Auto Circuit Again - Oil Dummy Light

[Nishmaster]

Hi gurus,

My bro is doing a bit of his own DIY, putting together a snowmobile from a blank chassis with a bigger engine, more power, etc. The oil tank he has contains a float with a switch that activates a LED to indicate when the oil is low. However, the chassis he has is older and doesn't have this light or space in the harness. I'm going to build one in. The problem is that the power lead from the engine (according to the service manual) is 14.3v AC 10-12A on a single wire from a magneto on the flywheel, and then ground on the other wire. So, I came up with this:

I'm a total noob (well, not total maybe). SW1 is the oil float switch, D1 is half-wave rectification, C1 smoothing, D2 maybe a 12v zener for regulation (probably overkill here, but a pulsating LED from noisy power will annoy me to no end), and of course the resistor and LED. What size cap should I use for smoothing? My big worry is the 10-12A. Is that going to blow diodes, cause general havoc, etc.? Thanks folks.

Matt

 

[JohnRoberts]

You won't get amps through that circuit.. current will be defined by R1 and ohms law (voltage drop across R).

The cap needs to be connected to ground across the led- resistor- zener leg.
You probably don't need a zener at all, just make resistor a little bigger.

JR

 

[Nishmaster]

Of course.

Just had a number of "duh" moments there. Thanks JR.

Matt

 

[PRR]

> a pulsating LED from noisy power will annoy me to no end

You WANT that!!!!!

This is a LO OIL light, right? You never want to see it. But when it does come on you WANT to see it NOW!!!!! Before that oversize overprice engine welds itself shut.

And snomobiles run in sometimes bright sun, with highly refective ground (snow). You can't make an LED bright enough to FORCE attention in sun. Pulsation may attract attention where a steady glow may be overlooked long enough to melt babbit.

> My big worry is the 10-12A. Is that going to blow diodes, cause general havoc, etc.?

Same as the 20A in your walls. Plug in a high-resistance night-light, only 0.05A and 6 Watts will flow. Plug in a low-resistance heater, 2,000 Watts flows. Plug in a too-low-resistance load, 30 Amps tries to flow, walls burn from the inside unless a fuse goes first.

So don't use more than 12V*12A= 144 Watt lamp for your LO Oil indicator. 34W car headlight would be acceptable, and might even work with sun/snow glare. (But check the switch rating: it was probably scaled for a 0.1A-0.5A idiot lamp.)

> C1 smoothing, D2 maybe a 12v zener for regulation

Neither will do what you say. They need to be in shunt, not series.

Regulation be dammed. If the light is on, you stop the engine.

But anyway: you take 12V (beware: one-banger magneto regulation may be 10V-17V!), deduct ~~2V for LED drop, you gotta waste 10V. Put a reverse diode across the LED to damp the backswing of the AC. OR: use two LEDs back to back. Redundant, and 2 colors flickering 47Hz could be extra attention getting. Even your hind-brain knows that isn't a "reflection".

Pick an LED current: for this "EMERGENCY!" indicator, I'd aim high of the normal upper limit, typically 20mA, because half of it is wasted, and because this is a SEE ME NOW! lamp. 10V/20mA= 500 ohms, use 470 or 390 ohms. Check: dynamo soars to 17V on downhill over-rev, 17V-2V= 15V, 15V/390= 38.5mA. Idle, 10V dynamo, we have 8V across the resistor, ~20mA. Actual LED current is half of this, 10mA-19mA. Very safe for the LED.

Resistor dissipation is tricky. Assume the entire dynamo voltage, maybe 17V, across the resistor. 0.74 Watts, use at least 1 Watt. Actually, because snomobiles run in cold air, and the dry engine bearings will heat faster than the resistor, a 0.5W may out-live the engine.

 

[Sorr]

Pardon my ignorance in these matters, but dont you want the low oil lamp to indicate low oil before you start the engine?
Why dont you hook it across the battery?

 

[JohnRoberts]

Why kill the battery when off... engine doesn't burn oil while off, and if it leaks there will be a different indication to see.

JR

 

[Nishmaster]

Thanks PRR for the immense food for thought. I might go with a 555 timer and make the LED really obnoxious. I'd use a larger lamp, but it would look out of place, and I'm not sure the dash has appropriate space for it.

Sorr, the main reason for not hooking the light across the battery is that this snowmobile has no electric start, and hence no battery.

Should I be putting a cap to ground after the switch to eliminate bouncing of the float switch when oil is at a medium level but the vehicle is thrashing about?

EDIT: Of course that's pointless as the voltage has yet to be rectified there...

Matt

 

[PRR]

> dont you want the low oil lamp to indicate low oil before you start the engine?

Nice, not essential. The engine will run several minutes without oil without major damage. It only takes a second for the alternator to light the light, and the pilot should be looking for that light to be off.

