Reading / Calibrating late Bay fuel gauge

[SGKent]

[mikedjames]

Initial signs are that the BBT sender has the correct nonlinear characteristic of the original.
_________________
Ancient vehicles and vessels

1974 VW T2 : Devon Eurovette camper with 1641 DP T1 engine, Progressive carb, full flow oil cooler, EDIS crank timed ignition.
Engine 1: 40k miles (rocker shaft clip fell off), Engine 2: 30k miles (rebuild, dropped valve). Engine 3: a JK Preservation Parts "new" engine, aluminium case: 26k miles: new top end.
Gearbox rebuild 2021 by Bears.

1979 Westerly GK24 24 foot racer/cruiser yacht Forethought of Gosport.
1973 wooden Pacer sailing dinghy

[SGKent]

wanted to add to this thread with a special kudos to some folks here who have assisted on this project.

There are also other folks who are all over the USA, Canada, the UK, the EU, who made suggestions, and we are thankful for all of them too. One member in particular whom I am grateful for is Telford. He hangs around with many of us here in California. Because he is older like many of us, some now gone, y'all probably have not gotten a chance to know him. Gary (Aeromech), and Robbie (Airschooled) know him well. I recently went to work on the bus gauge system mentioned in this thread. As part I tried to set up a late nos gauge, sender and regulator to see how they worked on a test bench. To do so I needed a stable power source equal to when the bus is running, so I grabbed a used 0-20V 10 amp regulated supply that I bought second hand. It did not work right so I used a 13.5V one instead. The regulated one is such that one can dial in current or voltage in fine increments. I hoped that by doing so I could figure out how to calibrate my existing system so it is reasonable accurate - something I no longer think is possible without the original senders. Even the VDO senders shipped for buses in the 2010 - 2015 period are not accurate in a bus. This is well documented. Anyway, before I went into the hospital I sent the defective regulated power supply to our California member, who said he'd take a look. I still owe him postage and for his time, but anyway, while I was in the hospital he let Cathy know it was done and he was shipping it back. It works fantastic. Just tested it. The prior owner had dropped it sometime and damaged some wires coming from a transformer. He solved the issue. I just want to give a special thanks to Telford in San Diego for helping out on this bus project. I just cannot say enough good things about him and the other San Diego crowd - which there are others who share many good technical suggestions at times. Thank you Telford.
_________________
"Most people don't know what they're doing, and a lot of them are really good at it." - George Carlin

[SGKent]

So here are the results of some measurements. The conclusion is that replacement VDO senders sold for the bus will never be accurate. These are appx measurements made with:

a NOS VDO sender
a NOS VW solid state instrument voltage regulator
a NOS late VW gauge
a regulated 13.85V DC power supply

The gauge will not read from full to empty as a new VDO sender ships. It is not a direct bus replacement unless one finds a NOS original $400 sender. They do come up for sale but that is a lot of money for something that does wear out. The replacement VDO sender stops must be adjusted until the gauge reads from Empty to Full over the full sender sweep. This would be the first step to take with jumper leads hooked up to the sender.

Regardless the depth of the tank, the total sweep of the float is only 8.25" when adjusted to go from empty to full, although there is an additional inch of fuel that would be in the bottom of the tank when the float is sitting at or near on the bottom of the tank. This is because the float is 1" tall. 1 inch equals about a 1/2 gallon or 7 - 8 miles range before bone dry. One would drain the tank, flow CO2 into it like from dry ice bubbling, and then one could use an endo scope safely in an empty tank full of gas fumes, now CO2, to see how close the arm is to the bottom of the tank. Do not adjust the stops again, adjust the bend in the arm.

Total drop from full to empty on the new VDO float sold for buses is 8.25" which when divided by 15 gallons (actually 14.7 or whatever) is about .55 inch that the replacement float moves per gallon. 8.25" \ .55 = 15 gallons.

