Fuel Strip to Float Conversion...Dummy Run

Last time I ran out of fuel (a few weeks ago) with the GSA I managed to put 35 liters in the tank (should be around 33).
2 blobs over the arrow with 20 liters to fill as you say seems consistent with a GSA tank capacity.

fucking_fuel_strip.jpg
I have put in 35,5 litres in the past on a few occasions when I haven't run out of fuel, must have been on vapours.
I like the way this fix is panning out, 3 fuel strips in the last 18 months is silly.
 
I've not email Henk about any of this yet...I am trying to get things straight in my head first before trying to share ideas in a language I am not familiar with!

From trawling the Internet this is what I think we have:

Fuel strip resistances:
  • Heater circuit (pins 1 & 4) = approx. 32Ω
  • Fuel level circuit* (pins 2 & 3) = 2.4kΩ
*No heating applied = equivalent to tank full.

From what I have found the fuel strip calibration does this:

The bike reads the 'un-heated' resistance of the strip and records that as 'tank full', then heats the strip and records the higher heated resistance as 'tank empty'.

So if I zap my old strip (to make sure it is at 2.4kΩ before I start) and then calibrate it, provided I am quick enough I should be able to get an idea of the resistance for empty. This is not a value I have been able to find anywhere, so presumably would be useful?

If during this I am able to observe when voltage is being applied to the heating circuit I will probably then have a go at calibrating with the float sensor attached...my assumption being I will just need to move the float between the extremes at the relevant time when voltage to the heating circuit changes.
It is a long shot but I am hoping that this will tidy up the inaccuracy (albeit within the constraints of the float only be effective from half full) and shift the gauge range, though I am not optimistic the ZFE will be able to cater for such a large shift.

Flaws in the plan?
 
Hmm, that did not play out quite as expected...

Resistance of the fuel level circuit on the strip I have here is 3.9kΩ not the 2.4kΩ that is mentioned elsewhere (a few zaps didn't make any difference either).
Anyone know if the base resistance varies depending on year?

The strip I am using would appear to be OK because a test calibration ran without generating any faults.

Unfortunately I can't monitor fuel heater voltage with the GS911 at the same time as running the calibration, so I will have to tap into the wire and measure with a meter to be able to determine the timing profile of the calibration.

Nonetheless I had a go at calibrating the float sender based on moving it from high to low at around midway through the calibration. Again no errors, but bizarrely afterwards irrespective of float position the gauge resolutely showed one block above the arrow/triangle, and no low fuel warning - even with the float in the lowest position.

Anyhoo to make sure I had not fried anything I re-calibrated using the strip again and the gauge has returned to reading full (the tank probably has about 17 litres in it), so it would appear the converter did not suffer during the calibration.

I had a go at measuring the resistance of the strip post calibration but to be honest I was not very quick getting it disconnected so it is probably not terribly accurate, but I measured approx 4.2kΩ decreasing slowly, at a guess I'd say maximum was probably close to 4.3kΩ based on the time it took me to get my act together.

When I get a chance to make up a better set of test connections/plugs/leads I'll have another go, but my ABS pump has gone FUBAR in the last week, so I'll be spending some time on that first!
 
The fuel strip removed from the bike showed 36ohm on the heater circuit but open circuit on the sensing strip. That's dry - tank empty.

The lads over the pond got it to display full with a resistor of 2K ohm.
I bought some trimmer resistors to have a play but just no time to make up a circuit.
 
The fuel strip removed from the bike showed 36ohm on the heater circuit but open circuit on the sensing strip. That's dry - tank empty.

Open circuit on sensing strip = knackered strip no?

Any hoo some further investigation reveals the following range of resistance on my dodgy sensing strip (observed during normal ignition on cycling of the heater circuit with the strip dry and out of the tank):
  • Lowest resistance approx 3kΩ
  • Highest resistance approx 11kΩ

Fuel strip calibration cycle (all timings and measurements approx. and done with the same dodgy strip...though with no errors):
  • 0:00: heater circuit 0V, sensing circuit 3kΩ
  • 0:40: heater circuit 4.2V, sensing circuit 8.5kΩ
  • 1:10: heater circuit 5V, then steadily increasing over 20 seconds or so to 5.6V, sensing circuit 11kΩ (no subsequent increase with voltage)
  • 3:00: heater circuit 0.65V, sensing circuit resistance drops (but I was not quick enough to note what it dropped to!)
  • 3:10: heater circuit 5.6V, sensing circuit 11kΩ
  • 3:35: heater circuit 0V, sensing circuit 3kΩ
  • 4:00: completes

Not sure what all this means but it must be useful somehow :nenau
 
Bendy, do you think any savvy DIY enthusiast could mimic Henk Poskam's converter and install it with float?
Henk has ceased trading now sadly.
It seemed to have worked for other members on here.
Wish I'd ordered one sooner!!

