Watch and Listen to me make a fuel guage move slower! SO EXCITING!

https://www.youtube.com/watch?v=_cw1_Q0MDBA

it's about as boring as it sounds, but yes, i kind of narrated a demo of how i solved a very long-lived issue with the 95MC.

it's REALLY common for the 95-99 monte carlo and 95-01 lumina fuel guages to act really jerky with anything less than a ~90% full tank and anything more than ~10% left in the tank, the needle will run into the minimum and maximum stops as the fuel sloshes around in the tank. i'm told this is actually a really widespread issue on GM vehicles in the 90s, but i don't have an exact list yet.

now, this was never an issue with the earlier W-bodies, as they used the exact same tank and fuel level sender from 88-96(97 on the cutlass supreme). the difference was in the fuel guage, the older stuff had some type of friction element in it that physically slows down needle movement and prevents it from jumping around when fuel moves around in the tank. when GM designed the 95+ monte carlo/lumina clusters, they decided to leave that out entirely and rely on the tank baffling to keep the guage relatively smooth. turns out, it doesn't work very well at all as i don't think i've ever seen one that wasn't jerky, even on low-mile vehicles. 97+ vehicles have a slightly different setup, as the sender is read by the PCM, which drives the fuel guage, but still have the same issue.

the solution i eventually ended up with was an electrical one, since i couldn't see any realistic and long-term method of slowing down the needle via friction, the guage just won't allow for it that i can see. i setup my anti-slosh module to have 64 selectable settings(set via DIP switches) since i don't know what i'll want to end up with when i take it off of the bench and put it in-car.



random thread topic, i know.

OK I watched it and have to say, yours is more interesting then paint drying! :laugh:

There's similar issues with 98 Chevy trucks where the needle shakes after years, guys have replaced the gauge and sure enough still shakes after a few years. Seems GM packed the stepper motor with grease when they changed to a PWM signal, works well for a few years. The earlier 97 trucks did not have the issue because of old style fuel level and 99 was different body style, so they just patched up this issue on this year truck. As the grease is pushed away or? Well the needle shakes. Mine does but I don't look at it that much.

Here's a thread, 18 pages, where some guys really looked into this and a fix and found one with electronics (capacitors or?)
http://www.pacificp.com/forum/viewto...er=asc&start=0

With that and yours you could go into gas gauge repair, or more accuaratly Instument Panel Repair and Rebuilding. As vehicles get older things like instrument panel repair/rebuild, light bulbs, steeper motor etc... seem to make a good business. Man some of the old muscle cars get big bucks to restore an IP.

that's not boring at all, instrument accuracy is a great thing to play with! i'd be really happy to have that fixed if something like that happened.

a fuel tank level tuning module is a great thing

the wedge shaped tanks in the f-bodies are a particular pain in the ass, they read full until about a half a tank, then drop quickly to half, then hang around 1/4 gauge until they're almost dry, THEN they go to empty.

i often thought of making some kind of input to output curve in my onboard dash computer to try to force it to be linear, just never got around to it....

i'm inspired to do that now.

Quote:

---

Originally Posted by EagleMark

OK I watched it and have to say, yours is more interesting then paint drying! :laugh:

There's similar issues with 98 Chevy trucks where the needle shakes after years, guys have replaced the gauge and sure enough still shakes after a few years. Seems GM packed the stepper motor with grease when they changed to a PWM signal, works well for a few years. The earlier 97 trucks did not have the issue because of old style fuel level and 99 was different body style, so they just patched up this issue on this year truck. As the grease is pushed away or? Well the needle shakes. Mine does but I don't look at it that much.

Here's a thread, 18 pages, where some guys really looked into this and a fix and found one with electronics (capacitors or?)
http://www.pacificp.com/forum/viewto...er=asc&start=0

With that and yours you could go into gas gauge repair, or more accuaratly Instument Panel Repair and Rebuilding. As vehicles get older things like instrument panel repair/rebuild, light bulbs, steeper motor etc... seem to make a good business. Man some of the old muscle cars get big bucks to restore an IP.

---

the 97+ W-bodies don't seem to have a needle buzzing/shaking issue like the truck clusters do, even though i'm fairly certain they're driven at the same ~60Hz frequency. or at least, that's what i'm told from other w-body enthusiasts. i'm wondering if needle weight or something else is contributing to that not happening, the truck clusters have pretty short/light needles, the W-body ones are probably twice the length and weight.

i've actually rebuilt quite a few clusters already..... the 88-93 grand prix clusters have a couple of rather interesting issues that are pretty widespread, including the low fuel level and check guages lights coming on when they're not supposed to and the speedometer and tach becoming wildly inaccurate(showing 120MPH when you're going 40 was what the wife's 90GP did). i've also been throwing LEDs in everything possible.

http://i.imgur.com/5yUTuC1.jpg

that's the stock cluster lit up. the greenish, yellowish tint to the backlighting, it doesn't really match the rest of the interior lighting at all, which is a fairly light blue. i put some white, blue and green LEDs in place of one to see how it looks and came up with these.

http://i.imgur.com/TvApaDf.jpg


http://i.imgur.com/AmC9JUf.jpg

http://i.imgur.com/3vviEwN.jpg


i ended up going with the center one and redid my needles as a rather vibrant red. it feels less cheap to look at now.

Quote:

---

Originally Posted by steveo

that's not boring at all, instrument accuracy is a great thing to play with! i'd be really happy to have that fixed if something like that happened.

a fuel tank level tuning module is a great thing

the wedge shaped tanks in the f-bodies are a particular pain in the ass, they read full until about a half a tank, then drop quickly to half, then hang around 1/4 gauge until they're almost dry, THEN they go to empty.

i often thought of making some kind of input to output curve in my onboard dash computer to try to force it to be linear, just never got around to it....

i'm inspired to do that now.