At full the gauge is at the top stop whether it is sitting against the top of the tank or on the stop. It would be preferable to adjust the arm so the float just sits almost on the bottom of the tank when on the empty stop. It will rise until it hits the full stop or the top of the tank. If the tank is out, one can use a ruler and look thru the neck with a flashlight to lift the float until it is against the stop. If it hits the top of the tank first before the top stop, you will never be able to read full unless you bend the arm down, in which case the empty will now be off. You want to bend the arm so the float sits against the bottom stop when it is almost touching the bottom of the tank.

Measurements:

At 1/2 reading on the gauge, the float dropped 3.25". That equates to only 6 gallons of drop, which puts the 1/2 mark on the gauge really at about 9 gallons of fuel in the tank (when it should only be about 7.5 gallons). At the 1/2 mark you would have 1.5 gallons extra in the tank or 30 more miles than you would guess you have.

At 1/4 on the gauge, the float has dropped 4.75". That equates to about 8.6 gallons of drop. 15 total minus 8.6 of drop leaves about 6.4 gallons in the tank. The actual number should be about 3.75 gallons in the tank at the 1/4 mark, so the gauge is off by 2.65 gallons when it sits on 1/4 - which is about 50 miles more fuel than anticipated.

At the empty mark and the beginning of red R, the float has dropped 6.25", which means it is seeing 11.4 gallons of drop to get to this point. The factory specs say there is one gallon left at reserve, but the sender has only dropped 11.4 gallons so there is about 3+/- gallons in the tank when it hits reserve. The owner's manual would imply that a one gallon reserve would mean 13.7 gallons would have been used from full, but the replacement senders didn't read the owner's manual, and don't know that. This is a range of about 50 miles.

So no matter what adjustments are made, when the float is set up to sweep empty to full, the gauge will never be all that accurate.

Where this is a PITA is that after you do your best, sometimes you will still be seeing that needle enter the red zone with your next gas station 30 - 50 miles away. Will you make it, or is the gauge lying?
_______

Spare gas gauge, regulator and sender, 13.85V regulated supply

Full = 0” about 14.8 gallons or more
½ = 3.25” about 9 gallons
¼ = 4.75” about 6.4 gallons
R = 6.25” about 3 gallons
E = 8.25” about 1” of fuel

Above are measurements of float height with stops set to read Empty to Full.
Suggest set the stops then the float to the bottom of the tank and let stop set the rise. Use dry ice for CO2 or CO2 tank and regulator if using endoscope inside tank with fumes.
_________________
"Most people don't know what they're doing, and a lot of them are really good at it." - George Carlin

[SGKent]

Hmmm. This bus is a German bus delivered on the overseas program then later shipped to the US when they came home. I suspect someone stationed in the Air Force as we had both Mather and McClellan AFB here until the early 1990's. We also have Beale just north of here where spy planes and stealth aircraft fly out of. I would think the US and German tanks would be the same. I'll have to calibrate this better I think if I get the time. Otherwise I have my cheat sheet.
_________________
"Most people don't know what they're doing, and a lot of them are really good at it." - George Carlin

[kreemoweet]

I've run into several gas stations that have quite inaccurate pump displays.
Always in favor of the station, of course.
_________________
'67 bug: seized by the authorities
'68 bug: seized by the authorities
'71 kombi: not yet seized by the authorities

Stop dead photo links! Post your photos to The Samba Gallery!

[dodger tom]

sorry, steve. i’m on my third broken sender in the last five years. i am counting miles until i get a new engine, this summer, i hope. and, yes, i’m gonna put in an access port for next time.

checking my maintenance logs, the 78 ce2 has never taken more than 14.7 gallons with three top-offs at each fill.
_________________
1978 Champaign Edition 2 Westfalia
Would never find the time to keep up another classic air-cooled.

[SGKent]

back to deciding if I want to correct the sender rather than carry a cheat sheet. I cannot get a feel what gas I am using with the sender as it is. We all know that the replacement VDO senders lie.

Can anyone with a stock sender that still works tell me roughly how much gas their bus takes to fill when it is on 1/2 tank? 1/4 tank? Hitting reserve?

This bus takes about 16 gallons even though the book says 14.6 is the capacity.

