Fuel level sensor resistance?

[Bernd Felsche]

My 1990 Golf GTI has a float-style fuel sender.

The fuel gauge appears to be simply an ammeter as the sender unit is
a variable resistance (of, as yet, unknown range) and the instrument
panel simply feeds 10V, via that fuel gauge, to the fuel sender. The
internals of the meter are unknown ... so I don't know if there's a
large resistor in series, inside the meter without skinning my
knuckles.

I cannot fathom why the resistance ranges of fuel senders aren't
published in workshop manuals. It'd make fuel gauges easy to test
properly, wouldn't it?

From what I've read from unofficial sources, minimum resistance is
about 50 Ohms when empty and about 240 Ohms when full (other
manufacturers like it upside-down). Does that look about right?

Time to open my box of pots to see how the fuel gauge responds over
the range of a potentiometer and the actual current/voltage.

The reason for wanting to know these things to look at the
feasibility of faking the fuel level indication using the stock
meter. I can measure the resistance of the sender using one network
node and then send a scaled value to the network node that will
proxy the value to the meter.

But faking a resistance value is "complicated"/expensive. It needs a
precision current-sink/-source if the application uses an ammeter.

OTOH, if the meter is sufficiently slow, one can easily sink enough
current for the full scale of the meter, varying the amount by
pulse-width modulation in the kHz range. OR; one can isolate the
meter from the PCB backplane and drive both ends which might be
needed if it really needs an estimated 200 mA (10V into 50 Ohms);
something that the microcontroller won't be able to do
internally.

Similarly, coolant and oil temperature sensors also need to be
proxied. Their temperature coefficients (NTC) are documented so
there's less guessing.

Oil temperature is displayed is not by analogue meter; it's one of
the multi-function-accessory (MFA) digital functions - so there's a
can of aliasing worms awaiting the opening, should I "drive" that
directly with PWM. A voltage is probably what it measures and a DAC
can do that. But I need to know the scale.
--
/"\ Bernd Felsche - Innovative Reckoning, Perth, Western Australia
\ / ASCII ribbon campaign | For every complex problem there is an
X against HTML mail | answer that is clear, simple, and wrong.
/ \ and postings | --HL Mencken

[Rob]

On 30/11/2011 3:22 AM, Bernd Felsche wrote:
> My 1990 Golf GTI has a float-style fuel sender.
>
> The fuel gauge appears to be simply an ammeter as the sender unit is
> a variable resistance (of, as yet, unknown range) and the instrument
> panel simply feeds 10V, via that fuel gauge, to the fuel sender. The
> internals of the meter are unknown ... so I don't know if there's a
> large resistor in series, inside the meter without skinning my
> knuckles.
>
> I cannot fathom why the resistance ranges of fuel senders aren't
> published in workshop manuals. It'd make fuel gauges easy to test
> properly, wouldn't it?
>
> From what I've read from unofficial sources, minimum resistance is
> about 50 Ohms when empty and about 240 Ohms when full (other
> manufacturers like it upside-down). Does that look about right?
>
> Time to open my box of pots to see how the fuel gauge responds over
> the range of a potentiometer and the actual current/voltage.
>

The testers are variable pots which are connected to the gauge.
Manufactures like VDO supply a chart full mid and empty. You could go
from here.

BTW are you sure there are no values for the fuel sender unit?

full - 263 ohms
half - 167
empty - 57.

[Trevor Wilson]

**FWIW: My VL COmmodore manual states that an empty tank should read 294
Ohms +/- 7 Ohms. A full tank should read 34 Ohms +/- 5 Ohms.

--
Trevor Wilson
www.rageaudio.com.au

[Bernd Felsche]

Rob <mesa...@gmail.com> wrote:
>On 30/11/2011 3:22 AM, Bernd Felsche wrote:
>> My 1990 Golf GTI has a float-style fuel sender.
>>
>> The fuel gauge appears to be simply an ammeter as the sender unit is
>> a variable resistance (of, as yet, unknown range) and the instrument
>> panel simply feeds 10V, via that fuel gauge, to the fuel sender. The
>> internals of the meter are unknown ... so I don't know if there's a
>> large resistor in series, inside the meter without skinning my
>> knuckles.

>> I cannot fathom why the resistance ranges of fuel senders aren't
>> published in workshop manuals. It'd make fuel gauges easy to test
>> properly, wouldn't it?

>> From what I've read from unofficial sources, minimum resistance is
>> about 50 Ohms when empty and about 240 Ohms when full (other
>> manufacturers like it upside-down). Does that look about right?

