Modern Vespa

Can you repeat the test, but jump the float to ground, before you connect the +?

By "jump", do you mean bridge? (Wire from battery - to both float and - )

If that's the case, the needle went left.

Ok...
A conundrum indeed... Ok, I'm gonna remove the instrument from my scooter and do some tests locally. It's an LML instrument, but should be useful.

I'll come back to you.

In the meantime, because my hunch still screams check polarity, try this test please. Just a makeshift battery across two terminals, and observe the motion of the needle.

No reaction at all observing the polarities on your sketch. I made sure the battery was still delivering (it's still at 6.3V), and even moved the needle to the right to see if it was still moving freely, but no reaction.....

EDIT: I tried again, and it does move to the left a tad; just not as abruptly.

But reversing the polarities: - from battery to right of resistor, and + from battery to "-" post, the needle moves to the right.

Acts as if it's an open circuit, but we have resistance (223R) so we know it's not. Maybe 6V is not enough to generate the field required to move the coil to E (empty). In which case, we need more voltage. Perhaps two more batteries, or to use a scooter battery.
Mathematically it equates to about 28mA of current @6.3Vdc, you could safely measure it with your meter, if you wanted to confirm. But, keep in mind that unlike voltage, which is measured in parallel, current is measured in series. So you'd have to insert your instrument between the battery and the gauge.

Danger involved with el.current measurements is associated with passing higher current that the multimeter can conduct. This inevitably fries the instrument, or if the internal fuse is properly dimensioned, that's a minor failure. Usually not, in economy-friendly multimeters.

Another thing we can try is this:

The needle moves briskly to the right. Could this be the key to the issue?

Should I try the measurements prescribed anyway?

How far did the needle go? To Empty or to Full (or somewhere in between)?

The needle is always on empty (to the left) by default. It went to full (right), at a rapid pace but without banging on the other side.

https://streamable.com/vwm3bm

Beautifu, we're making progress! 👏🏻

The way I see it should behave,
- without power, needle is at a hard stop. This is the position at which the needle is pushed against, by means of a built in spring.
- with power, but without the float, needle is magnetically positioned to Empty (a gentle movement to right)
- with power and with the float sender, needle is magnetically positioned to anywhere in between Empty and Full.

So we've got to a scenario in which we're driving the magnetic field in such a way as if the float sender is indicating full tank. Now we need to find the empty tank condition. 😁

Gotta admit, I'm impressed with your presence. So you sleep at all? 😂

In this test, do you remember how far to the right it went?

Agreed on the principle, except that it positions itself to the Left (empty indeed). My understanding (I could be wrong) was that under full power (12V plugged in), the gauge went all the way to the right. Introducing the fuel sender added a resistance, which increased as the float went down.

Full tank -> high float -> low resistance -> more current to gauge -> needle on full
Empty tank -> low float -> increased resistance -> less current to the gauge -> needle on empty
Both the sender and gauge being calibrated so that in between, current fed to the gauge matched more or less the fuel level in the tank with the needle position.

Would that jive with the needle pushing towards Empty while revving when the gauge polarities are reversed? I stayed up a little late last night (not really for a Friday night), and woke up this morning at a reasonable time. I just had the battery setup and gauge handy and was doing stuff around the house, so I was able to get on it rapidly when instructions were given.

Thanks for making my problem your problem for the past 48 hours , and providing all the sketches. I'm not very fluent in electrolese, so the pics certainly helped a lot.

So, what's next? Should I go ahead and plug another wire for the sender to see if it still works in the frame under AC power? I guess that's the other factor missing here, the power coming as AC instead of DC....

AC conundrum is not that confusing at all. The diode is in the wrong place. We will have to have it removed eventually, and put it where it belongs.

You are welcome, I'm enjoying the process. 👏🏻😁

Almost to the end, but quite slowly...🐢

https://streamable.com/489op9

Ok, so lower current creates less magnetic field, resulting in slower movement of the needle, and inability to overcome the spring torque, past certain level.

I was really hoping it will settle on Empty 🙄
Back to the drawing board... If I was designing authority, I would mandate for the indicator to show empty by default. This way, if float fails, then it would always show empty, rather than full. But, I'm not saying that opposite is impossible, and that's what we're trying to figure out with your gauge.

At the sender unit, there's the wire for the gauge, a wire leading to the reserve light, and a wire to ground.

I've never experienced a gauge breaking down while correctly wired, but I've never heard of a gauge showing full by default when defective. They either stay stuck on empty, or dance around like ravers on speed. Is it possible that when or if they act up, the current gets dumped to ground and bypasses the gauge? That could be the safety mechanism to prevent false Full readings.

Common failure modes for magnetic coil gauges are:
- short in the windings, affecting their accuracy or
- open coil due to a wire break (usually due to overheating as a result of excess current)

I've been re-reading the thread, and I see at some point you had this gauge working. Admittedly, it didn't really respond well, but I attribute that to either faulty float, or incorrectly dimensioned float (incompatible resistance). But we're not at that stage yet, we're struggling to get it working in basic (default) conditions. I hope the gauge hasn't failed when it was used in the interim...

