4 hours ago, BilTechnik said:

 The airbag sensor isn’t used as acceleration data is broadcast on CAN.

 

Which is exactly the same way that the ABS module transmits data to the instrument module including your percieved acceleration data.

 

Humouring you and accepting your premise that the fuel level guage has additional damping to cope with acceleration and braking forces how do you reconcile that the axis of movement of the float is at 90° to the vehicle fore and aft axis and how would your ABS calculated accelerometer data cope with cornering forces which are the ones that would in theory affect the guage were it not as damped as it is?

Are you sure it is at 90° on a Fabia @J.R.? 

@BilTechnik where did you read about this feature? I'm presuming you didn't make it up.

Removing the relevant cabin fuse for ABS on my Polo on the way home demonstrated this interaction. Braking made fuel level drop, accelerating made it increase.

Not having a gearbox speed sensor meant I had an alarming array of warning lights and no speedo!

 

Yeah several people online have hadt when their ABS module has broken. Thanks for confirming Pete

6 hours ago, J.R. said:

Which is exactly the same way that the ABS module transmits data to the instrument module including your percieved acceleration data.

Yes, ABS transmits the speed data on CAN, for your speedo and such as Pete mentions. There's no need for the Airbag controller to expose information about acceleration, so it doesn't. The processing for the "calculated acceleration" happens in the cluster. I didn't expect this to cause so much discussion, so I wasn't clear in my original post. The cluster is estimating if you're accelerating or breaking, based on the rate of change of speed, and when doing so, slows down/stops its reading of the fuel sender to prevent erratic readings.

The position of the sender in the tank is irrelvant in this case. Ideally VW would put the sensor in the centre of fluid, and the tank would be a hemisphere, and fuel wouldn't splash, that way you'd have a perfect reading regardless. But they can't do that obviously. If they have put the sensor in at 90 degrees to the direction of travel, I would assume this has been done to get closer to that perfect scenario of a central float. The sender is on the far left/right of the tank to begin with, so cornering forces are already going to have a much greater affect than acceleration of deceleration. By pointing the float inwards, you make little difference to the fore/aft position but you get much closer to the centre, thus minimising the cornering affects. It could also simply be packaging. To correct my previous statement, the Fabia dash cannot compensate for steering angle as there is no real accelerometer data, nor is there a steering angle sensor as these cars didn't come with ESP. Golf MK4 on the other hand, that may be different.

I did find it in an SSP a while ago Pete (And confirmed by taking the fuse out), will try and find it

I think the float pivots in an arc in roughly the direction of travel, and behind it's axis of rotation, judging by the polarity of my observations.

 

People may be keen to howl indignantly that Fabias do have steering angle sensors, but they'll be meaning the turn rate sensor instead.

6 minutes ago, BilTechnik said:

I did find it in an SSP a while ago Pete (And confirmed by taking the fuse out), will try and find it

Try SSP028 Octavia ESP / SSP042 Skoda ESP?

 

There are other SSP's with nuggets buried here and there

4 hours ago, Breezy_Pete said:

Removing the relevant cabin fuse for ABS on my Polo on the way home demonstrated this interaction. Braking made fuel level drop, accelerating made it increase.

Not having a gearbox speed sensor meant I had an alarming array of warning lights and no speedo!

 

 

Very interesting 👍

I guess if @TMB were to disco his gearbox speed sensor, he'd see the same sudden variability?

I was vaguely aware that the fuel gauge sender might not be permanently 'excited' having looked at the circuitry in mk1 and mk2 Fabia clocks. Easy for the relevant microcontroller to pause sampling if some threshold acceleration is calculated to have been exceeded.

 

Snippet from SSP082 Seems there is 100 degrees of angle between empty/full translated into a Voltage

 

 

SSP082 Fuel Gauge sender.pdf

I find this really interesting as I have often wondered how they get the gauge needle to stay so steady. Never realised it was relatively complex.

17 minutes ago, Breezy_Pete said:

I guess if @TMB were to disco his gearbox speed sensor, he'd see the same sudden variability?

