Poked around in my IC's I did not have any lm117, but I do have a few um78 auto rated ones but these are all fixed voltage.

Wuffa. If you do try, becareful and monitor the heat, you'll amlost certainly need a heat sink of some form even abit of PCB. If scope was correct 850mw is going to be a bit above the power(heat) dissipation rating w/o a heat sink it's at its limits at 833mw.

 

OK, thanks for the warnings. I found a LM350T that I think should work for this. I'll take the cluster apart tomorrow and see what it looks like in there.

 

I couldn't find a voltage regulator in mine either.
And it doesn't look like yours.
Here's what I found:

 

Lets try sending this again and see if I remeber what I typed the first time, computer eeds capacitors replaced

Can you get clear pics of the front and back of that PCB?
Did you do a voltage check on A-C(+/-)?

I think maybe they changed the way these work, from the diode thread guys. His was likely only using the 7v, and these appear to use the "7v" and battery voltage.

So what happens is in the dual voltage ones with a resistor based voltage divider.. It doesn't need a properly regulated voltage supply just uses the ratio of divided and non divided voltages. When the B+ drops the "7v" drops accordingly, eg a fixed ratio. There is little to no tension on my spring loaded "horseshoe" coils, sam with the wound coils. Infact I know my side to side fuel guage stays where ever it was when the power is turned off.

After I get a look at the circuit board I'll have a better idea of whats going on. It'll be a few days before it's warm enough to pull my dash apart. It's a bit UV brittle and I have to work outside sans garage also.

 

Trying to work this out in my head.. I was reading the other circuit description again it has this part..
I tryed flexing that bit of metal(right side of my pic) and it wasn't budging. I can try applying power and watching it, but I don't think my DMM reads but maybe 2x per second so it's not going to show the opens.

 

being wrong and corrected is good, being wrong and figuring it out yourself *priceless*

Applying power from F to GND does produce a heat reaction bend in my non sr5 guage. My little test 9v battery wasn't enough power to get the switch to open when running between 12-7v or 7v-gnd.

 

Here are what I have now. Almost clear anyway.
As you may be able to see, there's some darkening from heat around some of the resistor connections.
The resistors measure:
Red Blk Brn - 92 - should be 200
Brn Blue Brn - 150 - should be 160
Grn Blue Blk - 58 - should be 56
So might replacing the big resistor make a difference?

 

This is interesting. But not having taken mine apart I can only contribute a little.

Historically, gauges like this have been "hot-wire ammeters" http://en.wikipedia.org/wiki/Ammeter#Hot-wire_ammeters. Very rugged, and intrinsically damped (slow-moving). Co_94_PU's looks like what I would expect for a hot-wire, but Wuffa's has far too many turns of magnet wire (so it looks like a moving-magnet meter).

I assume Wuffa measured the resistors with the board removed from the meter (like picture 2) (otherwise, the resistance of the coil is in parallel with the resistor). I read the color-code the same way. So I would definitely replace the resistor. These are only 10% resistors, but 92 is too low for the label. It looks to my untrained eye like a 1watt resistor, so I would either get a 1watt or, better yet, bring it along and match it by size. These look like metal-film resistors, but any 200 ohm resistor from wire wound to composition should work fine.

Thanks for sticking with this, and reporting what you see.

 

I think the big resistor is 3w and the littler ones 1w.
I ordered a 3w and 5w replacement. Since it looked like it was getting a little hot I thought I'd go a little bigger since there's a lot of room.
I'll let you know how it goes in a couple of days.

Replaced the 200 ohm resister with new one and the gauge still doesn't work right.
The measurements at the test points are a little better but not close to spec.

 

If swapping the resistors and cleaning up contacts doesn't get it in spec, maybe comparing the winding resistances thats not going to be an easy DIY fix but it's more info.

So I had my dash apart the other day to adjust the TACH.

The mid 90's "side to side" guages, have individual boards for each sensor. It's the same design as Wuffa's only smaller and more compact.

The Oil pressure guage is the hotwire type, the fuel guage has the large resistor as pictured above, the Coolant and Voltage are marked "temp/voltage" on the right-hand side and are interchangable by swapping the face plates. You can also swap the fuel guage for one of these but you'll need to reconfigure the resistors, it's the same circuit board.

For my arch type fuel guage pictured above. To work with the external regulator you need to desolder the wire at the 7V connector, then slide a piece of heatshrink over the arm this wire wraps around so it no longer contacts the screw(bottom-right side of picture). That will disable the mechanical hot wire voltage regulator.

On a side not, holy crap that thing wastes a lot of power. 50% duty cycle at ~14v and 1 Ohm. 7A 98W