'So my old Blazer (190K +) has been playing games with me for some time now, instrument wise. The fuel gauge has always been something of a mystery. Once the needle gets below 1/2 tank it can't be trusted. It will often read down near empty , then creep up. Sometimes it will drop to zero and the low fuel light comes on. A minute or two later it starts coming back up and may go back to the 1/2 full mark. That sort of thing. It's not a serious problem, I use the odometer trip meter and zero it out when I fill the tank. That way I know how many miles I have on each tankful and can juggle the gauge readings accordingly.

Problem is this. A couple weeks ago the instrument cluster had somewhat of a brain fart. I started it up as usual, and as soon as it went through the "warning lights on and back off" sequence a strange thing happened. All the warning lights came on, and most (i think) of the gauges fell to zero. The tach needle went full to the right and pegged. The radio and temp display went out, so did the shift indicator and odometer. I watched all that happen, shut it off and restarted it. No change. I didn't know if I should trust it like that and wondered if I should be driving it, not knowing why it was doing this and what else could I expect to go wrong. After a minute or so sitting there idling while I was deciding what to do next it all went back to work. It did that twice on me, both times a day or two apart, right after the first start of the day, and hasn't done it since.

The digital displays have all played the games where they come on and go off at will. The radio and temp display as one pair, the shift display and odometer as another. This has been off and on for a year or two. Minor inconvenience, but not much of a threat.

At issue today is the odometer. When this happened and things came back on the meter jumped to some number like 1808.xx or something. The trip meter went to 0.0 Neither one has changed since. So now I have no trip meter, no odometer, and still have the wonky fuel gauge that is only truly dependable when the tank is full or near full. The gauge is always on target then.

All this leaves me wondering if this could be mainly an issue with the ECU. I'm just guessing here, but I think most all of these items and circuits come from or pass through the control module. Right or wrong on that? I did some quick searching for a replacement recently, not knowing what to expect. Among other options was one source that listed a "refurbished" replacement unit tailored to the vehicle vin number for under $200. Price seems reasonable enough, but what do I know? Something like that could be just what I need. Then again it might be nothing but more trouble.

Having zero experience with this sort of thing I pose this question to all, Am I even looking in right direction here, or is my problem more likely something else entirely?

 

The ECM on these trucks are pretty robust. I would first consider either an ignition switch problem or a ground problem before replacing the instrument cluster. Fuel gauge may be it's own issue. From what your describing I'd lean more towards ground. However, if you have a lot of weight on your key ring it could be the switch.

A couple visual/wiggle checks you can do. Turn the ignition to RUN and gently wiggle your key around to see if the display does wild stuff. That could indicate bad switch.

For ground, on my '98 the ground for the cluster goes thru the BBEC then the UBEC then terminates on the back of the driver's side cylinder head (maybe someone with '02 can verify it's the same). That's where I'd check first. For my truck it's best to go up from underneath. You'll see a couple black wire bolted onto the head. Kinda about 5" below where the spark plug wires route. Little tough to access (getting your hand between the firewall and the head). See if any of those wires are broken. Maybe have someone watch the display with the ignition in the RUN position (KOEO (Key On Engine Off)) and wiggle those wires some to see if the display changes. There could also be a rust problem and if you're feeling froggy you can remove the bolt and clean that area and the terminals with a wire brush/sand paper. Bit of PIA because of the limited access.

If you have and know how to use a volt meter, there are some checks you can do that may help point in the direction of the problem.

EDIT: Also check the braided ground strap between the firewall and the engine block on the passenger side. I'd also disconnect and clean where the B(-) cable connects to the engine on the lower front passenger side of the engine. Let us know if you have a volt meter.

 

I have a voltmeter, but admit I don't use it all that much. I find a simple test light pretty much shows me where I have power and where I don't. In most cases (other than charging systems) that covers most of my bases.

The ignition switch is a good suspect. My key ring is pretty heavy, and I've had problems with the steering wheel lock for years. It always takes a few seconds of wiggle and jiggle to get the key to turn towards the on position. This is my work truck, so I'm in and out of it all the while. I leave the key in the switch pulled out just far enough to stop the beeper.

So maybe the most practical first step would be to replace the switch, eliminate (hopefully) the lock/unlock problem with the key and see what else that does for me?

 

I put a video together real quick because I thought the visual would be more helpful. The purpose is to confirm or eliminate pointing you at the ignition switch as the potential problem. You need a meter for this one because you're measuring for possible Voltage Drop (Vd). I'm a huge fan of the incandescent test lamp, but since you have parts of your cluster working, for this test, doesn't work as well.

If you post a pic of the face of your meter, we can help you set it up.

Couple other things:

While optimally using the B(-) post is the way to go for testing, you also ground the meter to a good ground on the truck. Insure you test to make sure you have a good ground.

Check the label on your fuse panel...the fuse numbers I was giving may be different on your '02.

If you get battery voltage (~12v) on the cluster fuses at first, have someone jiggle the keys while you watch the meter to see if it changes.

 

Thanks. Looks like a good place to start. I think I can do that.

Just an aside, the truck has been on the road nearly every day for the weeks since those "instrument brain fart" incidents. There have been no repeats nor anything other than its usual quirks during that time. When I say usual, I mean things it's done for years, not just recently The odometer and trip meter still haven't changed. where does that signal come from if not the ECU? Is the memory all stored in the dash?

 

The signal does come from the ECM via a Class 2 message. It originates with the VSS (Vehicle Speed Sensor). But you haven't stated that you've had any issues with the speedometer, so I would assume there's no issues with your VSS. It could very well be that you need a new cluster. What I failed to point out in the video was that specific test was to give direction (is the problem upstream or downstream of where you're testing). It's also just a general easy tip when suspecting the ignition switch for any component that gets it's power thru the key in RUN.