Also: the very wise pilot will unscrew the cap and LOOK for oil before he starts. Maybe not every time, but after any long run or long summer idleness. Switches fail.

> Why dont you hook it across the battery?

There may be (isn't) no battery. Batteries can not be trusted in cold, which is when snowmobiles usually run. You have to have a hand/kick start. Batteries and start-motors are heavy, snow is weak, weight may be an issue. "Fun" snowmobiles may have battery-start for convenience; we sense that Matt's bro is building a Special which may be all-GO no-frills.

Magneto engines don't need a battery for spark or lights: my snow-thrower runs the headlight on magneto AC, flickering at low RPM. Cigarette lighter, cider-warmer will heat fine on AC. AC is the natural output of rotating electrical machines, DC means brush wear or fat rectifiers. Since 1959 rectifiers have been affordable... but if you don't need em, why buy em?

> eliminate bouncing of the float switch when oil is at a medium level but the vehicle is thrashing about?

Interesting point.

Go thrash it some and find out.

Worst-case: if flashing when thrashing, not staying ON, then it is just thrash-flash and may be safely ignored until you can invent a fix.

I'd suspect the switch is set low in the bottle, so you'd have to be getting near the end before it flashed. Then it is like an old Ford Windsor: It'll flash occasionally but stay mostly out, but flash more and more, before it stays on and bearings go dry. In the Ford, a hard stop while hot and a quart down would blip the light. 2 quarts down while hot, any slow-down would light the light. 3 quarts down, lamp stayed on at idle. But cruising, the oil still held pressure, so you could gingerly get back home for more oil.

Also, the elegant trick would be to use oil to damp the float's bounce. A large-area pivot will have considerable viscosity damping. The risk is that very cold or dirty oil could bind the pivot and hang the float high after oil goes low. If what you have was made for snowmobile use, someone already thought about this. If it came off a lawnmower, cold-temp and high-thrash issues were not considered.

I'd think that thrash cycling can only be a second or so, while engine disaster takes many seconds. So a 3-second delay between "switch On" and "lamp on", resettable by switch-off, would seem to ignore thrash and still give engine safety.

Where is this oil going? Simple 2-Stroke adds it to fuel so the tank is sure to go empty but at a predictable rate (200 miles per gallon oil). If the mixer is screwed, you know by the excess smoke. 4-Stroke usually keep their oil inside, mostly, and usually should not run out of oil between oil changes. But stuff happens (rock through sump) so you want to know. Some engines just splash, but high performance engines often have a pump. An oil pressure switch in the pump line is the usual way to monitor these systems. Unless it is an old Ford with pre-worn bearings, you will "always" have oil pressure (even in thrash) so a no-pressure situation is real bad, a good warning. In such systems, an oil level sensor just tells you "you will lose oil pressure soon, check it now".

> the voltage has yet to be rectified there...

AC-powered systems violate all our usual assumptions. "Use a 555" is not a simple answer. Yes you can use "DC" parts on AC: an LED which is lit 40% of the time on 47Hz-200Hz AC gives a perfectly clear "ON!". At another extreme, megaWatt oscillators run raw AC to melt metal: each tube is off half the time, so you load it for double power, and run 2 tubes on 2 pots of metal to balance the AC line. (If you may not have two pots ready, you splurge for a rectifier, but no filter.) But it is Think Different.

 

[JohnRoberts]

Yup 4 stroke (autos) usually had oil pressure not oil level indicators, but I did have an oil level light in my '86 stang. I think they had a design problem with the oil rings in that particular block/piston/ring combination (5L) and added the oil level light to deal with the fact the customers don't routinely check their dip-stick between oil changes.

The customer is always right, even when they don't check the oil and if the motor burns a quart every 1-2k miles you better keep track of oil level.

JR

 

[Nishmaster]

> There may be (isn't) no battery. Batteries can not be trusted in cold, which is when snowmobiles usually run. You have to have a hand/kick start. Batteries and start-motors are heavy, snow is weak, weight may be an issue. "Fun" snowmobiles may have battery-start for convenience; we sense that Matt's bro is building a Special which may be all-GO no-frills.

This is entirely correct. More weight means less acceleration, which in this project is A Bad Thing. A battery and starter motor would increase weight by at least 10-15 percent.

> Also, the elegant trick would be to use oil to damp the float's bounce. A large-area pivot will have considerable viscosity damping.

No such luck here, the float is a simple hollow plastic cylinder on a stick with metal contacts at the bottom.

> I'd think that thrash cycling can only be a second or so, while engine disaster takes many seconds. So a 3-second delay between "switch On" and "lamp on", resettable by switch-off, would seem to ignore thrash and still give engine safety.