Possibly, I assume the clocks will use any speed signal they get. Really nice retro diagram in that SSP. It was just a passing sentance, rather than a chapter as I remember. Writing the bit of code to utilise the hardware they already have and provide a subtle improvement to the customer would be very cheap for them to implement for a good reward. When I tried it, the fuel was near reserve, and it fluctuates enough that you can turn the fuel light on/off, along with the acompanying beep, with fairly normal driving. So it would be very annoying if they didn't smooth it. As to why they don't apply the strong smooth costantly, I cannot comment. I guess this is probaly only marginally more diffuclt, and is tangible better.

Thanks for all the info, BilTechnik

And to add to the subject my car with the so called saddle tank has twin sender units VCDS giving a fuel volume for each one.

 

The damping on my Yeti and previous Octavias, which I have no doubt is software driven as the guages are stepper motor driven, is very heavy throughout the whole range, I have no experience of Fabias.

 

I am anorak enough to have watched the rate of guage rise when refuelling but I do now appreciate the need for special consideration at the point the fuel warning light comes on, I'm still not convinced that it does or needs to compensate for acceleration/deceleration as these are relatively transient compared to the existing damping of the guage, I have had the fuel warning light come on just before decending a mountain pass and it did not go out, conversely on ascending a long climb at the point it would come on it did not make it come on, the trigger point is always when the "remaining miles" are at a threshold commensurate with the average fuel economy, twice on my 500 mile overnight journey at precisely 80 miles remaining indicated, in truth 80 miles plus 5+  litres virtual reserve.

 

I have also put in 20l from a jerrycan when the fuel light had just come on and it did not go out until I switched the igition off and then on again.

 

Compared to the temperature guage and its fiddled with readings the fuel guages are as honest as a Nun!

I also understand that the vehicle has inclination sensors, however it achieves it the fuel warning light effectively latches on which is a very good thing.

Just given you a groan for your 'BS' post @J.R., since you don't seem inclined to apologise to BilTecknik for it. 

You are right, I apologise BilTecknik, I will leave the posting uneditted for my shame 😳

I'm extremely surprised that anyone would bother to implement such a fiddly feature, especially using the ABS module, I guess the coding team got bored.

 

However I'll reserve the most surprise for the OEM that anyone bothers with tank senders any more, pretty much every other transportation sector uses mass rather than volume and I would imagine it would be fairly trivial to incorporate a strain gauge or load cell instead, this would allow for much more accurate and consistent measurements when the vehicle was stationary and the rest could be calculated on the fly.

Thread that we could now answer: https://www.briskoda.net/forums/topic/217059-fuel-gauge-drops-under-hard-corneringbraking/ 😁

 

Interesting that he mentions hard cornering too. Presumably the cluster could get 4 individual wheel speeds instead of, or as well as the average of them from the ABS module's broadcasts.

 

Reasonably sure I haven't noticed fuel gauge being affected when cornering hard.

 

Aiming to have a quick look at the float travel orientation in my tank in the near future, it might be a compound angle, with a bit of fore and aft and side to side, I guess. Then if speed was constant but level changing improbably fast, output to the gauge stepper could be suppressed on the basis "must be cornering"?

 

Edit: A quick look at the top of my fuel pump today versus ebay pics suggests the float does indeed pivot on an axis  about 45° off the car centreline.

 

 

 

 

Edited 25 August, 20223 yr by Breezy_Pete

9 hours ago, J.R. said:

And to add to the subject my car with the so called saddle tank has twin sender units VCDS giving a fuel volume for each one.

 

Skoda Estelles & Rapids had a saddle tank with just one sender unit making for amusing fuel level readings when in motion.

Remember the Jags like the XJ6 although it probably started well before that, with twin tanks and rocker switches on the dashboard to engage the other fuel pump when one tank ran dry?

 

I used to think they were so cool! Can you imagine the price of filling both those tanks today? - You would have to pay a minder to watch your car whenever you parked.

 

The 70's fuel crisis put most of the really desirable jags & their ilk off the road, I dont see that happening these days with the lifestyle statement Range Rovers and the like but I can't recall the last time I saw anything like a Jag or a big Yank V8 being tooled around as a daily drive.