Dodger Tom? What say yee.
_________________
"Most people don't know what they're doing, and a lot of them are really good at it." - George Carlin

[SGKent]

I have no idea what is going on.

The 1977 book clearly shows 14.6 gallons with 1 gallon of that total capacity as reserve. The Dutch manual shows 14.7 gallons.

My tank with a genuine VDO sender being sold for our buses and adjusted to show full at the top of the tank shows the following:

Reserve = 4 gallons
1/4 = 7 gallons
1/2 = 12 gallons
1/1 = 16.7 gallons roughly

I measured 4 gallons accurately and put in one at a time to get to the top of the reserve line, about 1 cup remained in the Gerry Can. I then drove to the station and put in 4 gallons more. That gave me the 7.7 gallons in the tank and the gauge showed slightly above 1/4. I drove to get the bus smogged, about 30 miles round trip and the gauge was sitting at 1/4 on the way home. Added 5 gallons more to get to the center of 1/2. Then it took 4.7 more gallons to full. That is a total of about 6 - 7 in the tank plus 9.7 put today in for 15.5 - 16.5 gallons total. The gauge reads a tiny bit below full now but it is hard to know if that is because I did not top it off or the gauge is a little off at full. Regardless, there is more gas in the tank than 14.6 gallons.

This bus was delivered in Germany. I don't know if that matters. Below are the 1979 specs and they show the same

This is the Dutch owner's manual. It shows "56 liter of which 5 kiters is reserve."

Maybe this is correct:

[SGKent]

well, since no one has chimed in with an original, which is understandable, I am reverse engineering it and see a float adjustment in my future. 4 gallons was top of the red line for reserve, and 7.7 - 8 Gallons was about a third of the way above 1/4 tank. In the meantime I'll make a chart and carry it in the bus so I know how much gas is in the tank. 50 year old car without access to the original factory sender.
_________________
"Most people don't know what they're doing, and a lot of them are really good at it." - George Carlin

[SGKent]

[mikedjames]

The original late bay senders have a curved former under the resistance wire in the sender. So that each turn is a different length.. that produces best accuracy.

12 years ago I bought an aftermarket sender and it had three blocks of windings with a different pitch so it was close but not so perfect.

Last year I bought an aftermarket sender and it was wound in a linear fashion.. so all of the range of the gauge vanishes in about 100 miles then you are deep in the reserve for the next 100 miles or so on the gauge.
_________________
Ancient vehicles and vessels

1974 VW T2 : Devon Eurovette camper with 1641 DP T1 engine, Progressive carb, full flow oil cooler, EDIS crank timed ignition.
Engine 1: 40k miles (rocker shaft clip fell off), Engine 2: 30k miles (rebuild, dropped valve). Engine 3: a JK Preservation Parts "new" engine, aluminium case: 26k miles: new top end.
Gearbox rebuild 2021 by Bears.

1979 Westerly GK24 24 foot racer/cruiser yacht Forethought of Gosport.
1973 wooden Pacer sailing dinghy

[metahacker]

here are a couple microcontroller "adapter boxes" to sit in between a sending unit and a gauge - particularly useful just for eliminating slosh on early bay, i'd imagine.. but basically let you read the sender actual reported range and then decide what values are actually sent to the gauge
https://www.tanksinc.com/index.cfm/page/ptype=product/product_id=748/mode=prod/prd748.htm
https://www.opgi.com/air-fuel-delivery/fuel-tanks/...23120.html
https://classicdash.com/product/universal-fuel-signal-adapter/
they're all pretty neat and the option for the low fuel light on some of them is rather cool IMO

here are some VW specific ones that include a new gauge movement/needle to be mounted on the original gauge face ... it uses a stepper motor for the movement (so it's all software positioned) and appears to have RGB LED to go amber and red as fuel is getting low
https://www.c-e-g.eu/product-categorie/gauges/fuel-gauges/
he also has an updated microcontroller driven OE style sending unit
(it's not listed on the site - you have to ask)
to order that Classic EVO stuff in the states u will need to message him
https://www.facebook.com/harold.fransen.3

telford's approach is the coolest, though :)

also, i believe the curves of the output created by original senders and _all_ replacement senders are also different?

