006, while on the road none the less, and it was replaced at a GM Chevy dealer. I wonder if the stepper motors are different It would be interesting to get a 2001 van cluster and see what happens and if its built the same way as the GMT400's or more like the GMT800's.

TJ

That's right JR....mine has doe it through 2 pump swaps and probably 3 different PCM's (one old BB and two 0411's).

Howard

Found this a few minutes ago:

2000 GMC Yukon 5.7L fuel gauge vibration

"Called the Tech Line and they said that the PCM PWM's the signal to the fuel gauge after receiving the information from the Sending Unit in the tank. They said that the PCM has an internal buffer that will "clean-up" the PWM signal to the fuel gauge. According..."

I cannot get to the rest of it because the site requires a membership. Here is the link:

http://www.iatn.net/techmail/summaries/2006/06/05/gm.html

The PCM has an internal buffer to clean up the PWM signal? This sounds like what Patrick was saying except that the buffering takes place in the PCM and not in the gauge cluster.

Hmmmm....

Hey Howard,

The PCM swaps you've done... were they used PCMs? If so, do you know how many miles had been put on them? Right now my thinking is along the lines of a PWM buffer piece of hardware within the PCM which begins to fail around some typical mileage (apparently just after 100,000 miles on average).

I also found this last night while digging around, but I believe it's for the S-Trucks and not full size. However, there may be enough similarity here to aid in what we're looking at. Sounds like the wire colors used on the 4.3 are the same as the full-size trucks. This seems to be just an expanded version of the diagnostic procedure I posted above several days ago.

---------------------------------------------------------------

98+ Fuel Gauge Diagnosis
FUEL GAUGE INACCURATE OR INOPERATIVE

NOTE: Corroded connections or severed ground paths generally cause fuel gauge to register past FULL position. Short circuits to ground generally result in fuel gauge resting below EMPTY position.

NOTE: After each repair procedure has been completed, reconnect all components. Perform INSTRUMENT PANEL CLUSTER (IPC) SYSTEM CHECK under SYSTEM TESTS.

1. Perform INSTRUMENT PANEL CLUSTER (IPC) SYSTEM CHECK under SYSTEM TESTS. Go to next step.
2. Turn ignition switch to OFF position. Disconnect 6-pin chassis-to-engine in-line harness connector, located under vehicle, leading to fuel tank. Connect Signal Generator and Instrument Panel Tester (J33431-C) between male terminal side of connector terminal "E" (Purple wire) and terminal "D" (Orange/Black wire on 2.2L engines, Black wire on 4.3L engines). See Fig. below. Turn ignition switch to ON position. Vary resistance on J33431-C while observing fuel gauge on IPC. When resistance is near 40 ohms, fuel gauge should read EMPTY. When resistance is near 250 ohms, fuel gauge should read FULL. If operation is as specified, go to step 4 . If resistance is not as specified, go to next step.
3. Turn ignition switch to OFF position. Connect scan tool to DLC. Turn ignition switch to ON position. Vary resistance on J33431-C while observing FUEL TANK LEVEL REMAINING parameter in PCM data list. See FUEL TANK LEVEL SPECIFICATIONS table. If FUEL TANK LEVEL REMAINING parameter is according to specification, go to step 11 . If parameter is not as specified, go to step 5 .
4. Check for poor connection at fuel level sensor harness connector. Check for high resistance in Purple wire and Orange/Black wire (2.2L) or Black wire (4.3L) between 6-pin chassis-to-engine in-line harness connector and fuel level sensor. Check for a misaligned fuel level sensor. Check for a damaged/deformed fuel tank. Repair as necessary. If problem was found and repaired, go to step 13 . If problem was not found, go to step 7 .
5. Check for high resistance in Purple wire between 6-pin chassis-to-engine in-line harness connector and PCM. Repair wiring as necessary, then go to step 13 . If no problems are found, go to next step.
6. Check for high resistance in Orange/Black wire (2.2L) or Black wire (4.3L) between 6-pin chassis-to-engine in-line harness connector and PCM. Repair wiring as necessary, then go to step 13 . If no problems are found, go to step 9 .
7. Remove fuel level sensor. Check fuel level sensor for mechanical fault, damaged level arm or for fuel strainer interfering with level arm. Check fuel tank for foreign material. Repair as necessary, then go to step 13 . If no problems are found, go to next step.
8. Connect DVOM between fuel level sensor terminal "A" (Purple wire) and terminal "D" (Orange/Black wire on 2.2L, Black/White wire on 4.3L). Monitor resistance while moving level arm. If resistance changes smoothly (40-250 ohms) as arm is moved, system is functioning properly at this time. Check connections and wiring for intermittent faults. See WIRING DIAGRAMS . If resistance does not vary smoothly as level arm is moved, go to step 10 .
9. Check PCM connector for poor connection or corroded terminals. Repair connector as necessary, then go to step 13 . If no problems are found, go to step 12 .
10. Replace fuel level sensor. When repair is complete, go to step 13 .
11. Replace IPC. When repair is complete, go to step 13 .
12. Replace PCM. When repair is complete, go to next step.
13. Using scan tool, clear PCM DTCs. Operate vehicle to verify proper fuel gauge operation. If fuel gauge operation is normal, repair is complete. If fuel gauge operation is faulty, go to step 2 .

Some interesting reading on filtering PWM signals:

http://www.proaxis.com/~wagnerj/PWMfil/PWM%20Filters.pdf

The PCM I'm using now JR is the one I installed my Roadrunner into and is brand new. Sorry....another dead end. Lol!