Harley Davidson V-Rod Forum banner

1 - 20 of 24 Posts

·
Premium Member
Joined
·
6,248 Posts
Discussion Starter · #1 ·
Comment for cbdane or any other expert on the Harley serial data protocol concerning the persistent VRSC fuel gauge sender problems:

A fairly reliable fix for this problem is possible using the fuel use information that is available from the VRSC ECM serial data stream. This 1/10 cc fuel consumption data could be used to operate a resistor network that would provide a variable 50 to 260 ohm load (from full to empty) on the fuel gauge circuit. User selectable fuel tank volume could be used to "linearize" the fuel gauge motion. Fairly simple jumper wires could be supplied to read the serial data line and tie in to the fuel flange electrical connector. I know it's possible with the older fuel tank float system and may also be adaptable to the newer sonar sensor (I haven't seen a 2007 VRSC wiring diagram).

If someone made this device I think you could sell them.
 

·
Posting From The Pub
Joined
·
7,850 Posts
V-Gauge. Already does it.
 

·
Premium Member
Joined
·
6,248 Posts
Discussion Starter · #3 ·
Louis said:
V-Gauge. Already does it.
There are many people who don't want to mount a new instrument cluster and don't like "digital" displays. I'm talking about a device like a speedometer calibrator that corrects the current analog fuel level gauge.
 

·
1130cc Ninja
Joined
·
3,972 Posts
i think the moco could get some great ideas from the vgauge
 

·
Dr. Laser Guy
Joined
·
997 Posts
Interesting idea.

stever975 said:
This 1/10 cc fuel consumption data could be used to operate a resistor network that would provide a variable 50 to 260 ohm load (from full to empty) on the fuel gauge circuit.
Do you know if the new "ultrasonic" sensors also just provide a variable resistance back to the cluster? I wouldn't be surprised since I assume that there is not a special cluster design used only for the new level sensors. Not sure though.

I like your idea. Only thing is there would need to be a button/switch somewhere to tell the system when you refueled and perhaps some dip switches to set the tank volume.

Finally, can you point me to a reference that has the definitive resistance range mentioned above? Did the 50 to 260 come from your own measurements or is that indeed what the standard is supposed to be?
 

·
Premium Member
Joined
·
6,248 Posts
Discussion Starter · #6 ·
cbdane,
I don't know any details on the new "ultrasonic" sensor but I would think you're right since the low fuel warning and level indicator appear to function (or malfunction) in a similar way.

Maybe someone with a 2007 wiring diagram can answer with better information.

You're right about the need for dip switches for setting fuel tank volume and a reset switch for zeroing after a fill up but the module could be designed to fit near the fuel flange and the zero switch would be easy to reach.

The 50 to 260 ohm values are based on my measurements. The HD Electrical Diagnostics manual lists the range as 50 to 261 ohms but my experience indicates that the gauge wouldn't know the difference between low values from 46 to 53 ohms and high values from 255 to 261 ohms.
 

·
Dr. Laser Guy
Joined
·
997 Posts
stever975 said:
You're right about the need for dip switches for setting fuel tank volume and a reset switch for zeroing after a fill up but the module could be designed to fit near the fuel flange and the zero switch would be easy to reach.

The 50 to 260 ohm values are based on my measurements. The HD Electrical Diagnostics manual lists the range as 50 to 261 ohms but my experience indicates that the gauge wouldn't know the difference between low values from 46 to 53 ohms and high values from 255 to 261 ohms.
What we would need is for someone with a 3.7 and a 5.0gal tank to fill up in about 0.5 gallon steps (or smaller the better) and record the sensor resistance. I could then build an interpolated table of fuel remaining versus resistance for both setups.

Do you happen to know the input resistance a the cluster? I can check mine tonight.
 

·
Premium Member
Joined
·
6,248 Posts
Discussion Starter · #8 ·
cbdane said:
What we would need is for someone with a 3.7 and a 5.0gal tank to fill up in about 0.5 gallon steps (or smaller the better) and record the sensor resistance. I could then build an interpolated table of fuel remaining versus resistance for both setups.

Do you happen to know the input resistance a the cluster? I can check mine tonight.
Do you really need to know the incremental readings? I thought you could get the amount of fuel consumed from the injector values.

Anyway, I have one of the beta testing UE tanks and this is what I had for data when I first installed it and had just cleaned my sending unit:
total______Pumpable___level
fuel in____fuel in____sender
tank_______tank_______reading
(gal.)_____(gal.)_____(ohms)
0.063______0.000______252.9
0.563______0.500______251.9
1.063______1.000______207.4
1.563______1.500______163.9
2.063______2.000______138.2
2.563______2.500______108.6
3.063______3.000______87.0
3.563______3.500______71.1
4.063______4.000______53.8
4.563______4.500______45.7
4.813______4.750______45.7
4.938______4.875______45.7

I don't know the cluster resistance right now.
 