[bsairhead]

As I am not sure about my gauge.This spring when I do my 5 year fuel line change I will note where 4 liters reads. I'am less than a 1/2 mile to my filling station and will take notes.

[SGKent]

[bsairhead]

I see the brain trust have spoken Fill till pump shuts off, drive till 1/2 tank, fill till pump shuts off. That and 5th grade math should be enough.

[mikedjames]

Another way to measure is to put a low pressure gauge below the tank and measure the height of fuel by pressure.

Alternatively you can buy an ultrasonic pinger that measures the distance to the top of the fuel.

Both of these work by using a user-provided depth versus surface area model to get the volume of fuel above the pressure sender, or below the ultrasonic sender.

I have used both (made by Maretron) on boats, we rejected the ultrasonic sender because we had an single cylinder diesel engine running at 3000 rpm (generator) for 50Hz AC .. and the ultrasonic sender went wild unless we isolated it from the vibration.
_________________
Ancient vehicles and vessels

1974 VW T2 : Devon Eurovette camper with 1641 DP T1 engine, Progressive carb, full flow oil cooler, EDIS crank timed ignition.
Engine 1: 40k miles (rocker shaft clip fell off), Engine 2: 30k miles (rebuild, dropped valve). Engine 3: a JK Preservation Parts "new" engine, aluminium case: 26k miles: new top end.
Gearbox rebuild 2021 by Bears.

1979 Westerly GK24 24 foot racer/cruiser yacht Forethought of Gosport.
1973 wooden Pacer sailing dinghy

[telford dorr]

If you really really really really want an accurate fuel gauge, there's a way to do it (but it's not trivial...)

Use a microcontroller (like an AtTiny85).

This requires that the fuel sender and gas gauge indicator are repeatable (always giving the same resistance reading for a given fuel level, and indicating the same fuel level for a given amount of gauge current). These values must be mathematically "well-defined".

You create a circuit to pass a fixed current through the fuel sender, and measure the resulting voltage with the microcontroller A/D converter. You power the gas gauge with a fixed voltage and drive it with the Pulse Width Modulated output of the microcontroller. You select the sender current to use the maximum range of the A/D converter, and the gauge voltage to use the maximum range of the PWM drive.

In software, you use a method known as piece-wise linear approximation to calculate the PWM output for any given reading from the A/D converter. This is done by finding the voltage readings for, say, 9 tank levels (empty, 1/8, 1/4, 3/8, 1/2, 5/8, 3/4, 7/8, and full. You find the pulse width needed to make the gas gauge read these levels, and put them in a look-up table. For any reading other than these fixed points, you assume that the value is linear between the known points, and so is the PWM drive. The error is usually insignificant.

Using this method, you can interface ANY gas gauge to ANY fuel sender. Also works for temperature and pressure gauges. It's how you make retrofit engines work with old gauges in old cars.

Takes a bit of work to do, but it works!

FWIW
_________________
'71 panel, now with FI
'Experience' is the ability to recognize a mistake when you're making it again - Franklin P. Jones
In theory, theory works in practice; in practice, it doesn't - William T. Harbaugh
When you're dead, you don't know you're dead. The pain is only felt by others.
Same thing happens when you're stupid. - Philippe Geluck
More VW electrical at http://telforddorr.com/ (available 9am to 9pm PST)

[SGKent]

[my59]

Wow, I have to admire the desire for accuracy and the time expended.

Given the 59 beetle came with a decades old notebook with mileage and gallons added, when I got the bus I simply instituted the same process. This was before I knew that the bus gauge reads empty after the first 60 miles or so.

SGkent, rock on!
_________________
my59: Well son, my grandfather died before I got to drive it, so does that answer your question?
our79: sunroof bus w/camper interior and 2.0 FI
Other:'12 Jetta, '77 Benz 300D, and a 74 MG Midget.