TheSamba.com :: View topic - Water level indicator LED always green / sensor troubleshooting

Sun Apr 19, 2009 9:04 am

The green led on the kitchen panel for the water tank level is always on regardless of the actual level in the tank. Is there a way to fix this or should I be looking at a new magnetic float rod assembly or panel?

Sun Apr 19, 2009 11:32 am

Err, ain't no float system used. There are sensors stuck in the side of the tank. This is a simple circuit. Do the other LEDs work? You may wish to try another panel. Double check your power. Ditto the other connections.

Sun Apr 19, 2009 11:35 am

Later Westys have a float based level sensor mounted from the top, not the probes in the side of the tank.

Mark

r39o wrote: Err, ain't no float system used. There are sensors stuck in the side of the tank. This is a simple circuit. Do the other LEDs work? You may wish to try another panel. Double check your power. Ditto the other connections.

Sun Apr 19, 2009 12:30 pm

Open the right-hand top loading cabinet that's between the hanging closet and the stove unit. Unscrew the metal cover - this will reveal the top of the fresh water tank. You'll notice that there are wires connected to the tank cover. Remove the tank cover and you'll find a float system attached --- perhaps the float is stuck at the top?

Good luck!

Sun Apr 19, 2009 1:04 pm

The level switch works of 2 miniature reed switches and 2 resistors, 560K IIRC. The lower reed switch is a Normally closed switch, the upper reed switch is Normally open. When the tank is full, the mag float will close the upper reed switch and give a low resistance reading across the 2 wires to the level switch. As the level drops, the mag float slides down the tube, allowing the upper reed to open, this gives a resistance circ 560K. As the level falls further, the mag operates the lower (NC) switch, which opens and gives a combined resistance of circ 1120K. (the 2 resistors are wired in parallel to the reed switches, so that when the switch is open, the resistor comes into play!)
The display panel senses the change in resistance and operates the respective LED on the panel.

Sun Apr 19, 2009 1:36 pm

My '91 has the float probe suspended into the water from the top of the tank. I'll open up the tank and see what's going on, so thanks for the descriptions. It's possible the probe has gone bad, so I'm wondering if there's a source for a replacement?

Sun Apr 19, 2009 10:40 pm

singler3360 wrote: My '91 has the float probe suspended into the water from the top of the tank. I'll open up the tank and see what's going on, so thanks for the descriptions. It's possible the probe has gone bad, so I'm wondering if there's a source for a replacement?

maybe here?

http://www.busdepot.com/view.jsp?model=43&category=40&group=97&prodgroup=923

Mon Apr 20, 2009 8:48 pm

madspaniard wrote:
maybe here?

http://www.busdepot.com/view.jsp?model=43&category=40&group=97&prodgroup=923

Thanks for the link MadSpaniard. Wow, that is spendy! Anyone know of any alternatives?

I played around with the float and different water levels in the tank via the cabinet access and the light remains on GREEN regardless. Should I look at the panel?

Tue Apr 21, 2009 6:43 am

When you slide the float up & down the tube, listen for 2 very tiny "click"s, one at each end of the tube - this is the reed switches operating. Quick way to prove the workings of the led's would be to get 2x 560k resistors and strap then in series across the black & green wires to the tank gauge. this should light the red LED. Then remove one resistor and check again - this should light the amber LED. If this happens, there's nothing wrong with the indicator panel, it's just your level gauge which is goosed.

Tue Apr 21, 2009 6:50 am

singler3360 wrote: madspaniard wrote:
maybe here?

http://www.busdepot.com/view.jsp?model=43&category=40&group=97&prodgroup=923

Thanks for the link MadSpaniard. Wow, that is spendy! Anyone know of any alternatives?

$79.00 spendy? to fix it right?
Alternatives are why some of these vehicals are in the condition they are in now.

Tue Apr 21, 2009 8:41 am

badger7777 wrote: When you slide the float up & down the tube, listen for 2 very tiny "click"s, one at each end of the tube - this is the reed switches operating. Quick way to prove the workings of the led's would be to get 2x 560k resistors and strap then in series across the black & green wires to the tank gauge. this should light the red LED. Then remove one resistor and check again - this should light the amber LED. If this happens, there's nothing wrong with the indicator panel, it's just your level gauge which is goosed.

I'll try this. Thanks.

Tue Apr 21, 2009 9:25 am

singler3360 wrote: Thanks for the link MadSpaniard. Wow, that is spendy! Anyone know of any alternatives?

I heard that some people connect a clear plastic hose to the bottom of the water tank then route it outside of the cabinets vertically. That way you can always see the water level with a quick glance. I wouldn't do it because I would have to drill a hole in the tank, and probably in the cabinetry, but it may be an option for someone.

Thu Apr 23, 2009 6:57 pm

badger7777 wrote: When you slide the float up & down the tube, listen for 2 very tiny "click"s, one at each end of the tube - this is the reed switches operating. Quick way to prove the workings of the led's would be to get 2x 560k resistors and strap then in series across the black & green wires to the tank gauge. this should light the red LED. Then remove one resistor and check again - this should light the amber LED. If this happens, there's nothing wrong with the indicator panel, it's just your level gauge which is goosed.