·
Dr. Laser Guy
Joined
·
997 Posts
stever975 said:
Do you really need to know the incremental readings? I thought you could get the amount of fuel consumed from the injector values.
I am worried that the cluster already has some sort of correction table in it making the needle position not necessarily linear with float height because of changes in the tank cross-section with depth. It's an attempt to reverse engineer any correction algorithm in the stock gauge. Even though we'll know the exact fuel used, we have to trick the fuel gauge into displaying the correct needle position. On the other hand, the tank is "mostly" rectangular and the HD engineers may not have even bothered to try to add a correction. Am I making any sense? (not a rhetorical question :))
 

·
Premium Member
Joined
·
6,248 Posts
Discussion Starter · #10 ·
cbdane said:
I am worried that the cluster already has some sort of correction table in it making the needle position not necessarily linear with float height because of changes in the tank cross-section with depth. It's an attempt to reverse engineer any correction algorithm in the stock gauge. Even though we'll know the exact fuel used, we have to trick the fuel gauge into displaying the correct needle position. On the other hand, the tank is "mostly" rectangular and the HD engineers may not have even bothered to try to add a correction. Am I making any sense? (not a rhetorical question :))
You're making sense, however, from what I've seen there's been no attempt by HD to correct for the shape of the tank. It wouldn't hurt for you to look at the data also since I didn't try correlating ohm readings to gauge position. There may be some non-linearity in the markings on the gauge but I suspect the gauge moves in direct correlation to the ohm readings.

Someone else will have to get you data for the stock tank. I have my O.E. tank but it would take a little more time than it's worth to switch the sensor to the old tank and fill it with fuel.

The other problem that the current float based sensor produces that could be corrected with this technique is inadequate range, especially with the taller early release UE tank, but still a problem with the O.E. tank. The ultrasonic device could also have this "feature" but it's other problems overshadow it.
 

·
Color me Gone
Joined
·
27,333 Posts
I don't think the sonar functions as you think it does. For one thing it is similar to a data buss and only checks in on intervals. My understanding is that the interval is 10 minutes. If your going to go to that much trouble with switches that would have to be reset (tough to monitor and activate) then just stick a float slide back in the tank. While people complained about the old float it didn't give nearly the trouble the new one does. The old one was also easy to repair when it did give problems. The sonar is a great idea I just think the small tank volume, fuel additives and possibly weather are all playing a factor in poor readings. It may even have something to do with the check in time also.

Brent on a side note I found out from the engineers that 07 and 08 bikes have a built in telltale system that can be activated. It will give a great deal of feedback on riding habits, speeds etc. It is a voluntary monitor that has to be activated I assume by a scanalyzer or such and usually people that allow it are compensated by the moco for the data.
 

·
Premium Member
Joined
·
6,248 Posts
Discussion Starter · #12 ·
Max,
I can't argue concerning the 2007 models except that I would still think that the fuel level indicator is a simple analog meter controlled by a voltage supply with a regulated load. Reverse engineering to determine this I can't address since I don't have access to a newer bike.

Concerning the older float system, sure we can take them apart and clean them and they'll work as well as new:
1) insufficient float travel
2) non linear behavior (relative to fuel quantity)

As you suggest this fix might be too much work = expense; I'll leave that question to others like Brent. If it can be done for about $100 (as much as a speedometer calibrator), I think there are a few people that would spend this to have a fuel gauge that provides useful information.
 

·
Dr. Laser Guy
Joined
·
997 Posts
cbdane follow-up

Follow-up: I've thought a lot about steve975's excellent suggestion and it is almost a done deal. Can't predict whether it is a month or 3 months but I will develop and test this circuit. My analog gauge was always screwed up on my 03 and now my 06 reads dead empty for the first 30 miles after fill-up. Yep, I have the V-Gauge (which of course is right on :)) but it still bugs me to have non-functional HD stock equipment staring at me and taunting me.

Here's what I imagine. A small plastic box (1.25" square) with an LED and a reset button. A wire lead with a pre-installed Deutsch connector plugs into the diagnostics port (temporarily unplug it for service, SERT downloads, etc.). Another two plugs will allow it to plug directly inline on the tank sensor harness. The power for the fuel pump will just pass through. Velcro it to the top of the tank flange.

Procedure: set a little switch the first time you use it to choose between 3.7 and 5.0 gallon tanks. After that, refill as normal. After each refill, press and hold the button until the LED comes on to indicate a reset. That's it. One small extra step and the needle always reads correctly. The float, sonar, or whatever other contraption is in the tank is not even used.

What'dja think? Stay tuned!
 

·
Banned
Joined
·
15,036 Posts
cbdane said:
Do you know if the new "ultrasonic" sensors also just provide a variable resistance back to the cluster? I wouldn't be surprised since I assume that there is not a special cluster design used only for the new level sensors. Not sure though.

