WAY WAY off topic but I need some help..

[orange_glow_fan]

Electronic design is not my strong suit so I thought I ask if someone here could offer some advice.. I'm not even sure if this is possible.

 I need to make a fuel tank sender and it's gauge compatible. The gauge is expecting to see a sender (basically a variable resistor) that has a value 0 to 90 ohms. The current sender in the tank has a value of 0 to 30 ohms. The original gauge, changed long ago is NLA and while an appropriate sender is available, changing it is going to be an issue.

 I was hoping for a simple circuit that would accomplish this. It would have to be slightly adjustable and not affected by the supply voltage that vary by several volts depending on the battery load and charge condition.

Thanks,

\Kerry

[John Rehwinkel]

> Electronic design is not my strong suit so I thought I ask if someone here could offer some advice.. I'm not even sure if this is possible.
>
> I need to make a fuel tank sender and it's gauge compatible. The gauge is expecting to see a sender (basically a variable resistor) that has a value 0 to 90 ohms. The current sender in the tank has a value of 0 to 30 ohms. The original gauge, changed long ago is NLA and while an appropriate sender is available, changing it is going to be an issue.
>
> I was hoping for a simple circuit that would accomplish this. It would have to be slightly adjustable and not affected by the supply voltage that vary by several volts depending on the battery load and charge condition.

There are a few ways to approach this. I'll address them in order of increasing complexity.

The easiest approach is to make the existing gauge look like the gauge the sender expects. To do this, parallel it with a resistor. In effect, you want a resistor that's half the resistance of the existing gauge, so it in parallel with the existing gauge, will be 1/3 the resistance. Then the (approximately 3x) current from the sender will be divided unevenly, with about 1/3 the current (equivalent to a sender with 3x the resistance) going through the gauge. Making part of this resistance adjustable will give you the adjustment you're after. Note that this is the approach I would use if it were mine. Note that there might be some odd effects on linearity, but fuel gauges aren't famous for linearity anyway (I had a VW where full to half barely moved the pointer, but it had lots of resolution near empty, where it mattered more).

The next most complicated approach is a lashup known as a "current mirror". It uses transistors to make the current through one transistor control the current through another (or several more). By playing with transistor betas and paralleling transistors, you can make it assymetrical, so the transistor driving the gauge tracks the one driven by the sender proportionally.

After that, you get to relay type circuits, where the sender operates something like an optocoupler (ideally one of the linearized kind), which then drives an amplifier that in turn drives the gauge. This can also be done mechanically, with the sender driving a galvanometer hooked to a potentiometer that acts as a surrogate sender, with the desired resistance range.

After that, you get to microcontroller-based arrangements, where the sender drives an analogue-to-digital converter, and then the microprocessor uses a lookup table to drive a digital-to-analogue converter which then drives a transistor controlling the gauge (PWM works just fine for automotive gauges, as they don't respond quickly).

Once you have a microcontroller and a lookup table, you could also drive a nixie display or a bargraph indicator if it'll fit (just to offer something on-topic).

A side note - some cars (like my old Volvo) actually have a voltage regulator in the dash so the gauges can run on a more stable voltage supply (I think it was 8 volts).

- John

[Joe Croft]

John,

In your first solution, I think you have your circuit backwards. If I read the issue correctly, the new sender has a resistance of 0 - 30 ohms and what the gauge expects is 0 - 90 ohms. I will be honest though, I don't have a clue of how a gas gauge and sender work.

-joe

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/D0DBC2B4-F25D-414E-A652-9D2B52D1C688%40mac.com.

For more options, visit https://groups.google.com/groups/opt_out.

[John Rehwinkel]

>>> I need to make a fuel tank sender and it's gauge compatible. The gauge is expecting to see a sender (basically a variable resistor) that has a value 0 to 90 ohms. The current sender in the tank has a value of 0 to 30 ohms. The original gauge, changed long ago is NLA and while an appropriate sender is available, changing it is going to be an issue.

>> The easiest approach is to make the existing gauge look like the gauge the sender expects. To do this, parallel it with a resistor. In effect, you want a resistor that's half the resistance of the existing gauge, so it in parallel with the existing gauge, will be 1/3 the resistance. Then the (approximately 3x) current from the sender will be divided unevenly, with about 1/3 the current (equivalent to a sender with 3x the resistance) going through the gauge. Making part of this resistance adjustable will give you the adjustment you're after. Note that this is the approach I would use if it were mine. Note that there might be some odd effects on linearity, but fuel gauges aren't famous for linearity anyway (I had a VW where full to half barely moved the pointer, but it had lots of resolution near empty, where it mattered more).