---

for a while, i've been looking to get around the fact that the needle completely overshoots the marks for empty and full, i think a microcontroller would be the best way to go about it since after some initial code is written, all that is needed to be done is drain the tank and add in measured amounts of fuel and keep track of where the sender is during the process to calibrate the fuel curve.

For many years, one of the differences between a GMC and Chevy truck was in the amount of fuel gauge damping. Somewhere in the '60s the Chevy gauges started to respond slower while GMC's remained a direct representation of the current flowing through the sender. Those old needles would really get to bouncing. But you got used to it. And accuracy? A linear response sender to a non linear tank shape is just part of the game with a GM car. Many years ago one of the first things I'd do when I bought a new GM vehicle was throw 5 gallons of gas into a can in the trunk then drive the car until the tank emptied and the engine died. That way I knew where empty really was. Of course you learn some tricks like bringing a squeeze bottle to prime the carb because the old mechanical pump system took a lot of cranking to get fuel back up to the engine.

The Sunbird clusters develop problems in the r/c calibration circuit for the tach and/or speedometer. I've added resistors in the past to adjust for this and it's worked well enough. I've done a few other clusters but for some reason I never found thermally stable resistors. So I never felt the repair was good enough to charge for. At cold temps the circuit always produced high readings.

99-2000 6.5 powered diesel vans have a fuel gauge driver module (diesel pcm can't drive gauges). It provides so much damping that you almost think it isn't working. Part number is 12135192. Circuit has an NTP18N06 Power mosfet, LM2902 op-amp, MC705P6 (MC68HC05P6) processor, and sundry caps, diodes, and resistors. Refreshes on key on, then very slowly after that.

Quote:

---

Originally Posted by 1project2many

The Sunbird clusters develop problems in the r/c calibration circuit for the tach and/or speedometer. I've added resistors in the past to adjust for this and it's worked well enough. I've done a few other clusters but for some reason I never found thermally stable resistors. So I never felt the repair was good enough to charge for. At cold temps the circuit always produced high readings.

99-2000 6.5 powered diesel vans have a fuel gauge driver module (diesel pcm can't drive gauges). It provides so much damping that you almost think it isn't working. Part number is 12135192. Circuit has an NTP18N06 Power mosfet, LM2902 op-amp, MC705P6 (MC68HC05P6) processor, and sundry caps, diodes, and resistors. Refreshes on key on, then very slowly after that.

---

seems the sunbirds have the same issue as the GP clusters.... the laser-cut resistors for needle angle deflection seem to have their resistor material fall apart. i remove that part of the circuit and stick in 25 turn pots, i've never had a complaint about cold operation though... maybe i'll stick one of the 4 clusters i have on-hand in the freezer for a while and setup a quick test while they're still cold, that should bring them down to about as cold as they'll ever get in most vehicles(i want to say i keep our freezer near 5*F?).

that diesel fuel module sounds interesting.... the nearly 3 minutes that setting 63 on my module ended up with going from empty to full already feels like an eternity. i ended up with a LM358 to deal with the smoothing, a quad op-amp like the 2902 seems like overkill to me for this kind of circuit(especially when i only used one of the 358's two circuits)? or is there some advantage of if that i'm not realizing?

Quote:

---

Originally Posted by 1project2many

For many years, one of the differences between a GMC and Chevy truck was in the amount of fuel gauge damping. Somewhere in the '60s the Chevy gauges started to respond slower while GMC's remained a direct representation of the current flowing through the sender. Those old needles would really get to bouncing. But you got used to it. And accuracy? A linear response sender to a non linear tank shape is just part of the game with a GM car. Many years ago one of the first things I'd do when I bought a new GM vehicle was throw 5 gallons of gas into a can in the trunk then drive the car until the tank emptied and the engine died. That way I knew where empty really was. Of course you learn some tricks like bringing a squeeze bottle to prime the carb because the old mechanical pump system took a lot of cranking to get fuel back up to the engine.

The Sunbird clusters develop problems in the r/c calibration circuit for the tach and/or speedometer. I've added resistors in the past to adjust for this and it's worked well enough. I've done a few other clusters but for some reason I never found thermally stable resistors. So I never felt the repair was good enough to charge for. At cold temps the circuit always produced high readings.

99-2000 6.5 powered diesel vans have a fuel gauge driver module (diesel pcm can't drive gauges). It provides so much damping that you almost think it isn't working. Part number is 12135192. Circuit has an NTP18N06 Power mosfet, LM2902 op-amp, MC705P6 (MC68HC05P6) processor, and sundry caps, diodes, and resistors. Refreshes on key on, then very slowly after that.

---

My 97 Van has always had a whacky fuel level sensor/gauge. Maybe I need to look into that module. I know where a 6.5TD van is in the wrecking yard.

Quote:

---

Originally Posted by Fast355

My 97 Van has always had a whacky fuel level sensor/gauge. Maybe I need to look into that module. I know where a 6.5TD van is in the wrecking yard.

---

The GM bucket pumps are notorious for wearing the 6 (pair of three) fingers off the sending connector, some of them get worn to the point you cant see the fingers any more. If you watch the voltage return on a scanner in graph mode you can see the voltage jump around quite a bit causing the crazy fuel gage readings. Craziest one I've seen was a Cadillac with digital dash, the gauge was all over the place with random readings.