New to the forum and have seen this subject discussed already but unfortunately some time ago so that some images are no longer available.

Anyway, the title is self explanatory but the problem began just recently, one day apparently working fine and then suddenly registering full, even with the sender disconnected.

I've attached a chart of voltage and resistance readings taken from the cluster under various conditions which I hope may help to throw some light on the cause of the problem.

Any help or suggestions greatly appreciated and feel free to query my amateur terminology.

Thanks.

Caddy Fuel Gauge Analysis 250629.pdf

Hi, welcome.

Sorry I don't know the (over complicated) VW system and your attachment doesn't download for me so total generalisations which you might already know.

If you have fitted any new (modern made) parts recent then check the parts and their installation as many modern made parts can be very poor quality and if the part was replaced was copied from what was fitted and the way it was fitted one or both could be wrong or not fully working.

ETA: live suppliers are check but also checking the the earths are often about

Wires and connections -

"If the gauge increases in resistance from Empty to Full, such as 0 to 90, 0 to 30, or 16 to 158, and assuming that the gauge is properly matched to the sender then read on. The most common cause of this type of gauge to read over full is typically due to no ground at the sender, an open circuit, or break in the wire going from the gauge to the sender." - https://www.autometer.com/blog/faq-post/why-does-my-fuel-level-gauge-always-read-past-full/

HTH.

Edited 1 July, 20251 Jul by nta16
ETA:

Thanks for the reply, Nigel.

This is the chart I tried to attach.

Nigel,

I'm still no further forward with this but I saw on another post you had mentioned a potential cause could be an issue with the voltage stabiliser which sounds logical to me. Do you know where I can locate the stabiliser itself, what it looks like and what readings I should find on it?

Thanks,

Kieran

Hi Kieran,

I vaguely remember his but I think I was wrong about it at the time anyway. The German engineers, for many decades and still, take some sort of pride and pleasure from taking what should be straightforward things and over-complicating them and this applies as far as I know to the fuel gauge reading for Felicias. I am only used to very simple 1960s and 1970s electrics on very simple old British "sports" cars. Those had simple (but very long lasting) bi-metal voltage stabiliser that took the nominal 12v to 10v, I note 8v and 6v readings with Felicias so have no idea about the voltage stabiliser they have.

I have noted before that even those used to the German VW marque stated the dials, gauges and systems were over complex for what is really required.

Though well presented I couldn't make sense of your table of figures (or scale to resistance figures) so hoped the others would dive in to the thread.

There may be other threads and posts with more details to the numbers you should be seeing.

I guess (don't know) that the gauges will be quite common in the vehicle range but there can be wiring and other variations for model year or part year, change of parts of components during manufacture life, perhaps LHD to RHD even or VWŠkoda to VW badged vehicles but I don't know and then there might be changes made since the vehicle left the factory by previous owners and overs in 25 years so all this has to be taken into consideration.

With electrics and electronics, you want clean, secure and protected connections, wires, switches, gauges. The other posts on threads on the gauge cover this. Many times physically cleaning connections and adding protection can sort poor or intermittent supplies (be that live or earth). Sometimes connectors need firming up or replacing, wires re-ended or replaced or further protection to or on them.

Have a good look at all wires and connections, tug and wiggle connections and wires to test them , make sure all wires and connections are secured, firm and wires not hanging loose when they should be supported.

If there might be a poor earth connection one way to test this is to put in a good solid temporary connection from negative battery post (or clamp) to item. As only an example for suspect engine or main body earthing for starting the car using a jump-starting cable, this may be impractical for what you want but any suitable cable/wire with sold connections both ends would do.

To me, in your figures, the one that stood out to me was sender ground, connected with ignition on, it didn't look right at 3.46V.

Cheers for that, Nigel.

I currently have the instrument cluster removed and all 4 clocks separated from the PCB. Now I'm trying to check the PCB for defects but, as yet, have found none. Obviously there are numerous components attached to the PCB and I'm wondering if one of them is the voltage stabiliser or is it mounted separately from the PCB. Unfortunately I have no idea what it looks like which doesn't help matters. Do you have or know of an image which shows the voltage stabiliser?