I'm drawing on paper everything we tested so far, and it's inconclusive at the moment...

So, run this by me again. It's a Motovespa, 200, without a battery?

It's basically an early P200 without indicators, a simplified wiring harness with no battery, no electric start, all AC, but with an EFL fuel gauge sender installed, and a PT Danmotor headset from an Exclusive 2 with its speedo installed. I made the loom as simple as possible (lights on by default etc) but complete enough to accommodate the lights in the speedo; and apart for this gauge issue, all works.

You mentioned I'd managed to get it to work previously; rigged to a battery it did seem to respond to float movements at one point, but so did other gauges at different moments, and inconsistently. Other issue at one point was the sender's windings were broken and that gave bad readings as well. The current sender is brand new. The best I've been able to get the gauge to work in situ and AC powered has been the current compromise: accurate at idle but low with revs.

Here's a diagram that shows how the headset is wired with the principal components it interacts with:

Well, that plan is a dud. Firstly, my LML's gauge is a 5Vdc, air coil instrument (no spring). It also has a 7805 voltage stabiliser, a 470uF capacitor, 28ohm wire resistor and a generic 1N400x diode. Also, because it was floating in water for years, it's all corroded. It's not worth spending time on, to learn anything about your case.

Back to square one...

According to this schematic, the fuel gauge is fed by two wires only. And colours match the designators at the back of the instrument case. Two wires are Float (indirect Gnd) and +12Vac. Third wire is reserve light (indirect Gnd). Where did this schematic come from?

I was going to mention that: please don't mess up your speedo on my account! But I guess you know what state it's in on the inside now

I was wondering about either adding a capacitor to have a battery-like current delivery, or another gizmo to limit the input to say 10V or something in that range, which is what seems to be delivered at idle. But none of that should really be necessary if the gauge was made to work as is; which I'm guessing should be the case because PX/T5 gauges don't have all that equipment either, and they've worked for me on a battery-less scooter before...

I made the schematic from my own harness, and pics of the actual parts on the scooter . The gauge has a black wire running between the green and the red/green as well, which is the ground we've been working with.

The reserve light is totally separate, with the brown and red/blue wires feeding it. It is fed 12V directly (brown), goes to the sender (red/blue), then to ground on the sender side.
Wires tend to change color on either side of connectors, but the schematic is accurate: it has been gone over many times

I couldn't mess it up worse than it already is, so no worries there. 😁

But, it does make me wonder if your instrument is actually meant to be powered by 12V. We know it's a DC instrument already, but at what working voltage, that's the question. We've been testing it with 6Vdc, but just look at mine, it's 5Vdc unit. And yours is quite responsive at 6Vdc. Let's not forget, fuel gauge is not a fast responding instrument.

In order to limit the instrument's supply, you need to use a voltage stabilizer. These are simple to wire up, and are pretty robust. One diode, couple of capacitors and job done. But, they require some headroom. For an example, if you wanted 10Vdc, you'd need to supply the stabilizer with 12Vdc. You can't get stable 10Vdc, if 10V is all you have. You could get stable 5Vdc, 6Vdc and maybe 8Vdc (but it's a stretch), from what you have available at idle.

Edit: Ah, ok, got it.

One more question re. the gauge itself. I would say by looking at one of the earlier photos, that the metal case of the gauge is not grounded. Can you confirm if any of the posts are electrically connected to the case?

Unfortunately, I don't know much about what kind of current damper might have been used on the Indonesian bike it came from. Pretty sure the stator delivers 12V, but that's about it.

I get the conundrum with using a stabilizer. I guess the risk would be transferring the issue from bad readings at revs to bad readings at idle. Might as well stick with my current issue if nothing else: I'm used to it now

Regarding your question about the posts, I took a good look at it and there seems to be nothing in place to assure a purposeful connection between gauge body and metal case. The body itself, white plastic, feeds through the speedo bottom and the screws screw into the metal "nuts" inside the posts, which are in turn in contact with the winding ends. The metal case is far enough away from the gauge body to assume even the thin wires incidentally running outside the body have no chance of ever coming in contact with it. Hope this helps.

Some test results: ( post | batt. wire | needle direction | post | batt. wire )

1+ ---> 4-
1+ ---> 3-
1+ ---> 2-
2+ <--- 3-
2+ <--- 4-
2+ <--- 1-
3+ ---> 2-
3+ <--- 1-
4+ ---> 2-
4+ <--- 1-

Let's try this:

The idea above is that if the float is connected to post 2, and as the fuel level changes, so will the amount of current being shunted to GND through it. This will change the combined magnetic field and the needle will follow.

But, there may be some difficulties, mainly the resistance of the float potentiometer. Can we have the float for testing? I presume it's in the tank. 😁

Hi Rikko,

Ground to both 1 and 2 and power to 4 makes the needle go right.

I could pull the sender from the tank, but that wouldn't give us actual working conditions (aka AC power from the stator). Mentioning this because the current setup was decided upon after tests with 12V battery seemed to work - only to see the needle move with the revs once powered by the stator...

