Fuel gauge and sender mismatched

Dedicated to the memory and knowledge shared by Jim Mettler - All things relating to the flow of electrons in a Pinz.
Post Reply
redbeerd
United States of America
Posts: 6
Joined: Thu Aug 08, 2019 10:59 pm
Location: Corvallis, Oregon

Fuel gauge and sender mismatched

Post by redbeerd »

My recently inherited 710M has some issues, the one I'm focusing on now is the fuel gauge reading. I knew that the gauge needle moved, but it never seemed accurate, and we never trusted it. I want to get this thing dialed so I don't have to worry about it in the future.

EDIT: I write pretty long technical posts for the sake of clarity. I didn't think this post would end up to be this long when I started. :shock:
Also in case you're curious, the Ω symbol can be inserted by holding the "Alt" key and then pressing in sequence, the number keys 2 3 4.

SHORT VERSION: I had mismatched ohm-ranges on the sender and gauge. And I haven't ordered "matching" replacement parts yet. If you have mismatched sender and gauge, one option might be the "MeterMatch" http://www.tanksinc.com/index.cfm/page/ ... prd748.htm for about $70. The specs state that you can use any sender range with any gauge range (90 degree needle sweep), and will work with 7 to 24VDC. I haven't tried it though as I just found it last night. It probably uses a digital pot (rheostat) to map whatever your sender range is to whatever the gauge range is. Sounds slick.

LONG VERSION...

I had drained the fuel tank and dropped the tank (about 2 weeks ago) to inspect for tank liner peeling (red-death) and was relieved to not find any. It's actually very clean!

I really didn't know anything about how the fuel sender and gauge actually worked until I started reading some posts here and online. I've found that people have quite a few issues regarding fuel gauge and tank accuracy with mentions of different sender and gauge ohm ranges. So, I pulled the fuel level sender and the fuel gauge and did some electrical testing with a bench-top power supply and a multi-meter.

The fuel sender is a VDO with marking,s "817/8/2, 217286, 6 85". It has two male spade connectors on it, which means it has a switch in there too.
The fuel gauge is also a VDO with markings, "301.471/2/20 24V TWW 20/156, Made in Germany 9.77, BUND66 80-12-156-9730".

First, testing the fuel-level sending unit with an ohm-meter, I found the ohm range to be roughly 90/10 Ω empty/full.

Second, I used a bench-top DC power supply, the fuel-level sending unit and the gauge to simulate the sender/gauge circuit. This way I could physically move the sender from an upright vertical position to an inverted vertical position which would simulate empty to full movement.
I found that when the sender was upright (float at bottom of tube so the gauge should read "empty" or "L" in my case) the gauge was showing about 3/4 full.
And when I inverted the sender (float to top should be "full" or "V") I could watch gauge needle move to EMPTY!

This could explain why we never trusted the gauge. We'd fill up and the gauge would read close to empty, then after driving around for a bit, the gauge would show an increase in fuel level.

Based on some other forum posts, I tested the fuel gauge with the power supply and some resistors I had available. 10, 51, 100 Ω. Corresponding gauge levels were 1/8 full, 1/2 full, 3/4 full. So, you might see where this is going. I then used 2 legs of a (500kΩ) potentiometer to simulate the sender (which is just a rheostat or variable resistor) the fuel gauge and the DC power supply to watch the needle sweep as I turned the pot. Based on this, I found the gauge ohm-range to be roughly 10/180Ω empty/full.

I also noticed that there appeared to be a slight "hiccup" in the gauge when the needle passed through the 1/4 full mark (~30 ohm) when using my pot for testing. The needle would jump quickly when passing through this value. Initially, I thought it was a symptom of a bad gauge. This actually turned out to be my 500 kΩ pot. I checked the ohm values with an ohm-meter and could see the value jump from about 20 ohms to 55 ohms even when I turned the pot very slowly. Using a 500kΩ pot is not the best device to test an ohm range of 0-180 ohms, and it also appears that there are some dead spots in pots. I just figured this out today during more testing.

So, not only do the gauge and sender have different ohm ranges, they're "almost" exactly opposite in their empty/full values.

Trying to narrow my options for a fix has been an experience, to say the least, so I thought I'd share some of what I found.

SAV (Swiss Army Vehicles) offers VDO fuel senders and gauges, though the sender only has one connector on it (no switch for dummy light). When I contacted them last week to confirm the ohm-range on the sender, they replied with "0/90 Ω empty to full". They also mentioned that the gauge is "matched". So, I ordered these 2 items last week, and received them a few days ago...and did some more testing on them.