Many thanks,

Kieran

I'm not sure it has a voltage stabiliser as I've no idea how the complex VW system works, you would need the wiring diagram to see what goes where. Again wiring layout and possibly colours could change variously.

You are perhaps breaking a primary rule of starting with the basics by jumping on to the sexy stuff, I'd have started at the other end with the sender and it's earthing, wiring and connections rather than going on to 25 year old wunderbar VW plastics and electronics at the instrument end.

I'm sure wiring diagrams have been posted before, how much they relate to a 2000 VW Caddy might be another matter but they would at least give an idea of how VW had things laid out how wunderbar clever they were with their electrics and electronics.

I'm almost sure someone went through all this before but can't remember what the answer(s) was/were, as well as wiring diagrams I'm almost sure there were some photos put up too @D.FYLAKTOS (and others) often has such stuff.

You might as well now keep cleaning, securing and protecting, also testing at whatever stages you can.

Thanks Nigel.

Rest assured I have already been through all the basics, the sender unit taken out & tested (10 - 160 ohms), the wiring to & from the sender (yellow positive in & blue negative out), the blue connects to a brown before making a verified earth, the yellow is verified intact at it's connection to the PCB i.e. the solder behind the 10 point connector block. So I've ruled out the sender & all the wiring which is why I'm now concentrating on the PCB (which I believe can be somewhat suspect on these models).

I have the wiring diagram for the instruments which indicates that the power supply for the fuel gauge is via a voltage stabiliser. My reading of the diagram is that this power then flows through the low fuel warning lamp (in series), then to the fuel gauge, then to the yellow wire which goes to the sender which then supplies varying degrees of earth resistance depending upon the float position. At empty the resistance becomes low enough to illuminate the warning lamp.

Although I've been chasing after the voltage stabiliser I have no confidence that it would provide the answer as I can measure 12v at both sides of the low warning lamp which would be correct for proper illumination. So I will eliminate the stabiliser as a suspect. I have already replaced the gauge itself but with the same result - always full. I am reading 7.93v at one side of the fuel gauge which coincides with one side of the temperature gauge. The same 7.93v also goes to the yellow wire for the sender so long as the sender is disconnected but this drops to 3.46v when the sender is connected, presumably due to the resistance within the sender.The other side of the fuel gauge reads 0.8v which appears quite low in comparison with the other side of the temperature gauge which reads about 5.5v when cold while dropping to about 4.5v as the engine warms and the hand rises. I now wonder if the low voltage on the other side of the fuel gauge is the cause of the hand reading full. This will be the focus of my investigation going forward and I'll post my findings in due course.

Great that you're ahead already. I am so used to those that jump to the sexy stuff ignoring the basic stuff. As I put I am almost sure (never really sure about anything much now) that I have read posts on here about fuel gauge issues and some similar voltage figure but I could be remembering wrong. A few others might know a lot more detail on this subject but they don't seem to be around here so much now.

Took me second time to realise hand was needle, I thought first time was a wrong autocorrect. 😄

I have no idea in this case but it isn't unknown for manufacturers' wiring diagrams to have errors or updates missed and if gained from other sources even more likelihood of being incorrect for application, if you can find other sources of the wiring diagram to confirm (course they all could have errors or be wrong application as such is life and Sod's Law), it's like following a TwatNav religiously, or Google maps for delivery to a certain address, I guess, wouldn't really know. 😄

Let's put a shout out and see who can reply.

Anyone able to help Kieran with this, @Papez @Breezy_Pete @Thefeliciahacker @R_U_AFA ?

Hope this helps as it's as much as I can do.

Cheers.

On 13/07/2025 at 13:12, Koponop said:

I have the wiring diagram for the instruments which indicates that the power supply for the fuel gauge is via a voltage stabiliser. My reading of the diagram is that this power then flows through the low fuel warning lamp (in series), then to the fuel gauge, then to the yellow wire which goes to the sender which then supplies varying degrees of earth resistance depending upon the float position. At empty the resistance becomes low enough to illuminate the warning lamp.

