My sender works but is a bit off, says I am slightly above "R" when I am empty. Is there any way to adjust these? My sense is no but just asking. I am going to ask an EE buddy about this also.
Sender Unit Adjustment Possible?
- Thread starter bavbob
- Start date
I don't think the gauge is adjustable, it is possible to change the resistance curve I suppose. Seems more likely the sender isn't sized for the tank.
They usually read too low which isn’t a bad thing.
Fill up when there's a quarter tank left, then you'll not run the tank empty.
My hope is to change the curve, wondered if anyone had tried. Sender is original to the car, moves freely and is the correct length. I have a longer sender which I will hook up outside the tank and set it at Empty and see what my gauge shows. Gonna bet it's the gauge and not the sender.
x_atlas0
Well-Known Member
Typically these are a straight up resistance vs tank % function with a float, then the gauge uses a voltage drop going through that variable resistor to determine the presented value. If you wanted to get really nuts with it, you could use a microcontroller to perform a resistance correction using digital variable resistors, but that's a good bit of work and overkill for this problem. The easiest thing would probably be to bend the float arm slightly to correct your offset. I suppose you could also 'calibrate' it using a variable pot resistor to 'zero' it out when the tank is empty by adding it in-line with the sender. That won't help though if the resistance decreases when the tank is empty, since you can't add 'negative resistance'.
No float arm to adjust. Both the E3 and E9 suffer from phantom gas...........no gas when the unit says there is. I agree with your assessment and still going to ponder such a adjustment.
I think there is little you can do, as the resistance of the detection system is giving less ohms when the tank drains.
The only thing that would work, is to replace the resistance wire with a different, say 10% lower ohm-ed wire.
But that is quite a fiddly job.
The only thing that would work, is to replace the resistance wire with a different, say 10% lower ohm-ed wire.
But that is quite a fiddly job.
Are you sure about that?
I have not measured one, but in looking at the circuit the sender is a variable resistance to ground (makes sense as you do not want to send 12V to the gas tank area...).
To move down with more current requires two coils, one always pushes it up, and the one for the sender pushes it down, therefore the needle goes up with more resistance (The E24 uses such dual coils). The other option is that the circuit has a resistor between 12V and the instrument to create a voltage divider, so an empty tank has low resistance therefore the voltage divider has lower voltage and that lower voltage is a lower reading into a high impedance gauge using minimal current. I see 12V directly to the gauge in the diagrams, so the resistor would be internal to the gauge. In any case a high reading with an empty tank suggests there is too much resistance in the sender path to ground, and that may be grounds, dirty connections, etc. It can be fixed by cleaning all that or a lower value internal resistor, which requires opening the gauge.
Last edited:
I have a bunch of 2 ohm resistors and I plan on hooking up an extra sender I have and getting current on full and empty and then with 2ohm increased resistance. This will give us a hint as to how this is setup. So sad, today's cars would have 4-5K electrical damage doing something like this.
Don would say that plan is irresistible...
I think @Arde you're right.
And because resistance to your sound arguments is futile, my current thinking now is this;
The sensor has two strands running to the bottom, over which the float glides.
The further down the float goes, the more length of resistance wire is taken into the system.
Does it work like this (?):
Adding an additional 2 ohm resistor will 'fool' the system, in thinking the float is a bit /a lot further down then it is in reality. If that is correct, then your low fuel light will come on earlier with an additional resistor compared to today's situation.
Regards,
And because resistance to your sound arguments is futile, my current thinking now is this;
The sensor has two strands running to the bottom, over which the float glides.
The further down the float goes, the more length of resistance wire is taken into the system.
Does it work like this (?):
Adding an additional 2 ohm resistor will 'fool' the system, in thinking the float is a bit /a lot further down then it is in reality. If that is correct, then your low fuel light will come on earlier with an additional resistor compared to today's situation.
Regards,
That is what I anticipate.
To get you in the ballpark on resistances, you might want to look at the VDO sender catalog. VDO sells several 90 to 0 ohm Reed switches, and a bunch of 240 to 33 ohm switches, for fuel levels, so these would be my best guess as to the most common ranges in these types of senders.
If ours are a match to either of these, we may want to see if a new VDO sender would work in our cars.
If ours are a match to either of these, we may want to see if a new VDO sender would work in our cars.
@bavbob
Fyi, here's a link to the resistance specs of a correct functional NOS system:
Post in thread 'Fuel Tank Sender - 62161352617' https://e9coupe.com/forum/threads/fuel-tank-sender-62161352617.33116/post-272910
So, empty should be, depending on the info source, bmw tech manual @ 73 ohms, or my kitchen table measurement @ 60 ohms. Now for this exercise that 13 ohm delta don't really matter, we just need to know the number of zeroes to make some rough calculations.
As the system is linear by physics, it's possible to make an educated guess (provided I get my math right):
There is about 60 to 73 ohms added over 70 liters of fuel level drop. Without going into decimals, that's about 1 ohm per 1 liter of fuel.
Thus +2 ohms will adjust your gauge by only 2 liter (about half a gallon). As You write that your gauge is slightly above 'Reserve' when empty in reality, you may want to try something like 5 ohms as a start.
