TheSamba.com :: View topic

Sat Aug 23, 2008 9:06 pm

I know nothing about electrical...
When I turn on my headlights, I get a strong electrical smell in the cabin. Any ideas? 1970 Deluxe Sunroof, 1600 DP

Thanks in advance

Sat Aug 23, 2008 9:10 pm

I'm not familiar with the bus wiring, but the vanagons run the power for the lights THROUGH the light switch. This has the effect of causing an overload when the switch gets dirty, burning up the switch.

The cure in the vanagon world is to use relays in order to run the power through the relays, which means the switch is only running the relays, reducing the load.

Best,

Lanval

Sat Aug 23, 2008 9:10 pm

An Electrical Smell first time I knew electricity smells!!!!

But do you mean the smell of the cables burning?

Sat Aug 23, 2008 9:18 pm

has it rained on your bus? maybe water got in to a relay?

Sat Aug 23, 2008 9:39 pm

no, it doesn't rain in CA in the summer

Sun Aug 24, 2008 1:10 am

Feel around all the wiring, switches, fuses, etc. until you find something real hot.

Thu Jan 21, 2010 11:40 pm

Maybe it is Sacramento? I have this too - it has happened twice and there is always a small sense of panic when that smell comes up. Both times it was with the headlights on. It goes away as soon as the lights are turned off. What is weird is that I know all the wiring is good as we went through the whole bus. On mine it has something to do with the gauge dimmer. I think that it may be that stock lights plus 4 VDO and a tach are too much for it but I can't figure why it does it one time and not another. I am wondering if it is the actual switch. I bought what I believed to be a well made switch but maybe it is not.

Fri Jan 22, 2010 11:39 am

It is common for the cement which holds the coiled resistance wire making up the panel light dimmer to fail with age. When this happens, turning the headlight switch knob wads up the resistance wire, causing it to short against the shaft, causing the smell (burning plastic / Bakelite). Turning the knob will be un-smooth (crunchy) and the panel lights may or may not work. Verify by inspection. Fix: replace the headlight switch.

Word to the wise: DO NOT CLEAN THE HEADLIGHT SWITCH WITH CONTACT CLEANER! It dissolves the cement holding the dimmer resistance wire in place, leading to switch failure. Ask me how I know...

Fri Jan 22, 2010 11:55 am

Telford - we bought NOS switches where we could find them. It seems to me that some had a small resistor on top where as others didn't. Do you know from experience which switch is least prone to failure? The original switch was in pieces when we bought the bus. There are various models of the switches and various places in the world where they are made. This is a 1977 bus. I am driving the stock gauges, 4 VDO and a VDO tach off the switch and my concern is that the load is too great.

Fri Jan 22, 2010 4:31 pm

SGKent wrote: Telford - we bought NOS switches where we could find them. It seems to me that some had a small resistor on top where as others didn't. Do you know from experience which switch is least prone to failure?
Unfortunately, I do not. My bus is a '71, and the headlight switch has a contact to drive a dash indicator when only the parking lights are on. [Yeah, it's a useless feature, but that's the way it came.] This particular switch is (apparently) NLA, and the substitutions are missing the extra terminal. As such, when I needed one, I found a good used one at the junkyard.

Never seen this 'resistor' component on a switch before. As I recall, the late bay switches (thru '79) are all the same, but I could be wrong on that.

Quote: The original switch was in pieces when we bought the bus. There are various models of the switches and various places in the world where they are made. This is a 1977 bus. I am driving the stock gauges, 4 VDO and a VDO tach off the switch and my concern is that the load is too great.
That's a valid concern.

Try the following circuit:

This circuit will power the additional gauge lights without significantly increasing the load on the stock dimmer circuit. It uses the property of a transistor whereas the transistor emitter voltage will track the voltage applied to its base junction (emitter follower) by allowing collector current to flow as needed. In short, it's a power amplifier.

Now in doing this, it will waste some power (caused by the voltage drop across it) as heat, so you need to mount the transistor on a heat sink (aluminum plate) with an insulating mica 'mounting kit' to insulate it electrically. Use a little silicone grease on the mica to facilitate heat transfer. If you can touch it without burning your finger, the heat sink is big enough. Worst case rule of thumb: the heat dissipated by the transistor can't exceed that of the panel lights themselves.

I used a 2N3055 transistor in a TO-220 type package (TIP3055), as it's very common, but any NPN power transistor will do as long as it can handle a couple of amps and, say, 30 watts or so. Shouldn't be critical at all.

In practice, you probably don't need the resistor. I put it in by habit.

Supply the +12 vdc power from the headlight switch tail light terminal. It's where the gauge light power initially comes from anyhow.

I assume from your posts that you have electronics knowledge, but if you need any more info or clarification, let me know.

Fri Jan 22, 2010 4:36 pm

maybe unrelated, but my bus and my parts bus both had the rheostat or resistor or whatever that coiled contact is for dimming the dash lights completely burnt out and gone.

Now it either dash lights on, or off. I have to crank it to the max to get them "on"

Fri Jan 22, 2010 4:42 pm

SGKent wrote: It seems to me that some had a small resistor on top where as others didn't......

Some hazard switches have a small resistor to make the knob glow with the dash lights, sure it was a headlight switch?