Do you have the diagram? This definitely isn't correct. The lamp cannot be operated directly by the float, because the current would be too high and the lamp would illuminate gradually instead of going off instantly.

Unfortunately I wasn't able to find schematics of the instrument panel, I only found information, that the gauge is operated through cluster of operational amplifiers in the LM2902N chip. Something like bad resistor can cause output to be stuck at full voltage, or the chip could be burned. Unfortunately I cannot say more without full schematic diagram.

The diagram I have is the attached pdf file which I downloaded from Erwin.

Hopefully you can open it?

Any thoughts greatly appreciated.

Caddy Dash Panel.pdf

21 minutes ago, Papez said:

The lamp cannot be operated directly by the float, because the current would be too high and the lamp would illuminate gradually instead of going off instantly.

My understanding is that the lamp is actually operated by the gauge and not directly by the float. However the float must regulate the resistance going through the gauge which in turn provides sufficient earth / ground to illuminate the lamp?

Just for information I can't download the pdf on my machine but then the other day I couldn't download a pdf I'd put up.

Koponop have you tested the pdf downloads for you?

If there's a problem with downloading the pdf you, i or anyone else can report this, either by using the 'Report' button (on dropdown from pressing the three-dots menu at top right of post) or (I think) posting on the 'Bugtracker' section. - https://www.briskoda.net/forums/briskoda-bugtracker/

Even if I could download the pdf and open it I'd have no idea if the information was correct or even read it or understand it properly but do bear in mind what I've put before, double-check and cross reference any information you get from any source for accuracy, even from VW.

I think professionals may use more than one source of wiring diagrams for various reasons, same as they may use various scan tools as each has its good and bad points, I'd imagine even at laboratory levels of equipment there'd be different makes and models to choose from.

Another rule of mine is always to test the test equipment before doing a test as just one false reading can give a lot of hassle and waste a lot of time, I've had that with borrowed simple leak testing equipment, the leak actually being on the tester hose. I once retested my (not expensive) digital multimeter as I didn't believe it's reading as it was outside on the coldest day of the year, but it was correct and this bloke had got his car's 12v battery so low, lowest I'd ever seen in my very limited but decades of experience with car batteries.

Watched a video the other day of some clever(?) bloke because he thought it had a design and build fault testing a cheap battery charger with a piece of laboratory electronic equipment then "modifying" the cheap charger - if he'd had read the cheap battery charger instructions he'd have known the charger was actually faulty and had it replaced for free if it was still under its 3(?) years warranty. Some blokes believe it's against some law to read instructions, particularly professionals in my experience.

Hope you get this sorted.

Otherwise do as we used to when the fuel gauge wasn't working or inaccurate, fill the fuel tank, reset the tripometer, and you gauge how much fuel is left in the tank by how many miles (km) you've done and type of driving since last refill. 😄

Ah, you're right Nigel.

The file won't open for me either so I've saved it again as an oxps.

Here goes again.

Caddy Dash Panel.oxps

Opened fine for me as it was.

4 minutes ago, Breezy_Pete said:

Opened fine for me as it was.

Very frustrating that some members have more difficulty than others with opening files.

Is this a browser problem or a host issue?

How do we get around this ?

Managed to take a screenshot of the pdf.

Hopefully this will be accessible.

18 minutes ago, Koponop said:

Very frustrating that some members have more difficulty than others with opening files.

Is this a browser problem or a host issue?

How do we get around this ?

Probably just some folks not spotting that the file has landed in their download folder, and that they need to open it from there; at a guess.

3 minutes ago, Breezy_Pete said:

Probably just some folks not spotting that the file has landed in their download folder, and that they need to open it from there; at a guess.

👍

Well it's obviously not a Briskoda site error.

The oxps file (wotever that is) did download and open for me on my machine. The pdf file I notice has 11 downloads (so far) so it's probably program errors elsewhere.

For me on my machine, with mine and your pdf the pdf boxes in the posts didn't show a download on the post even after going away from and returning to page and site, and wasn't downloaded to the Download or any other folder (you don't have to download to the Download folder) and wasn't stuck in any virus filter. not that I'm clever enough to know anything about such stuff I just click the mouse and use my eyes.