>> Time to open my box of pots to see how the fuel gauge responds over
>> the range of a potentiometer and the actual current/voltage.

>The testers are variable pots which are connected to the gauge.
>Manufactures like VDO supply a chart full mid and empty. You could go
>from here.

>BTW are you sure there are no values for the fuel sender unit?

Not to be found in the workshop manuals.

>full - 263 ohms
>half - 167
>empty - 57.

Thanks. Where did you get those figures?

[Albm&ctd]

In article <sedfq8x...@innovative.iinet.net.au>,
ber...@innovative.iinet.net.au says...

> My 1990 Golf GTI has a float-style fuel sender.
>

Capacitive?

Al
--
I don't take sides.
It's more fun to insult everyone.

[Bernd Felsche]

Albm&ctd <alb_mand...@connexus.net.au> wrote:
>ber...@innovative.iinet.net.au says...
>> My 1990 Golf GTI has a float-style fuel sender.
>>
>Capacitive?

Resistance.

[Bernd Felsche]

Bernd Felsche <ber...@innovative.iinet.net.au> wrote:

>My 1990 Golf GTI has a float-style fuel sender.

>The fuel gauge appears to be simply an ammeter as the sender unit is
>a variable resistance (of, as yet, unknown range) and the instrument
>panel simply feeds 10V, via that fuel gauge, to the fuel sender. The
>internals of the meter are unknown ... so I don't know if there's a
>large resistor in series, inside the meter without skinning my
>knuckles.

[snip]

>From what I've read from unofficial sources, minimum resistance is
>about 50 Ohms when empty and about 240 Ohms when full (other
>manufacturers like it upside-down). Does that look about right?

>Time to open my box of pots to see how the fuel gauge responds over
>the range of a potentiometer and the actual current/voltage.

Got out the multimeter. Fetched my box of pots which has about 200
in it of various scales. Turned on the desk lamp and settled down
for some rifling trhough the box to find the right one.

Picked one from the top in the middle of the box. Looked at the
somewhat "smudged" embossing which said B#$%@. Turned it to
half-scaled and whacked it on the multmeter ... 347 ohms. Check
scale on meter... OK. Turn knob to "max" and it shows 0.534
kiloohms.

Polish back of pot and look at the embossing under the light: "B500"

Dangit... got a suitable one, first try. I was expecting to take at
least half an hour. Checked a handful of others ... all are 20k or
more; or log-scale.

So I'll waste 10 minutes posting here. More than just a little sad.

[Albm&ctd]

In article <30fjq8x...@innovative.iinet.net.au>,
ber...@innovative.iinet.net.au says...

"Now in '14 and in '39, war raised its ugly head,
The bombs they fell on England, and one fell on my shed,
But we fought and beat the Germans 'cos we knew just what to do:
We stuck our fingers in our ears and went ting-a-ling-a-loo."

http://bestuff.com/stuff/stick-em-in-your-ear-and-go-ting-a-ling-a-loo

[Jason James]

Linear pots are around>>plse see this site. It shows the idiosyncrasies
of dash instruments,..especially the damping needed for a fuel gauge,..and the reistance to battery volts changes. NB this vibrator type of regulator is very common up until the mid '80s so your car may use a 3 lead reg.

Jason

[Bernd Felsche]

Ermmmmm.... did you READ what I wrote?

What "vibrator type" are you talking about? The one that you're
sitting on?

It's a 1990 GTI,

READ what I write.

[Jason James]

On Sunday, December 4, 2011 11:36:41 PM UTC+11, Bernd Felsche wrote:

> Jason James > >On Thursday, December 1, 2011 4:13:07 PM UTC+11, Bernd Felsche wrote:
> >> Bernd Felsche > >>

To measure the sender resistance, its easier to pull it out of the tank. The total resistance will include [in series] the bi-metal strip heater element in the gauge. This strip is the needle in the gauge. It gives damping action to the set-up. More modern examples probably use slew-rate as a damping effect, in a simple single transistor amplifier.

>
> Ermmmmm.... did you READ what I wrote?
>
> What "vibrator type" are you talking about? The one that you're
> sitting on?

Its a very common way of producing regulated volts for instruments [10v] It works on the same principle as early genny/alternator regulators. The relay type devicehas its coil connected to battery via its own make set of contaqcts. They are ajustable. The closer the armature to pole-end gap,..the higher the reg-volts out. Your car being a 1990 may still use one of these,..a std 3 lead regulator will do the job. 1 amp or higher to give goode safety margin.