> In your first solution, I think you have your circuit backwards. If I read the issue correctly, the new sender has a resistance of 0 - 30 ohms and what the gauge expects is 0 - 90 ohms. I will be honest though, I don't have a clue of how a gas gauge and sender work.

I could be wrong, but I don't think I am in this case. The way I look at it is that the signal is a current, driven through the meter and the sender from the vehicle's battery.

The meter is basically an ammeter - the amount of current flowing through it determines the position of the pointer (many gauges do this by running the current through a resistance element that acts as a heater wrapped around a bimetallic strip attached to the pointer).

So what the sender is doing is adjusting the amount of current flowing through itself and the meter, and the meter is measuring this current. The current is the vehicle voltage divided by the sum of the meter resistance and the (varying) sender resistance. The maximum current flows when the sender is at minimum resistance, and the minimum current flows when the sender is at maximum resistance. The old gauge (for the 90Ω sender is expecting a larger range of current, as its minimum is less), so you need to find a way to have less current flow through it somehow - the 30Ω sender has less maximum resistance, and therefore a smaller range of currents (its minimum current is larger). You just need to translate this larger minimum current into the smaller minimum current your meter expects, or you'll only get part of the range. My approach was to divert the "extra" current around the meter, so the meter would see the larger range of currents that it expects.

It's not a cut and dried solution, however, for three reasons:

1) the meter resistances could be different, causing a different range of currents to flow.

2) there are no rules on whether maximum current means "full" or "empty", so your meter could read backwards.

3) since the current/resistance relation is reciprocal, it's non-linear, and some senders and some meters are made so the change in resistance or reaction to current is also nonlinear (many senders use a variable pitch resistance winding to accomplish this).

As you can see, there are rather a lot of unknowns (in particular, the resistance of the meter for the 90Ω sender and the resistance of the meter for the 30Ω sender).

- John

[J Forbes]

interesting...I've had to do the opposite, when I put a later tank in an older truck. Sounds like you're using a newer gauge in an older GM vehicle. I would find a sending unit from a late 60s vehicle, and swap the resistor onto the original sending unit. It's not very difficult.

Jim

[orange_glow_fan]

Hi, Ok I tired your method and it worked. I ended up needing around 24 ohms. The only concern I have is that the resistor runs a bit warm and it's a 20 watt. (Actually two 50 ohm, 10 watt resistors in parallel) I measured the voltage drop across them and got 9.6 volts. (taken with the sending unit grounded to simulate an empty tank.) I should have measured the resistance of the gauge...    I'm concerned about the damaging the sending unit.. should I be?

[John Rehwinkel]

> Hi, Ok I tired your method and it worked. I ended up needing around 24 ohms. The only concern I have is that the resistor runs a bit warm and it's a 20 watt. (Actually two 50 ohm, 10 watt resistors in parallel) I measured the voltage drop across them and got 9.6 volts. (taken with the sending unit grounded to simulate an empty tank.)

Does that mean they're dragging the battery down to 9.6 volts, or is there something else in the circuit that could account for the remaining 2.4 volts or so? That is a scary amount of power dissipation - automotive stuff is famously low-impedance and power hungry (a blower moter can often eat 30 amps), but that seems a bit extreme for a dash gauge. Maybe measure your sender to see if it's really 0-30 ohms or perhaps it doesn't go all the way down to zero?

> I should have measured the resistance of the gauge...

I'm guessing it's about twice your resistance, which works out to 48 ohms or so.

> I'm concerned about the damaging the sending unit.. should I be?

It could be the lower-resistance sending unit you have is intended for use with a higher-resistange gauge. If so, it'll be carrying more current than it's designed for. Unfortunately, without knowing the gauge resistance it's supposed to work with, it's hard to tell. If you want to be really safe, you'll have to go for a more complicated solution that measures the sender resistance and then drives the gauge directly. Alternatively, you could power the gauge with a lower voltage, which would also lower the current and allow it to work without the parallel resistors. This would have an unknown effect on the linearity, and there's still an outside chance the sender would want even lower current.