Sent from my SM-J510FN using Tapatalk
 
Bendy, do you think any savvy DIY enthusiast could mimic Henk Poskam's converter and install it with float?
Henk has ceased trading now sadly.
It seemed to have worked for other members on here.
Wish I'd ordered one sooner!!

Sent from my SM-J510FN using Tapatalk

Hi!
I have got one unused converter bought last month from Henk. Tried out, but won't fit the RT. So it's for sale.
110 GBP + shipping (10 GBP)
 
Thanks, but it seems a little pricey.
Henk knocked them out for €78.95.
Or is the price you charge, for the float assembly as well?


Sent from my SM-J510FN using Tapatalk
 
No, it's just the converter itself. Maybe you will find a cheaper one for sale.
 
Bendy, do you think any savvy DIY enthusiast could mimic Henk Poskam's converter and install it with float?
Henk has ceased trading now sadly.
It seemed to have worked for other members on here.
Wish I'd ordered one sooner!!

Sent from my SM-J510FN using Tapatalk

Its sad that Henk has ceased trading. Maybe it was product liability issue.

My stick float goes from 4Ω to 180Ω. Its a pretty standard unit that works with any 4-180 head unit. It reads 100% for about 100 miles, then slowly drops to 70% after which it quite rapidly drops to 20%. I've had the bike run out of fuel showing 20% but this time it kept going. It normally sits for a long time at around 50% (while the RHS is emptied) so this looked a jet pump issue. I didn't want to stop to add petrol on the hard shoulder so pulled off at Exeter services. As I waited at traffic lights the level gauge gradually moved up to 50%. With the engine ticking over there was now enough fuel flow to drive the jet pump. I filled up later (Tesco Momentum is cheaper).

The AdvRider thread on fuel pumps gauges etc suggests the fuel pump duty cycle changes with fuel level. At more than (say) 50% the pump runs at a low mark space - enough to feed the injectors with minimal return flow. I read elsewhere this is to avoid fuel heating and evaporation. Below 50% the pump cycle steps up so there's enough flow to pump fuel across the tank from right to left.

I believe the ECU really does change pump duty cycle according to what it sees from the fuel strip. I don't have a fuel strip so my pump stays on low cycle. However someone with a broken fuel strip indicating empty will be running their pump hard all of the time wearing out the pump. Funnily enough Ive had mine like that for at least 15K miles so no surprise my original pump wore out by 60K.

Running at "make progress" speeds on motorway the new (used) fuel pump empty the left side (and so drop the level reading) but not have enough left over to really drive the fuel transfer. Slow to low speeds and fuel gets pumped across.

To get my float sensor working with the ECU I need to know the fuel strip resistance values at full and empty. My fuel strip is broken so I have nothing to measure. Ideally the figures need to be real world in-tank values as the levels drop with use.

@AlexG - Do you think the fuel strip resistance readings you took are a linear scale? I imagine they pretty much are.
To get either my 4Ω to 180Ω or a 0Ω to 100Ω float to work with the ECU we would need a converter box such as an Arduino I/O processor. Now to find how to program it.
 
@AlexG - Do you think the fuel strip resistance readings you took are a linear scale? I imagine they pretty much are.
To get either my 4Ω to 180Ω or a 0Ω to 100Ω float to work with the ECU we would need a converter box such as an Arduino I/O processor. Now to find how to program it.

To be honest I have no idea, but like you I assume they are.

Given the size of Henk's converter I assume there is some way of using standard electrical components without the need for any sort of clever processing?
 
To be honest I have no idea, but like you I assume they are.

Given the size of Henk's converter I assume there is some way of using standard electrical components without the need for any sort of clever processing?

I don't know. Input of 4Ω to 180Ω (or 0Ω to 100Ω) is a long way from 3KΩ to 11KΩ used by the ECU. They all have low = empty but I think would need a microprocessor to do the resistance conversion. Does anyone know another way?

It would be nice to have the full range input and output values from Henk's device.
 
Hope this is not against forum rules, but if someone by any chance has a spare GSA float to sell... I'd happily get it out of your hands.
Had Henk's converter sitting in a drawer for a while now...
 
Fuel Strip solution

Bendy, do you think any savvy DIY enthusiast could mimic Henk Poskam's converter and install it with float?
Henk has ceased trading now sadly.
It seemed to have worked for other members on here.
Wish I'd ordered one sooner!!

Sent from my SM-J510FN using Tapatalk


Right now I'm busy designing a small electronic circuit for it.

Bert:thumb
 


Back
Top Bottom