If you have a working car with a dual sender guage, I would do the following:-

Disconnect where you have the red and blue circles. You might be able to do that at the back of the car, where the wire goes to the two-wire sender.

Turn on the ignition.

One side ofthe break will have a voltage on it. That is the one to look at for now.

Measure the voltage to ground. Write it down.

Measure the short circuit current to ground. Write it down.

Put everything back together.

The voltage devided by the current is the effective resistance that the pair of guages is supplied through. For one guage there should be half the resistance.

You can adjust the zenner voltage to give any ratio of resistance.

However, there is no guarantee that the guage will match the sensor at half tank even if it matches at full and empty.

 

Ill try that tomorrow, I was going to measure the gauge in the cluster itself tomorrow and reply, but I might as well update with what I know for now.

Although it isn't much, I took both of the senders out of my daily driver yesterday and measured them at full and empty. They're both roughly 4hm: to 60hm:. I think theyre close enough to the same for my purposes.
The actual lowest measurements I got from each of them were -
Primary - 4.0hmF) 59.0hmE)
Secondary - 3.4hmF) 65.7hmE)

Like I said, Ill try what you suggested tomorrow and see how it goes and post back.

Thanks

 

Okay, well I did a considerable amount of extra work till I went back and looked at your post again. I'll chalk it up to a brain fart and not thinking hard enough

Well, this is what I found-
V- 7.12 off
8.10 idling
A- 88mA off
99mA idling
Resistance through the gauge- ~81hm:
Since I'll only be using one sender, it should be 40hm:

As far as the senders go, they're linear, so I'd think they should match.

This is what I'm thinking, lemme know if I'm on the right track.
Reverting back to your schematic-

-Bottom R1 (R2) should be a 40hm: resistor.
-The 3.6v zener should stay as far as we know because I'm cutting the voltage in one fourth the supply voltage from a straight +12-14v on my project car, not cutting the 7-8v that I saw *at the sender* in my running car.
-The transistor should be a TO-220 transistor in order to be able to add a heat sink. A TIP31? I actually have one of those so that would be good.
-What should I be looking at for R1?

Thanks!

 

Well done doing all the hard work. I would agree that the sender pair see 80 hm: and that one should see 40hm:

As the lower resistor is 40hm: and there will only be 3V available, the current will be about 75mA

There will be up to about 9V across the transistor so about 0.7W of heat in the transistor, assuming low guage resistance. If the guage resistance is high, there will be less heat.

The TIP31 has a dissipation of 62.5 °C/W on its own, so that only gives a rise of 44 °C so you would probably get away without a heat sink. I would bolt it to a little bit of metal on the circuit board, but more to stop vibration breaking the legs, which is likely as they are in a straight line.

The gain of a TIP31 is poor, but better at low currents. http://www.onsemi.com/pub/Collateral/TIP31A-D.PDF page 5 says that 2 mA is enough base current for 0.3A collector current, so aim for that.

That gives the upper resistor as (12-3.6)/0.002 = 4200hm: but anywhere around there would be fine.

I hope it all works.

 

Much appreciated! If everything is on time I think I should be getting my gauges today or tomorrow. In the mean time I'll find the parts and put together the circuit. Since I have a running car and know that with both senders together it works as it should with the stock gauge, I plan to just have a secondary circuit to plug into the primary sender to test the new gauge. If the stock gauge and my new autometer one read the same, then we're golden.

FWIW, I'll post back with results when I get them. Thanks.

 

Looks like there MAY be a little problem, or I just don't see what should be changed.

Remembering-
Sender: 4hm: Full - 60hm: Empty

With the circuit built, Mock-up 12v battery hooked up, and hooked up the resistances look like this-
What the sender is turned into, between GND and the spade connector that plugs into the gauge, they read: 129.2Khm: Full - 288.7Khm: Empty
Current is 4mA at those same points.
Battery reads 12.25v
Voltage from GND to the gauge through the sender and circuit reads 0.29v with the sender at empty, and 0.13v when full.