> It's a 1990 GTI,
>
> READ what I write.

yes masser :-)

[Bernd Felsche]

Jason James <5sf...@hotmail.com.au> wrote:
>Bernd Felsche wrote:
>> Jason James

>> >Bernd Felsche wrote:

>To measure the sender resistance, its easier to pull it out of the
>tank.

No it isn't. And the meter depends on the total circuit's resistance
anyway.

>The total resistance will include [in series] the bi-metal
>strip heater element in the gauge. This strip is the needle in the
>gauge. It gives damping action to the set-up. More modern examples
>probably use slew-rate as a damping effect, in a simple single
>transistor amplifier.

What will you do if the gauge isn't a heated bi-metal strip type and
an ammeter with passive electronics fitted to slow the response?

>> Ermmmmm.... did you READ what I wrote?

>> What "vibrator type" are you talking about? The one that you're
>> sitting on?

>Its a very common way of producing regulated volts for instruments
>[10v] It works on the same principle as early genny/alternator
>regulators. The relay type devicehas its coil connected to battery
>via its own make set of contaqcts. They are ajustable. The closer
>the armature to pole-end gap,..the higher the reg-volts out. Your
>car being a 1990 may still use one of these,..a std 3 lead
>regulator will do the job. 1 amp or higher to give goode safety
>margin.

No. It uses a single, 3-terminal, solid-state, low-dropout voltage
regulator. All Mk2 Golf have, since 1983. That's well known. Widely
documented.

And can you PLEASE use a Usenet client that doesn't wrap
attributions out of context?

[Jason James]

On Monday, December 5, 2011 6:26:20 PM UTC+11, Bernd Felsche wrote:

> Jason James wrote:
> >Bernd Felsche wrote:
> >> Jason James
> >> >Bernd Felsche wrote:
>
> >To measure the sender resistance, its easier to pull it out of the
> >tank.
>
> No it isn't. And the meter depends on the total circuit's resistance
> anyway.

Ahh OK,..I hope you aren't implying it's easier to use a pot to determine same..

> >The total resistance will include [in series] the bi-metal
> >strip heater element in the gauge. This strip is the needle in the
> >gauge. It gives damping action to the set-up. More modern examples
> >probably use slew-rate as a damping effect, in a simple single
> >transistor amplifier.
>
> What will you do if the gauge isn't a heated bi-metal strip type and
> an ammeter with passive electronics fitted to slow the response?

Nothing :-) You are driving the discussion. I thought you wanted to know what resistance the tank rheostat was? These additional features were by the way..

> >> Ermmmmm.... did you READ what I wrote?
>
> >> What "vibrator type" are you talking about? The one that you're
> >> sitting on?
>
> >Its a very common way of producing regulated volts for instruments
> >[10v] It works on the same principle as early genny/alternator
> >regulators. The relay type devicehas its coil connected to battery
> >via its own make set of contaqcts. They are ajustable. The closer
> >the armature to pole-end gap,..the higher the reg-volts out. Your
> >car being a 1990 may still use one of these,..a std 3 lead
> >regulator will do the job. 1 amp or higher to give goode safety
> >margin.
>
> No. It uses a single, 3-terminal, solid-state, low-dropout voltage
> regulator. All Mk2 Golf have, since 1983. That's well known. Widely
> documented.

Excellent !

> And can you PLEASE use a Usenet client that doesn't wrap
> attributions out of context?

It's google,..

[Bernd Felsche]

Jason James <5sf...@hotmail.com.au> wrote:

>On Monday, December 5, 2011 6:26:20 PM UTC+11, Bernd Felsche wrote:
>> Jason James wrote:

>> >To measure the sender resistance, its easier to pull it out of the
>> >tank.

>> No it isn't. And the meter depends on the total circuit's resistance
>> anyway.

>Ahh OK,..I hope you aren't implying it's easier to use a pot to determine same..

Why not?

The pot will tell me what the gauge thinks is the fuel level
corresponding to a resistance. I'll also have an ammeter in series
as well as a voltmeter in parallel to the pot so I get the
resistance without breaking the circuit.

All I have to do is to adjust the pot slowly and allow it to settle.
How slowly I can gauge by shorting across the terminals and opening
them again.

VDO level sensor goes 283±25 ohms when full to 36±4 ohms when
empty, according to their sender replacement catalogue.
<http://www.vdo.com/generator/www/com/en/vdo/main/hidden/downloads/replacement_parts/flc_sensors_instrumentation_en.pdf>

Wiring resistance is probably within the ± at the low end.