- John

[Charles MacDonald]

On 13-11-19 12:46 PM, J Forbes wrote:
> Sounds like you're using a newer gauge in an older GM
> vehicle. I would find a sending unit from a late 60s vehicle, and swap
> the resistor onto the original sending unit. It's not very difficult.
>

Only trick is the original poster has rejected this approach - Possibly
as in many vehicles the only way to get at the sender unit is to drop
the gas tank, which is slightly difficult to do with the car on a jack.

--
Charles MacDonald Stittsville Ontario
cm...@zeusprune.ca Just Beyond the Fringe
http://Charles.MacDonald.org/tubes
No Microsoft Products were used in sending this e-mail.

[J Forbes]

The original gauge, for use with the 30 ohm sender (from a 57 Chevy) has a resistance from sender wire to ground of 200 ohms, and from sender to power of 50 ohms.

[John Rehwinkel]

> The original gauge, for use with the 30 ohm sender (from a 57 Chevy) has a resistance from sender wire to ground of 200 ohms, and from sender to power of 50 ohms.

Oho! Apparently these gauges are a little more involved than the 2-wire kind I'm used to:

http://auto.howstuffworks.com/fuel-gauge1.htm

I pulled a Chevy wiring diagram, and sure enough, the gauge has a connection to +V (tan), a connection to ground (direct), and a connection to the sender (brown). I'm unclear on how it works internally, so trying to fake it up with a parallel resistor like with a 2-wire gauge may well not work as expected. Sorry about that. Anybody know what's in the 3-wire gauges?

- John

[J Forbes]

Two electromagnets.

[orange_glow_fan]

I should have supplied more information with my original message but I was trying to brief and not bore you guys hehe..

This gauge problem is in a 1965 Ultravan (motorhome) which is a Corvair powered unit. As I said, it has the original sending unit but the gas gauge has been replaced. I used a 90 ohm resistor to verify that it needs to see a 90 ohm sending unit to show 'full' on the gauge.

 The tank is a 30 gallon unit that would be a huge PITB to remove. Also, since it's unlikely that the replacement sender will fit, it's going to require some modification and the thought of drilling/cutting into the tank (gas explodes hehe) after draining seems a bad idea. unless the tank is filled with water vented etc...

There is some confusion as to what resistance range the original sending unit had. Some report a 0 - 45 ohm version, some a 0 - 30 ohm version. Currently we measured the level of gas in the tank and got 1/2 a tank. The sending unit measures 14.9 ohms to ground hence I believe that it is a 0 - 30 ohm unit.

Here is a not so brief description of the gas gauge/sender operation :

The fuel gauge needle is mounted between two magnetic coils. One coil is on the "F" side of the gauge

and the other is on the "E" side of the gauge. Both coils are wired to the ignition circuit and receive

battery voltage whenever the ignition switch is on. The magnetic coil that pulls the needle toward F

(Let's call this the F coil) is grounded at the instrument panel and applies a constant magnetic force to

the needle. The magnetic coil that pulls the needle toward E (Let's call this the E coil) has more

windings for greater magnetic strength and is grounded through the variable resistor in the sender

attached to the fuel tank float.

When the fuel tank is full the resistance in the sender is high (30 or 90 ohms) which reduces electrical

current and magnetic strength in the fuel gauge E coil allowing the F coil to pull the gauge needle to 'F'

When the fuel tank is nearly empty and the float is near the bottom of the tank the resistance in the

sender is at 0 ohms. The fuel gauge E coil receives full current, has enough magnetic strength to

overcome the strength of the F coil and pulls the gauge needle to 'E'

 I hope that explanation helps clear things up a bit..

I am going to make some new measurements today, hopefully one of which will be the resistance of the gauge and voltages for reference. The gauge is also a bit of a pain to get to.

Thanks for your input so far!

Kerry

[J Forbes]

The only part of the sender that would need to be replaced, is the resistor. They are interchangeable, as long as you get a replacement that is not too many years away from the original. I've done the opposite a few times, swapping an early resistor onto a newer sender. The resistor is part of the stamped housing, it is held in place by a few bent tabs. The strap needs to be soldered to the old one. There really isn't much to it.

<http://www.summitracing.com/parts/opg-ch20871/overview/make/chevrolet/model/chevelle>

If you can figure out an electronic solution, you would save getting dirt in your eyes. Or just use the lower third of the gauge, it should work fine at the lower end of the scale as-is, it just won't have fine resolution.

[Terry S]

You could consider taking some turns off the gauge winding, reducing the field strength.

 

Terry