This causes the gauge to want to read at 1/4 tank at Full, and well past the E when Empty.
http://img.photobucket.com/albums/v2...f/IMG_3144.jpg
http://img.photobucket.com/albums/v2...f/IMG_3145.jpg

Next time i put this circuit together I'll be smarter and do it on my breadboard. Then I can more easily change out the resistance values. FWIW, I used a 39hm: to replicate the original gauge, where to be entirely accurate, should be about 41hm:, and instead of 4.2K in R1, the cloesst single resistor was a 3.9Khm:. Lowering that resistance by 300hm: only changed the base mA by about .3 IIRC, I have it all written down somewhere. Could that change the end result that much though? The resistance is in the K range not hundreds, but with reads due to the mA to drive the gauge being proper if I understand correctly. Keeping with that, I would need to adjust the resistance measured through the circuit must be lowered as a whole a total of ~50k to 240khm: and spread out by ~47khm: to reach down to ~33khm:, or something equivalent, I presume, assuming the battery voltage at only 12v didn't effect it too much. I'm wishing I didn't solder it all together. I bought some extra's online along with some other bits I need for a related but unrelated project, so when they show up I'll do the breadboard. Till then, well, that's where it stands. I'll try to figure some of it out tomorrow too. It's too late right now for me to be thinking logically about it

 

The resistance you quoted can't be correct. You didn't say how you measured it, but I suspect you used a multimeter set to read ohms on a the circuit that was powered by the battery.

That never works. Resistance is the ratio of voltage to current. It can only be calculated if you know the current and the voltage. The multimeter set to read ohms puts out a fixed current, measures the voltage and does the maths. If the circuit is powered, there will be other currents and the reading is meaningless.

129.2Khm: with 4 mA flowing would be over 500V.

I think that the voltage to ground at the guage points to where the problem is. My circuit is designed to give 3V at the sender. It relies on getting enough current through the guage, to power the sensor. The design current is up to about 70mA (that is 3V / 44 hm: ) and it seems that you are only getting 4 mA. 4 mA is also consistant with the 0.29 V and the 0.12 V that you saw.

With voltages like that, there is no way that you could get more current though the guage. You have just about shorted the wire to ground, and the guage reads less than half. I think that there is something wrong in the guage or the way that you have wired it. Have you grounded the guage? I guess so because the voltage seems to read correctly in the photos. You could check the guage with a 240 hm: resistor or a 33 hm: resistor and see if it reads empty and full. The voltages on the resistors would also be useful to check that the circuit is correct.

To check the circuit, put the guage to one side, and replace it in the circuit with an ammeter. It should read about 30mA empty and 70mA full.

 

Point taken. I wasn't quite thinking about how I was getting the numbers, and how they would be altered. Regardless, as explained, it comes back to me as a duh. I did check the circuit for any problems after I soldered it up and all seemed to be well. When I get home I'll check my numbers again. Particularly the mA.

 

I tried it again here at home and found something interesting, though kinda useless. The back of the gauge wants a Batt Pos 12v input, "Fuel" (from sender, which eventually see's the ground), as well as a direct ground to measure the battery voltage on another gauge. Without the Batt Neg connected to the ground on the back of the gauge, and when the sender says it's empty, it reads exactly Full. Hooking up the back of the gauge properly drops the needle at empty back to below E.

Anyway just for a recap for myself as much as anything else-
Sender-
Pin 1:Batt Neg
Pin 2:SNDR on board
Board-
SNDR:Pin 2 on Sender
GAUGE:Fuel on Gauge
GND:Batt Neg
BATT+:Batt Pos
Gauge-
Fuel:GAUGE on Board
+12v:Batt Pos
GND:Batt Neg

Having everything hooked up this way, and removing the gauge entirely, I measured the mA between the "Board" GAUGE terminal and Batt Pos and got this-
Sender @ Empty: 23mA / 11.94v
Sender @ Full: 51mA / 12.10v
A little bit off..