Connecting the pot at the sensor location will include the wiring
resistance.

>> >The total resistance will include [in series] the bi-metal
>> >strip heater element in the gauge. This strip is the needle in the
>> >gauge. It gives damping action to the set-up. More modern examples
>> >probably use slew-rate as a damping effect, in a simple single
>> >transistor amplifier.
>>
>> What will you do if the gauge isn't a heated bi-metal strip type and
>> an ammeter with passive electronics fitted to slow the response?

>Nothing :-) You are driving the discussion. I thought you wanted to
>know what resistance the tank rheostat was? These additional
>features were by the way..

Possibly down the garden path ... :-)

>> And can you PLEASE use a Usenet client that doesn't wrap
>> attributions out of context?

>It's google,..

Blame the tool.

[Jason James]

On Monday, December 5, 2011 7:37:17 PM UTC+11, Bernd Felsche wrote:
> Jason James wrote:
>
> >On Monday, December 5, 2011 6:26:20 PM UTC+11, Bernd Felsche wrote:
> >> Jason James wrote:
>
> >> >To measure the sender resistance, its easier to pull it out of the
> >> >tank.
>
> >> No it isn't. And the meter depends on the total circuit's resistance
> >> anyway.

It does. Re the pot method, you'll not be taking into account

>
> >Ahh OK,..I hope you aren't implying it's easier to use a pot to determine same..
>
> Why not?
>
> The pot will tell me what the gauge thinks is the fuel level
> corresponding to a resistance. I'll also have an ammeter in series
> as well as a voltmeter in parallel to the pot so I get the
> resistance without breaking the circuit.

A couple of things here,..tank senders have a reputation for not contacting the whole of the wire-wound resistance all the time. These intermittant contacts tends to hold the fuel level higher than actual. If this lack of good contact occurs near full or empty, it will stuff up your measurements. Also, you'll need a pot that can dissapate significant power. Awire-wound pot is indicated.

>
> All I have to do is to adjust the pot slowly and allow it to settle.
> How slowly I can gauge by shorting across the terminals and opening
> them again.
>
> VDO level sensor goes 283±25 ohms when full to 36±4 ohms when
> empty, according to their sender replacement catalogue.
> <http://www.vdo.com/generator/www/com/en/vdo/main/hidden/downloads/replacement_parts/flc_sensors_instrumentation_en.pdf>

This is all you need ie an accurate set of figures,

>
> Wiring resistance is probably within the ± at the low end.

Yep..

> Connecting the pot at the sensor location will include the wiring
> resistance.
>
> >> >The total resistance will include [in series] the bi-metal
> >> >strip heater element in the gauge. This strip is the needle in the
> >> >gauge. It gives damping action to the set-up. More modern examples
> >> >probably use slew-rate as a damping effect, in a simple single
> >> >transistor amplifier.
> >>
> >> What will you do if the gauge isn't a heated bi-metal strip type and
> >> an ammeter with passive electronics fitted to slow the response?

Like with a largish resistor-capacitor network cross the line-in? Yeh.

>
> >Nothing :-) You are driving the discussion. I thought you wanted to
> >know what resistance the tank rheostat was? These additional
> >features were by the way..
>
> Possibly down the garden path ... :-)
>
> >> And can you PLEASE use a Usenet client that doesn't wrap
> >> attributions out of context?
>
> >It's google,..
>
> Blame the tool.

Google will not post any statements with an @ in it. With some additional care [like this post] I can alleviate the situartion.

[Jason James]

Error in previous post [from me] Plse see last line this text >>>>>

On Tuesday, December 6, 2011 1:37:30 AM UTC+11, Jason James wrote:
> On Monday, December 5, 2011 7:37:17 PM UTC+11, Bernd Felsche wrote:
> > Jason James wrote:
> >
> > >On Monday, December 5, 2011 6:26:20 PM UTC+11, Bernd Felsche wrote:
> > >> Jason James wrote:
> >
> > >> >To measure the sender resistance, its easier to pull it out of the
> > >> >tank.
> >
> > >> No it isn't. And the meter depends on the total circuit's resistance
> > >> anyway.

I started to say something about using a pot, but made statement in the next chunk of text. Corrected line of text is below incorrect last line of text >>>>

> It does. Re the pot method, you'll not be taking into account

It does.

Thankyou,..Jason.

[jetblac...@hotmail.com]

Hi Bernd
Trying to reach you about a VW project, be great to hear from you.
Tried emailing but no dice,
Cheers John