Fri Jan 22, 2010 4:48 pm

[Rustbus] Yup. After the resistance wire all burns up and disappears, that's the remaining operating mode: when you turn the lights all the way up, the rotating contact touches a fixed metal contact which in turn powers the lights.

Fri Jan 22, 2010 4:54 pm

Telford - I can build that circuit easy with parts from Frys and mount the heat sink somewhere logical and dry.

I would assume that the 10,000 ohm resistor is just a shunt to ground so the circuit has some minimal load.

I guess one could even cheaply build a fused board with two of these on it in parallel and drive the stock gauges off one and the VDO off the other, with a pot in the second circuit to trim the VDO lights even with the stock gauges. If so could you guess at a potentiometer or resistors to place into the circuit that woud drop the voltage output so the lights could be trimmed to match the stock gauges.

Fri Jan 22, 2010 5:21 pm

SGKent wrote: Telford - I can build that circuit easy with parts from Frys and mount the heat sink somewhere logical and dry.
That'll work.

Quote: I would assume that the 10,000 ohm resistor is just a shunt to ground so the circuit has some minimal load.
Actually, it pulls the transistor base to ground if the wire to the stock dimmer is disconnected. In this application, you don't need it. No undesirable side effects.

Quote: I guess one could even cheaply build a fused board with two of these on it in parallel and drive the stock gauges off one and the VDO off the other, with a pot in the second circuit to trim the VDO lights even with the stock gauges. If so could you guess at a potentiometer or resistors to place into the circuit that would drop the voltage output so the lights could be trimmed to match the stock gauges.
No, you need the stock panel lights connected to load the headlight switch dimmer properly. Otherwise, the low resistance of the stock dimmer will be insignificant to the transistor, and you'll get full brightness all the time.

On the other hand, your idea about adding a pot to the input of the transistor to allow setting the added gauge lights could work nicely. And both light sets should track each other as you dim them. I'd add a second transistor to drive the first to really reduce the base current. Thus, a low wattage pot could be used as the adjustment.

Fri Jan 22, 2010 5:37 pm

Telfor - looking forward to it. I will drive the stock gauge lights off the switch and the VDO off the new circuit - correct?

One other item of interest. In my readings I discovered that on FI the primary pulse from the coil is essentially turned into a square wave that sets the pulse length in the fuel injection system on a late bay. Other factors such as temperature and load are used to lengthen that pulse and then the combined signal is amplified and sent to the injectors. I have a little tiny concern that the VDO tachometer could cause some interference with the fuel injection but I am using a diode right now at the coil because nothing would settle the tach down except that. I am bouncing ideas in my head how to get the signal to the tach without any risk of interfering with the inital timing pulse going to the ECU. I think the diode is adequate but not sure. Your thoughts.

Fri Jan 22, 2010 6:25 pm

SGKent wrote: Telford - looking forward to it.
[Edit] Built a prototype of the revised circuit above. Seems to work as expected. If you build one, feedback would be appreciated.

Quote: I will drive the stock gauge lights off the switch and the VDO off the new circuit - correct?
Correct.

Quote: One other item of interest. In my readings I discovered that on FI the primary pulse from the coil is essentially turned into a square wave that sets the pulse length in the fuel injection system on a late bay. Other factors such as temperature and load are used to lengthen that pulse and then the combined signal is amplified and sent to the injectors. I have a little tiny concern that the VDO tachometer could cause some interference with the fuel injection but I am using a diode right now at the coil because nothing would settle the tach down except that. I am bouncing ideas in my head how to get the signal to the tach without any risk of interfering with the inital timing pulse going to the ECU. I think the diode is adequate but not sure. Your thoughts.
Your FI operational description is basically correct.

On my bus, I was never able to get a VDO tach to work reliably, even with the diode fix, so I cobbed up a buffer circuit to isolate it from the ignition. Now it's steady as a rock. See http://www.thesamba.com/vw/forum/viewtopic.php?t=353088&start=20 for the circuit. This should eliminate any possible interaction with the FI, as well.

Sun Jan 24, 2010 1:36 pm

[bump]
Built and revised the buffer circuit; edited previous (up 3) post to show. This version works pretty good.

Sun Feb 28, 2010 1:54 pm

as we discussed - this is the circuit before it was revised. It does a great job of controlling the brightness of the aftermarket gauges so they can track at the same brilliance as the stock ones. I wired it so it comes off my aux fuse panel switched side so if the parking lights are on, the VDO gauges are off, thus preserving the battery even longer.

Will need to get it out on the open road at night to see the electronic reliabilty both of the componets and my assembly. I can see where someone could build these on a circuit board in mass and sell them to VDO as they would work on any car that is 12V negative ground.

Sun Feb 28, 2010 7:22 pm

How hot does the pass transistor get? Just curious...

Also (original circuit): it wasn't intended that the power to the unit be switched independently of the normal dash lights, as the dash light rheostat will still, in the parking lights only position, dump current into the transistors base and then into the aux dash lights - without the benefit of the transistor supplying additional power. The transistors probably won't care, but it could cause the adjustment pot to get hot / burn up.

Note that the revised circuit doesn't have this problem because of the 1K resistor in series with the first transistor's base connection. The original circuit doesn't have enough 'gain' to allow the addition of this resistor.