The way I screwed up the resistances and other stuff last night, I accidentally measured between the GAUGE terminal on the board and Batt Neg, which ended up just measuring through the whole thing back to the battery, rather than anything the gauge would have seen.

 

It looks like there isn't enough voltage on the base of the transistor.

With the guage replaced with an ammeter, can you measure the base and emitter voltages on the transistor with the sender at full, empty, and disconnected?

 

Hmm, unfortunately I only have one good multimeter (fluke 88v) around here that I can use for a good ammeter. I'd have to go searching for one to measure voltages at the base and emitter. I think I have a cheapo one in the garage. I have somewhere I have to be in less than an hour and I have yet to get ready so I'll measure that stuff when I get home in a few hours.

Basically, you want it set up the same way as I was measuring the mA, but while the ammeter is connected, measure voltage from the base and emitter to Batt Neg? Gotcha.

Just for fun, I had been working on drawing this in paint-

 

Voltages w/ ammeter hooked up in place of gauge-
Empty-
Emitter: 2.21v
Base: 2.78v

hah, I'm not going to say I just didn't get a chance to test it. I managed to cook it, conveniently arcing collector to base on accident. No more voltage out of it at all and the zener doesn't test =x. Awesome. I'll pick up a couple new pieces Monday since the place I got these is closed Sundays and retest for empty, and get readings for a full tank and disconnected sender. This time if I remember, with the breadboard. Boy, do I feel like a dope. It happens. At least there's one reading.

The battery is at 12.2v still.

 

Hmm, done up on my breadboard again with basically the same parts I got slightly higher voltages, so I'll try it hooked up to the gauge to see if it makes enough of a difference.

I also have a 3.9v zener, and a wide assortment of resistors to play around with if I have to.

These are the results tonight -
EMPTY-
Emitter: 2.45v
Base: 3.05v
FULL-
Emitter: 2.41v
Base: 3.02v
SENDER D/C'ed-
Emitter: 2.75v
Base: 3.08v

I'll edit this post soon to update what the gauges look like.
Edit-
They look pretty much the same as before. Full is just above the 1/4 tank mark, and Empty is considerably below E.

 

Managing to break the 3.6v zener in half, :P, I put in my 3.9 and slowly went lower on the R1 resistance while constantly measuring the flow across the zener. I found that I can get it to read up to about 3/4 of a tank, but as I went lower and lower, it made less of a difference. I think I need a higher voltage zener so I can keep the resistor a slightly higher value, once it was down to about 100hm:, R1 was getting really, really hot =x. I didn't leave it that way for long, just to measure the current on the zener and what the tank capacity wanted to read. Just over 3/4 tank with that, but using a 560hm: it reads exactly 3/4 with only 23mA across the zener.

This causes me to think what I just said above, what about using a higher voltage zener so I can keep the resistor high keeping current and temperature in check?

 

FWIW, these are the measurements just as above, but that I got with the 560hm: resistor and 3.9v zener, and the actual gauge connected-

EMPTY-
(reads a needle width below E, which is fine and will probably come up a little higher as it's more finely tuned)-
Emitter: 3.22v
Base: 3.84v
FULL-
(reads as exactly 3/4 tank)
Emitter: 3.18v
Base: 3.83v
SENDER D/C'ed-
Emitter: 3.54v
Base: 3.84v

Also, ammeter hooked up as the gauge between the "gauge" terminal on the circuit and BATT NEG, the sender reading EMPTY pulls 14mA. Reading FULL it's 8mA. Disconnected all together, clip fell off but to report it anyway, is 20mA.

Any help to figure what higher voltage zener and R1 it should be would be helpful, if that's what would turn this around to be just right.

Thanks for the help so far, I certainly wouldn't have gotten to this point without it. =X