Is there any risk to the windings in plugging the gauge back into its wiring with 1 and 2 reversed and the sender connected, and starting the engine?

Should have specified, apologies, I actually wanted you to see that when you engage the GND on post 2, you see the needle move from Full (far right) to Empty (far left). And opposite, when you remove the GND from post 2, you see the needle go back to Full. Then adding a potentiometer (float) for test makes sense.

I wanted to get the gauge to work on DC only, because that will be the final working condition. Once we have it working on a bench (DC), we can move to creating a AC/DC rectifier/filter suitable for use in real condition (AC supply from the generator).

I'm confused. The needle is by default on the left. During these tests, when the needle moved to the right (full), it went from left to right under power, then went back to its default position (empty, on the left) when I removed either wire.

Likewise, when it went to the left, it was actually pulled slightly to the left under power, and was released back to its position when power was cut.

https://streamable.com/rowy3i

https://streamable.com/iza2ds

So we've found two conditions at which the gauge behaves as expected, more or less. And there's a single action required to operate between the two. In theory we have the three aspects needed for the gauge to operate. We just need to confirm it with a test.

Test actually shows that under power (simulating production environment), we can alter the magnetic field. We do this by creating additional contact on post 2. Ideally, it would be a gradual change, and we will try that with a float. But only after we confirm that switch-on/off of post 2 creates desired motion.

It's not the same as switching the power off. Without power, the spring is the driving force behind the needle. With power, it's the magnetic force.

Hope this explains it a bit better. 👍🏻

Ah ok, still foggy but I see the conflict between spring action and magnetic field.

However, regarding the requested tests, connecting positive to 4 and putting 1 to ground doesn't make the needle go right; it moves slightly to the left. (+4 -2 would move the needle to the right).

I can probably get the float into the mix by adding a wire from the sender feed and putting it in contact with the relevant posts

You could, but don't forget, float is grounded too. So if you plan to use the float while it's inside the tank, you need to make a common ground between the two. So additional wire from post 1 to float ground (perhaps tank body). Also, you need the ability to move the float in its entire range.

In my opinion, it's more conclusive if you do the test on the bench, meaning taking the float out.

Ok, but if I pull the float out of the tank (easily done), what ground does its ground need to be linked to? Can I connect it to anything iron-based at home or does it need to be linked to the gauge or the battery - ?

I'm bored a bit, so here's some acts of being smart... 😁 Proper answer is at the end, who wants to skip, please do.

The term ground, comes from the land electrical distribution. Home electrical infrastructure does contain ground, as a separate conductor which as the same suggests, goes to ground (also electrically called earth). It's designed as a safety feature for electrical appliances. Ships have ground as well, albeit through sacrificial material imprinted on the outside of the hull, which is in contact with water. In automotive industry though, the term ground was carried over to, but in effect it is the negative pole of the battery which is the ground. Negative of the battery is connected to the body, therefore the body itself acts as a ground.

So when we talk about the ground in Vespas, it is commonly referred to the scooter body, simply because it's made from steel (modern ones have a frame which acts as a ground). But in effect, it's a negative pole of the battery (if one exists), or the negative on the rectifier (if it's DC without a battery). If the Vespa is pure AC, then the body is neutral (AC current is both positive and negative), but we still call it ground.

When I talked about the float needing a ground, I was in fact trying to say that the float needs to be connected to both the negative of the battery (or GND on the gauge) and Post 2 of the gauge (make sure please that polarity is correct). Whether the float is physically in the tank or not, is irrelevant. But if it is in the tank, then effectively the gauge, makeshift battery and the float (tank and/or scooter body) need to be on the same potential, otherwise common.

Real answer now. 😁
There's no need to connect any part of the circuit to in-house ground.

Thanks for the rather complete explanation!

I need to use the scooter for an errand, but I'll pull the float out when I get back and get some tests done in the evening.

So this setup is what we're talking about? And moving the float on the sender should move the needle accordingly.

I aim to please 😁😁😁

Joking aside, I'm gonna say yes, based on the colours of the wires in your illustration. Although, because it's out of the tank, it won't matter if the polarity is swapped. 😁

But, what I need you to do before you connect it, is a dry measurement of the float resistance. Put the ohmmeter across float, and measure with the float arm in the empty position, and then with the float arm in the full position. We will want to know two things:
- total resistance
- which end is 0 ohm (full or empty)

I tried my best to get the float out of the tank, with no success. I remember it being tricky to get in/out, but the body being plastic, after a while I let it go and decided to conduct the tests with the float in the tank rather than have a broken float

Wires taped to the multimeter electrodes and plugged into the relevant plugs in the sender:

-Float pushed all the way down (empty), I get 308 Ω .

-Float pulled all the way up (full): 2.3 Ω .

"Moving float" test with this setup:



Successful. Push the float down, left goes the needle. Pull it up, right goes the needle

https://streamable.com/03papr

Fun-funking-tastic! Finally👏🏻

Ok now we need to play with calibration. Can you repeat the float test, but with the gauge face mounted? Let's try to establish how far off it is, and whether we can improve it a bit.