I like your idea. Only thing is there would need to be a button/switch somewhere to tell the system when you refueled and perhaps some dip switches to set the tank volume.

Finally, can you point me to a reference that has the definitive resistance range mentioned above? Did the 50 to 260 come from your own measurements or is that indeed what the standard is supposed to be?
We've already had a couple of dip switches working on that!! :angeldev:
 

·
Scott Whitney
Joined
·
1,211 Posts
Interesting stuff. Good idea Stever!

One small bonus:
The float and a few other small pieces could then be removed from within the tank. That may give you volume for 0.05 more gallons of gas in the tank, and a couple more miles between fill-ups. Then Brent can advertise it as a fuel range extender upgrade too!

While you’re at it Brent, you can put on a remote “mode” button, and a few more lines of code in there, and allow that needle to also indicate some other cool stuff:

Coolant Temperature:
E = Cold.
1/4 = Temperature when the thermostat opens or when the rev-limiter bumps up.
1/2 = Temperature when the fans turn off.
3/4 = Temperature when the fans turn on.
F = Temperature when the red coolant temperature light comes on.
These indications require a very non-linear scale, but can be real useful.

Instantaneous Fuel Mileage:
Needs to be magnified a bit, so it has some visual meaning. This is more of a gee-wiz feature that tells you you’re twisting the throttle too much. Once a bunch of people buy one of these, we can have contests on the forum for who can keep the needle above F, or below E the longest, i.e. who’s going fast and who’s saving fuel the most. (that's a joke!)

Clutch Slip/Wheel Spin Indicator:
This one might be slightly trickier programming wise than the other features, but still doable just with more lines of code.

Gear Position Indicator:
F = 5th.
3/4 = 4th
1/2 = 3rd
1/4 = 2nd
E = 1st
Could be creatively combined with Clutch Slip/Wheel Spin Indicator so you don’t have too many modes. I.e if the needle moves away from one of the 5 marked points, you're getting slip/spin.

Battery Voltage:
I think this one requires an A/D, but it may be very helpful for some people.

With a single remote Red/Yellow/Orange LED you could indicate up to 3 different modes. Or skip the LED, and just make the needle temporarily go to one of 5 positions for two seconds then jiggle a bit just after switching modes (with pushbutton sequencing like V-Gauge). You would have to remember which needle position is which mode, but it keeps the device simple. I.e. push the mode button and if it goes to F for two seconds, giggles slightly, then moves to some new position, that means you’re in Fuel Gauge mode. If it goes to 3/4 for two seconds, jiggles slightly, then moves to some new position, that means you’re in Fuel Mileage mode.
F = Fuel Gauge.
3/4 = Fuel Mileage.
1/2 = Coolant Temperature.
1/4 = Clutch Slip/Wheel Spin Indicator.
E = Battery Voltage.

How do you deal with the aftermarket tank sizes, and unknown future tank sizes? You might not need the dip switches if you have a "learn" mode where you fill the tank full, push the "just filled" button, ride it until it runs out of gas, then push the button again to tell it "thats how many gallons the tank holds". Oops, that's a dumb idea, . . . but you have to admit it would work!

Love the new “V-Gas” gauge!
 

·
Registered
Joined
·
665 Posts
How about a pass though option for people iPods...so the fuel gauge bounces up and down like the gauges on old style amps? (joke)

I'd buy one as described by CD bane. (no joke)
 

·
Dr. Laser Guy
Joined
·
997 Posts
HogWild said:
While you’re at it Brent, you can put on a remote “mode” button, and a few more lines of code in there, and allow that needle to also indicate some other cool stuff:
Hey Scott! These are some great ideas but there may be one problem: the fuel gauge needle is controlled by the cluster and I'm pretty sure that there is some time filtering that goes on to prevent excessive needle movement due to fuel sloshing, etc. I need to do some tests to see how fast the cluster processor will actually move the needle in response to changes in input.
 

·
Scott Whitney
Joined
·
1,211 Posts
cbdane said:
Hey Scott! These are some great ideas but there may be one problem: the fuel gauge needle is controlled by the cluster and I'm pretty sure that there is some time filtering that goes on to prevent excessive needle movement due to fuel sloshing, etc. I need to do some tests to see how fast the cluster processor will actually move the needle in response to changes in input.
Yea, unfortunately that seems logical. To overcome that, you would have to dig deep and figure out how the whole protocol works so you can inject your own commands onto the data bus. If you overcame that large hurdle, then you would have control of LOTS more. Got a few hundred spare hours to do all that work?
 

·
Chrome is king!!!
Joined
·
20,269 Posts
I have a Dakota Digital fuel gauge on my bike. Do you think I could make this work for me? I am using the fuel sending unit output wire, which is yellow/ white stripe.
 
1 - 20 of 24 Posts
Top