The gauge had markings on it that indicated the ohm range was 90/0.5 ohms (empty/full). Bench-testing confirmed this. Opposite of what SAV had said.
I checked the sender with an ohm meter and found the range to be roughly 75/2.5 ohms (empty/full).

As the values show, this is not exactly "matched". And, testing these new units with a PS (power supply) I found that although there was good needle sweep from "empty" to about "3/4 full", it still wasn't accurate. So I started looking at other options.

What you need is a matched sender and gauge. It doesn't matter if the range is 0/90Ω empty/full, 240/33Ω empty/full, etc. as long as both sender and gauge are the same. The tricky part comes in when trying to find a sender that's the right length (~330mm stock length, and in stock) with an ohm-range to match a gauge.

I've mainly been using the VDO website as a reference for part numbers, then trying to find them on auto-part suppliers online, mainly Summit Racing. I've read that some of you have used the switch on the sender wired to a "dummy light" on the dash. I liked this idea, and wanted to find a gauge with a built-in warning light to minimize drilling/wiring on the dash.

My likely route will be the following...

VDO offers a reed switch sender, p/n: VDO 226-633, that has a 240/33 ohm range and is 330 mm long. About $57 online from Summit (not stocked, lead-time ~2.5 weeks).
VDO guage, Viewline series, p/n: A2C53412988S, it's 2 1/16 inch diameter (~52mm) that is also 240/33 ohm range and has a built-in LED warning light. About $48 from Summit.
And some extras needed for this would be an adapter ring for the smaller VDO gauge as well as a "stud mount" kit. I may be able to use my original stud-mount kit though. Again, Summit has or can get these.

I'll update this when I finally get some parts! May be a while.
Erik
1974 710M
Ham: KJ7DKD
User avatar
rmel
United States of America
Posts: 1395
Joined: Sun Sep 25, 2011 12:19 pm
Location: Woodside, CA
Contact:

Re: Fuel gauge and sender mismatched

Post by rmel »

The proper sending unit -- as you have deduced -- is Low impedance when Full and
Max impedance when the Tank is at the "Reserve Level". The reserve level is NOT Empty.
You probably realize this but the sender sits above the floor of the tank so when the
Gauge reading hits the minimum there is approximately 5 gallons of reserve in the tank.

Your observations on your Gauge is odd, e.g. inverse readings. Perhaps the PO can give
you some history having replaced the Gauge in the past -- and with the wrong model.

I have found that the Senders resistance characteristics are difference "dry" versus "wet".
I don't know why this would be the case but on the bench I have seen two different Senders
measure nonlinear readings on the bench dry but work normal in the tank when wet with gas.

I have found a source for senders with a warning contact at:
https://www.eurospares.co.uk/parts/ferr ... gI4oPD_BwE
The PN# is 224-817-008-002R. For reference the VDO sender catalogue can be found here:
https://www.vdo.com/media/182591/flc_se ... ion_en.pdf

I also suggest you give the SAV Gauge and your old sender a try with a "wet" tank. You may find it's
accurate enough. I'd also recommend you make use of that warning contact via an LED on your dash.
It's a foolproof way to know when your down to your last 5 Gallons and time to toss in the contents
of your Jerry can :wink:
Puller: 71' 710K 2.7L EFI aka Mozo
Follower: Sankey MK 3, 3/4 Tonne
Rescue Pinz: 73' 712MK

Driver: Ron // KO0Q
redbeerd
United States of America
Posts: 6
Joined: Thu Aug 08, 2019 10:59 pm
Location: Corvallis, Oregon

Re: Fuel gauge and sender mismatched

Post by redbeerd »

You probably realize this but the sender sits above the floor of the tank so when the
Gauge reading hits the minimum there is approximately 5 gallons of reserve in the tank.

Your observations on your Gauge is odd, e.g. inverse readings. Perhaps the PO can give
you some history having replaced the Gauge in the past -- and with the wrong model.

I don't know why this would be the case but on the bench I have seen two different Senders
measure nonlinear readings on the bench dry but work normal in the tank when wet with gas.
In response to your comments Ron, yes I saw the sender length vs. the tank depth, I sketched out the tank and dimensions when I had it down, thinking I may want to make another, or larger tank in the future.
The PO was my dad, who passed away a couple years ago. He bought it in southern California, just not sure from who.
When I graphed out the sender position to the gauge reading, they were all pretty close to linear. My table of graphs shows this pretty good (just need to figure out how to post it).


And a parts update:
I finally received some new parts, but not the ones I originally thought I would get.
Turns out that SAV doesn't accept returns on electrical items. So I figured I'd stick with the SAV sender and gauge, and order the MeterMatch (MM) directly from the manufacturer.