ETA: @Koponop did you check your Download (or other file) for the download?

I know from personal experience that computer folk and computer professionals often blame users for "user's error" when in fact in my experience and that of others, the masses sometimes, the error is often with the programming or equipment, scapegoating often for poor quality work and testing.

Also proves that those who are normally right can be wrong sometimes. 😁

Lessons for all of us, all the time, everyday.

The diagram looks like the horrible grey and black things presented by VW, in the VAG-COM, VCDS legacy type, appeals more to computer types than non-computer types like me.

At least you have it up for those that know to comment on, mine is only on its aesthetics. 😄

Edited 15 July, 202515 Jul by nta16

3 hours ago, Koponop said:

My understanding is that the lamp is actually operated by the gauge and not directly by the float. However the float must regulate the resistance going through the gauge which in turn provides sufficient earth / ground to illuminate the lamp?

This is diagram for wiring, the functional blocks are extremely simplified, just to show basic function of the block.

Both gauge and lamp are operated by the operational amplifier chip. Combination of resistors and capacitors around it provides delay, that prevents oscillating of the needle because of fuel sloshing around, and signal for the lamp to go off once voltage gets below set threshold. The lamp also requires some transistor, because OA isn't able to drive a lightbulb directly.

Any open solder or burnt component can cause bad signal for the gaige. But it will be quite hard to pinpoint the root cause without the complete circuit diagram.

On 13/07/2025 at 13:12, Koponop said:

Although I've been chasing after the voltage stabiliser I have no confidence that it would provide the answer as I can measure 12v at both sides of the low warning lamp which would be correct for proper illumination. So I will eliminate the stabiliser as a suspect. I have already replaced the gauge itself but with the same result - always full. I am reading 7.93v at one side of the fuel gauge which coincides with one side of the temperature gauge. The same 7.93v also goes to the yellow wire for the sender so long as the sender is disconnected but this drops to 3.46v when the sender is connected, presumably due to the resistance within the sender.The other side of the fuel gauge reads 0.8v which appears quite low in comparison with the other side of the temperature gauge which reads about 5.5v when cold while dropping to about 4.5v as the engine warms and the hand rises. I now wonder if the low voltage on the other side of the fuel gauge is the cause of the hand reading full

12V at the both sides of the indicator lamp is correct - it's operated by a transistor, which typically connects load to the ground, as this type of transistor is cheaper and overall circuitry is simpler. Same case with the gauge. There should be an 8V stabiliser in the cluster, so 8V measured on gauges is correct. As with the lamp, it's likely that top of gauges is connected directly to the voltage input and a transistor is operated between ground and gauge - this can be the failure point, a transistor can fail and be stuck in shorted state.

Voltage drop on the yellow cable is indeed caused by sender's resistance and it's correct function - the voltage should correspondent with the fuel level.

So I've managed to source a spare instrument cluster and plugged in the connectors. All appears to be functioning properly, especially the fuel gauge. Just to satisfy my curiousity I've taken voltage readings at the fuel and temperature gauges. Both indicate 7.95 at one pin, the temperature gauge reads 5.56 at the second pin and reduces as the engine warms. The fuel gauge reads 3.56 at the second pin as opposed to the 0.8 reading on the original cluster. The yellow wire going to the sender also reads 3.56 with the sender connected. I surmise that the voltage at the second pin of the fuel gauge should always correspond to the yellow wire and the fact that the second pin voltage was previously so low was the reason for the needle reading full. The problem would appear to lie somewhere between the yellow wire and the second pin of the gauge as the yellow wire previously had been reading 3.46 and the second pin 0.8. The cause of the discrepancy is a mystery to me and will remain so as I have no desire or need, (fingers crossed), to investigate further.

Hopefully this topic may help someone in diagnosing a similar problem and many thanks to those who contributed - especially Papez, as I see your last post just appeared.

Cheers,

Kieran

I'd still try to inspect all soldering joints around the gauge, there's chance one of them is cracked. If it is, you'll get free spare cluster 😁