And because I'm not sure if the " full" should read 60 or 73 ohms, the calculated value of 5 liter likely has an accuracy of plus minus 25%, which sounds good enough.
You might want to try and measure yours @ empty, just to know if your somewhere close to those values. If you're measuring e.g. 45 ohms when empty, then that would explain it, and show you the exact location of your problem.
Happy to assist from the couch.
Fyi, here's a link to the resistance specs of a correct functional NOS system:
Post in thread 'Fuel Tank Sender - 62161352617' https://e9coupe.com/forum/threads/fuel-tank-sender-62161352617.33116/post-272910
So, empty should be, depending on the info source, bmw tech manual @ 73 ohms, or my kitchen table measurement @ 60 ohms. Now for this exercise that 13 ohm delta don't really matter, we just need to know the number of zeroes to make some rough calculations.
As the system is linear by physics, it's possible to make an educated guess (provided I get my math right):
There is about 60 to 73 ohms added over 70 liters of fuel level drop. Without going into decimals, that's about 1 ohm per 1 liter of fuel.
Thus +2 ohms will adjust your gauge by only 2 liter (about half a gallon). As You write that your gauge is slightly above 'Reserve' when empty in reality, you may want to try something like 5 ohms as a start.
And because I'm not sure if the " full" should read 60 or 73 ohms, the calculated value of 5 liter likely has an accuracy of plus minus 25%, which sounds good enough.
You might want to try and measure yours @ empty, just to know if your somewhere close to those values. If you're measuring e.g. 45 ohms when empty, then that would explain it, and show you the exact location of your problem.
Happy to assist from the couch.
Last edited:
Ok, so now we agree that resistance gets lower as the tank fills. This means that the serial resistor will cure the empty tank offset, and the best is an adjustable resistor so you can calibrate the needle position at empty.
The system being linear is not the same as the sender resistance being linear vs. displacement by physics. There are logarithmic potentiometers, though for this use it makes more sense for the design to be linear. Unless the needle itself is not linear, but I digress, we were just trying to fix the empty offset. Whether half tank is half the needle is less important.
The system being linear is not the same as the sender resistance being linear vs. displacement by physics. There are logarithmic potentiometers, though for this use it makes more sense for the design to be linear. Unless the needle itself is not linear, but I digress, we were just trying to fix the empty offset. Whether half tank is half the needle is less important.
I think when we get down to the end of /beginning of the curve, even if it is not a 45 degree line, at the location I would be dealing with which is a short segment, one can assume linear. As I tell my kids " this isn't the space shuttle" meaning we don't have to go to 5 decimal points.
As an aside, any of you have kids who come home with an avg. grade in say, English class that is to two decimals? When I asked how a subjective class which grades in whole numbers can have such an avg and it's just stupid, I am looked upon as a moron. When your kid misses an A by .08 points, what???
As an aside, any of you have kids who come home with an avg. grade in say, English class that is to two decimals? When I asked how a subjective class which grades in whole numbers can have such an avg and it's just stupid, I am looked upon as a moron. When your kid misses an A by .08 points, what???
Maybe I confused the issue. The basic requirement for such an instrument may be relaxed from being linear to being "continuous and monotonic". That is, as the input changes the output should never jump nor reverse direction. I believe many fuel gauges are deliberately non-linear, where the needle moves little when the tank is on the half full range, and is very granular and fast as the tank empties.
You could temporarily stick a variable resistor into the gauge line and adjust to zero out your gauge at empty for your choice of reserve fuel level . Then measure the pot and substitute a fix resistor pack that’s close enough
So, rummaging around in the VDO stuff - this VDO catalog provides a cross reference on page 97 to the 2002 part. The electrical specifications for that part are on page 92. Resistance value when the tank is full should be 3.2 ohms; empty 80.6. Physical dimensions are on page 88.
This sender in the current parts catalog -
www.vdo-gauges.com
ranges from 0 to 90 ohms, is 250mm in depth and is designed to mount in a 54mm hole (which is the standard size for these senders). It mounts differently than the stock gauges, but might be made to work.
More importantly, the current gauge is a replacement for a VDO 224-225.
www.summitracing.com
That gauge appears very similar to ours, and Summit Racing indicates that these can be ordered. Check me on this, but if I were in need of a gauge, I might gamble $65 and give this a try.
This sender in the current parts catalog -
Reed Switch Fuel Sender 250 mm 90-0 Ohm | Reed Switch | Fuel Level Senders & Switches | Senders and Sensors | VDO Instruments and Accessories
Reed Switch Fuel Sender 250 mm 90-0 Ohm. Replaces fuel sender 224-225
www.vdo-gauges.com
ranges from 0 to 90 ohms, is 250mm in depth and is designed to mount in a 54mm hole (which is the standard size for these senders). It mounts differently than the stock gauges, but might be made to work.
More importantly, the current gauge is a replacement for a VDO 224-225.
VDO Gauges 224-225 VDO Tube Type Fuel Level Senders | Summit Racing
Free Shipping - VDO Tube Type Fuel Level Senders with qualifying orders of $109. Shop Sensors and Sending Units at Summit Racing.
That gauge appears very similar to ours, and Summit Racing indicates that these can be ordered. Check me on this, but if I were in need of a gauge, I might gamble $65 and give this a try.