For example: If you were having problems with the gauges and you measure and see a Vd on your meter that reads 8 volts. You now know your problem is upstream on the gauges circuit (at the point where your measuring) and you know the problem lies someplace between where you measured and the battery. So you can turn your attention away from the cluster and focus on further testing/repairs upstream.

If the measurement is 12 volts start heading downstream for further testing. In this specific case (after testing gauges themselves) I would then break out the test light (or meter) and start checking the ground path after the cluster/gauges. If that all checked out, then I would replace the cluster (if I couldn't fix it myself).

All of this is pretty tough (if not near impossible) w/o wiring diagrams. If you don't currently have the service manual for your truck, I highly recommend getting a subscription at Alldata.com for it. It's the same tech manual the techs use (plus it gives you color wiring diagrams, which can be very useful. A subscription isn't that much.

Let us know how you make out.

EDIT: I don't know for certain where the mileage for the odometer is stored. I believe you have to provide the odometer reading when ordering a new ECM, so I always assumed it was stored there. But maybe someone else can provide a definite answer.

EDIT #2: The EDIT above is incorrect. I corrected it in post #13.

 

I get the picture about voltage testing, looking for a voltage drop somewhere in the system. The problem I see there is if battery voltage is present at the various fuses and any other tested points while all the gauges are working then I haven't found a problem, but merely proven the absence of one. The incident I mentioned in my first post was (so far) a onetime occurrence. Granted, it did actually happen twice, but both times within a couple minutes of each other and hasn't happened since. The engine was running both times.

I do see the logic in that issue being caused by either an intermittent voltage drop or a ground connection. I follow you on that. Where I get lost is in suspecting the key switch for supplying battery voltage to some circuits and reduced voltage to others. I may well be wrong in this, but I would expect that the switch would have multiple circuits, yes, but there will be only one start circuit, one acc. circuit, one ignition circuit, etc. I wouldn't expect it to supply multiple ignition related circuits from the "run" position and do some from more than one place in the switch. I don't know that for a fact, but that's been my experience with ignition switches I've dealt with elsewhere. I can see where the "run" position supplies multiple circuits, yes, and those could have a bad connection or two here or there making for a voltage drop.

My reasoning for suspecting the control module from the beginning was based on which functions I've seen come and go over time. The digital displays for the radio and compass/thermometer, the shift and odometer displays, for example. Each pair have played cat and mouse with me for years now, yet when they come on or go out nothing else I'm aware of takes place. Gauges and warning light don't change, head. tail, and signal lights don't fail. Again I can't prove this but I would think the shift display is computer generated and based on signals from the transmission. The odometer would also be computer generated based on speed sensors.

 

If I understand your one question correctly, there are several circuits on the ignition switch. One or more of those circuits could have some form of resistance (corroded wire, etc) while other circuits are perfectly fine. I drew a very rough idea of how a corroded wire could prevent full source voltage from getting to a component.

P.S. My 'definitions' aren't literal, just a way to picture it. Also, this is only effective with the circuit loaded.

 

I’m going to respond to your last in multiple posts to break it down for a little easier reading.

#1

First, let’s assume your Instrument Panel (I/P) isn’t working completely (gauges, display, etc). The test tells you more than just the absence of a problem, it also gives you direction which way to go. Intermittents are the worse for obvious reasons, so you do what you can to recreate it (wiggling, etc.). Again, assume you’ve managed to recreate the problem OR something (anything) isn’t working on the I/P.


Using your I/P as the source of the load, you’re trying to figure out if it’s a power side problem or ground side problem. So maybe visualize the I/P as the center with power side to the left and ground to the right. On the left (power) we have the battery, a relay in the Underhood Fuse Box (UBEC), ignition switch, fuses and all associated power side wiring and connectors. On the right you have the ground side, associated wiring, connectors, terminals, etc. There are also power/grounds/data inputs & outputs from the ECM & the BCM (Body Control Module), but those are for a later stage of testing. Right now we are trying to determine if the I/P itself needs to be replaced. You want to make sure your power and grounds are good for the I/P by starting with the test I showed you. Also do it with truck running.



The fuse is the closest and easiest to access component to I/P. Why pull panels and such apart if you don’t have to, is my theory. If you show 12v during that test, you know that you don’t need to bother checking any of the components, wiring, connectors (i.e. ignition switch (power side)) all the way back to the battery.



You then move to the fastest and easiest testing to the right of your I/P visualization (ground side). For example, there is a large multi-pin connector behind the glove box that is easy to access by dropping the glove box door with the spring retainer.



After sourcing and identifying the correct wire terminal (wiring diagrams) you could use a test light connected to B(-) and touch the backside of that terminal (with connector connected, truck in RUN and ((again running)). If the bulb lights, you know somewhere after that connector is a bad ground. If the bulb doesn’t light, then the ground is good all the way from that connector to the battery (note if you used a meter instead of test light, you’d want to see no more that 100-200mV). You don’t have to start digging in looms for looking for a broken wire, etc under the hood.



You’ve now done all the easiest and fastest checks for powers and grounds for the I/P and shrunk down the problem to somewhere in the cab (caveat: we still are not dealing with the ECM yet or sensor inputs/outputs...we're trying to determine if the I/P is the problem).You’re now at the point where it’s time to start pulling panels off for deeper testing, and/or start testing input sensors. But you eliminated a whole bunch of components/wiring/connectors/terminals very quickly without having to check each individually.

 

Okay, let me try some of that and see what I can find out.

Thanks