I found 5.6K resistors and 560 Ohm at RS. Brought them both home and tried the above with no luck in getting red or amber to light. Just stays on green. I completely bypassed the wires to the float sensor in these tests. In other words, I unplugged the green and black male connectors from the clear connector array sitting on top of the water tank and wired in the resistors. The guy at RS wasn't sure they even carried a 560K resistor. Am I all wet here or could this indicate that the green wire is not working?

Fri Apr 24, 2009 1:16 am

The tank-side electrodes were sometimes a little flaky (thorough scrubbing and if necessary putting maybe a quarter- or half-teaspoon of salt in a full tank should help). So VW apparently asked for a more positive-acting sensor from Delta-Six, and that's the one with the internal reed switches and the magnetic float. It is meant to be electrically interchangeable with the tank electrodes and their associated little black box. Because of this, it's a high-resistance circuit, and yes, you really do have to use 560k or whatever the near value is to test it.

Unfortunately whoever designed that abortion should be skewered on one. It has two glaring weaknesses, and you can often tell which one got exercised by how the lights behave. If the light changes level after the panel has been running for a few minutes, you have gotten a drop or two of water (that's all it takes) through the seal where the wires enter the top. But if you have a solidly stuck light, you have a broken reed switch. The *reason* you have a broken reed switch is that the switches are soldered onto the long skinny board with no slack in the leads, so if you flex it even a little bit the wrong way, one or both of the reed switches will burst.

So if you're forced to use one of these monsters, A) seal up the top much better than it already is, and then keep water far away from it; and B) never never never bend the rod even a little. If you're scrubbing the tank out, remove the sender first.

Or convert to the side electrodes, which weren't so bad really.

Fri Apr 24, 2009 1:23 am

Aside from the sender getting flooded or busted, the most common problem with the LED panel is one or the other of the chips dying. The one on the left deals with the water level, and the one on the right is for the voltage lights. Both are LM324 Quad comparator chips, and Radio Shack used to sell them for 39 cents each. So when in doubt replace the chip.

Also resolder the pins of the switch every time you have the board out -- chances are it's trying to pull out of the board.

Couple odds and ends here about the panel: http://pws.prserv.net/synergy/Vanagon/LEDpanel.htm

Fri Apr 24, 2009 8:38 am

dbeierl wrote: Aside from the sender getting flooded or busted, the most common problem with the LED panel is one or the other of the chips dying. The one on the left deals with the water level, and the one on the right is for the voltage lights. Both are LM324 Quad comparator chips, and Radio Shack used to sell them for 39 cents each. So when in doubt replace the chip.

Also resolder the pins of the switch every time you have the board out -- chances are it's trying to pull out of the board.

Couple odds and ends here about the panel: <http://pws.prserv.net/synergy/Vanagon/LEDpanel.htm>

Looks like radio shack still has them:
$1.69

LM324 Quad Op Amp (14-Pin Dip)

Model: NLM324 | Catalog #: 276-1711

These amplifiers are designed to specifically to operate from a solitary supply over a wide range of...

Fri Apr 24, 2009 3:01 pm

dbeierl wrote: Aside from the sender getting flooded or busted, the most common problem with the LED panel is one or the other of the chips dying. The one on the left deals with the water level, and the one on the right is for the voltage lights. Both are LM324 Quad comparator chips, and Radio Shack used to sell them for 39 cents each. So when in doubt replace the chip.

Also resolder the pins of the switch every time you have the board out -- chances are it's trying to pull out of the board.

Couple odds and ends here about the panel: <http://pws.prserv.net/synergy/Vanagon/LEDpanel.htm>

Would a solid green light, or any light for that matter on the water level side of the board indicate the chip is working? If so, then this leaves the sender as the culprit and I'll probably have to bite the $80 bullet on a new sender/float unit. I don't have any experience with soldering on a circuit board, so I don't want the risk of having to replace the panel too, which may not even be available. Thanks for the posts.

Fri Apr 24, 2009 3:08 pm

I had the same problem when I bought my vanagon.

I bought a new sender from Bus Depot

Installs in less than 5 minutes

Has worked perfect ever since

:D

Fri Apr 24, 2009 5:21 pm

singler3360 wrote:
Would a solid green light, or any light for that matter on the water level side of the board indicate the chip is working?

Not necessarily. But it's easy to test. Get a 2-meg or 5-meg pot and wire the center to the green lead and the ends between +12 and grounds. By rotating the pot you should be able to get all LEDs to light. If so, then this leaves the sender as the culprit; if not it's extremely likely to be the chip.

Quote: I don't have any experience with soldering on a circuit board so I don't want the risk of having to replace the panel too, which may not even be available. Thanks for the posts.

This is a single-sided board, very easy to solder and desolder and not difficult to repair if necessary. The easy way to desolder a chip is to clip its leads off, then remove the leads one at a time. Some desoldering braid will be helpful for clearing the holes. The watchword in soldering is *clean* -- you need a bright shiny tinned iron to do good work. There are soldering tutorials on the web. For this work a 20-25 watt iron should be more than enough.

Tue May 12, 2009 10:03 pm

Just received a replacement Fresh Water Tank Level Sensor, plugged it in AND...
the green water level indicator light still stays on. Darn.

The float on the new sensor does not float, sinks like a rock. I don't hear the clicking of the reed switches as the float unit is moved up and down the rod. I've traced wires, connections and fuses and everything else looks good. When the led panel switch is ON, the green wire at the water tank reads 0.3V.

So, is this a tale of two bad in-tank sensors? Should I have the vendor send me a third?