With my "old" gauge and sender and the "new" SAV gauge and sender I had 4 possible combinations between all these parts. After playing with the MM for about an hour I decided to do some testing to see which combo would work the best. I did some "dry" testing again with and without the MM and all the possible combos, on 12VDC and 24VDC (with a variable DC power supply). Dry testing was the easiest option at this time for me since it allowed me to control the sender float location. There are a couple ways to "calibrate" the MM with and without a sender, but I think the easiest and most accurate is to use the an actual sender and gauge.

I used a 1/16 inch diameter welding rod with marks on it so I could physically push the float up inside the sender to 1/4, 1/2, 3/4 and Full positions. Then I could compare the actual location of the float compared to the gauge's reading. I drew up some tables to compare these values. And from those tables made a table of graphs all on a single page. (What's a good place to post pictures that I can link to from this site?) Based on these results, the 2 most accurate combos with the MM came down to both the "old" and "new" sender with "old" gauge. But, the MM's built-in "alarm" circuit wouldn't work with the "old" gauge. I think it has something to do with the gauge's backwards ohm range, 10/180 (empty/full). Though I could use the "old" sender's built-in alarm circuit.

The most accurate combo that allowed me to use the MM's "alarm" circuit was actually the "old" sender and "new" gauge. Again, I could use the sender's alarm circuit here.

I found that the needle sweep range on the "new" SAV VDO gauge would not peg out on "F", no matter what I tried. Even shorting the wire from "G" (back of gauge to sender) to ground which simulates no resistance and should read "full", the needle would only go up to just a hair above 7/8. So, even with the MM in there, there was no way I was going to get the new gauge to read all the way full.

On my "old" gauge, I was able to disassemble (and re-assemble) it including the pointer. This allowed me to re-attach the pointer so this makes me think that IF I took apart the new gauge, I could set the pointer a little more clockwise, which may give me the "FULL" reading. But, that would require me to take apart a new gauge, which I don't really want to do right now. The new gauge does have an adjustment screw to "dial-in" the needle when it points to "E", but this won't affect the reading on the FULL side.

Based on other posts, I'd like to use a setup that allows me to run a "dummy light" for an empty fuel tank. While I had my fuel tank down, I did run 2 more wires from the tank area back to the dash. One for a dedicated "sender ground" and the other for a "sender switch".

It might be a while before I can test this thing with fuel. I had the engine shrouds off for paint, and had also removed the fan/mount and alternator for access. After seeing the underside of the dash, I want to clean up some wiring runs there too. The aftermarket VDO tach has 5 connectors and all the wires are yellow! :(( I'd like to replace the stock speedo soon too, but that may have to wait a little.
Erik
1974 710M
Ham: KJ7DKD
redbeerd
United States of America
Posts: 6
Joined: Thu Aug 08, 2019 10:59 pm
Location: Corvallis, Oregon

Re: Fuel gauge and sender mismatched

Post by redbeerd »

OK, got some pictures uploaded to the 'imgbb' site so I can link to some pictures.

1. The MeterMatch unit.

Image

2. Close-up of terminal strip on MM. Works on 7-24VDC even though PCB says +12V.

Image

3. Main IC in the MM. PIC18F1320

Image

4. Test results with 2 senders and 2 gauges. First column is not using the MM at +12V, 2nd column is +12V and using the MM, 3rd column is +24V and using the MM.
Gauge #1 is my "old" gauge that was in the truck. Gauge #2 is the "new" unit from SAV. Same for "senders".
The circled notation on some of the graphs indicate if the sender's alarm switch is available (S.) or if the MM alarm is usable (MM).
The diagonal line is the ideal fit, 1:1.

Image
Erik
1974 710M
Ham: KJ7DKD
redbeerd
United States of America
Posts: 6
Joined: Thu Aug 08, 2019 10:59 pm
Location: Corvallis, Oregon

Re: Fuel gauge and sender mismatched

Post by redbeerd »

I did a little more tweaking with the MM today and found that if I also used the Mid-Hi and Mid-Lo calibration points (3/4 and 1/4 full) that the gauge reading becomes more accurate than when I only used the "high" and "low" calibration points. I still can't make the gauge read all the way "FULL", but the needle doesn't move down from 7/8 until the float passes that location roughly.

Based on this, I plan to install the new SAV gauge and sender and use the MeterMatch with it's built-in (blinking) "warning" light circuit.

Just thought I'd post this in case anyone else runs into some gauge/sender issues.
Erik
1974 710M
Ham: KJ7DKD
Post Reply