That safely ignores thrash for the first three seconds, but since thrash is more or less constant on a snowmobile, I feel that approach would cause reset either before the lamp lights or at very best right after. While the presence of oil in the tank exacerbates the thrashing, tests with the tank empty show almost just as much thrash, thereby rendering that approach somewhat suboptimal.

> Where is this oil going?

In these advanced snowmobile two-strokes, oil is taken right from the tank and injected directly into the cylinders. So while oil consumption itself is constant for a given RPM and throttle position, it is not directly predictable, although somewhat estimable with experience.

> Unless it is an old Ford with pre-worn bearings, you will "always" have oil pressure (even in thrash)

Oil pressure, as indicated above, is not really applicable here, and constant until the point of no oil.

Here's what I came up with. It's a simple timer circuit that flashes the LED at a predictable interval, I think a little over a second per cycle. I really am unsure what to do with the float thrash issue. I can't seem to put a cap anywhere to smooth the thrash that doesn't interfere with the RC timing network. Ideally and conceptually, I feel like a cap to provide temporary current (.5 to 1 second) when the switch is out of the circuit would do the trick, but I can't seem to figure out where to stick it. Would moving the switch to before the rectifier diode help any?

 

[Guest]

Since it is flashing and always cold outside you may increase current through a LED up to 100 mA.

 

[PRR]

> thrash is more or less constant on a snowmobile

Sure....

> that approach would cause reset either before the lamp lights or at very best right after. ... thereby rendering that approach somewhat suboptimal.

You could just say "bad". Even my boss doesn't render somewhat suboptimal. (HIS boss does....)

But I think you are wrong.

What happens for medium oil level and high thrash? Float switch is on-off-on-off-on-off-on-off.... How long can it be "on"? How long can you defy gravity? Not for very long. The common word "thrash" seems to suggest a second or less. An ocean liner "sloshes" with a period of several seconds due to vastly greater size and therefore delayed recovery to average.

Part of it is simple spring-rate. Car spring/mass is adjusted to a bit slower than 1Hz. Without showing all my ignorance, I suspect snowmobiles are sprung stiffer. Of course for some apps they are un-spring, or nearly so. Un-iced snow provides some yield.

OK, I figured a long-period "thrash". Gently rolling round-top hills. (I think skiers have a word.) Around the bottom you have high Gee. Over the top you have a low Gee. But there is a practical tradeoff between time and Gee. A 1,000 long foot round-top may take 10 seconds, but even with 100 foot rise/fall it will not be high Gee. A 10-foot crest may peak to zero Gee, but for less than a second before you return to earth. (If you find yourself in zero Gee for too long, you gonna hit really hard.)

What happens when oil is low? on-on-on-off-on-on-on-off-on-on-on-off-on... Since you still have some oil, the float is going to signal "off" for some part of every thrash cycle. Each thrash cycle is less than a second.

An engine which has been well-oiled will not die from a few seconds without oil. (I could be wrong for mist-oiled 2-strokes, and for engines working at insane speeds. I know a car V-8 will run several minutes at 2,000rpm before wipe-out; 6,000rpm on the same engine would be a lot shorter grace period.)

When you get an "On", hold that, but wait 3 seconds. If you get any "off" within 3 seconds, cancel the "on" (don't light). But if you have on-on-on-on-on-on for 3 seconds non-stop, it isn't a bump. It is a low float for sure.

We must also ask what the pilot will do when the light comes on. If he knows the float bottoms when the tank is 1/8th full, he's going to plot a 25 mile course to an oil supply and keep going. Of course that assumes normal oil consumption rate; if a rock tore the tank it will go from 1/8 to zero "now!". Or not a rock (which might be noticed) but a loose clamp (never noticed until it falls off).

There are other angles, though bro may know them. There is always 16:1 fuel/oil mix. If you have fuel, you have oil; no fuel you can't wipe bearings. And nobody can follow a 16:1 smokemobile. Modern oils allow 32:1, even 64:1(?) mix for the entire life of a strong engine. The fancy eco-friendly trick is that pump which adjusts for small oiling in light work and more oiling in heavy work, but now oil consumption is hard to predict. So maybe a 200:1 tip of good oil in the gas would allow just sensing oil pressure. If the smart pump goes to zero pressure, light the light to tell the pilot to back down to easy running and short route to the pits, the 200:1 fuelil may keep the engine alive at 1/10th full power.

 

[PRR]

> Since it is flashing and always cold outside

Summer is for working on the machine. After adding the supercharger, in July when wrenches don't freeze to fingers, bro is going to take a spin around the yard to check it out. They go on dirt and grass. It's hard on the skis so no long trips. But he may have some scrap skis just for summertime test runs. And the excitement of an engine job, after idle all spring, is a great time to forget to put oil in the machine, or fail to notice it dribbling behind. I would not want the LED to flash just once and die.

But this is a field adjustment. Some modern LEDs, tight beam at the eyes, may be fine at 50mA even in sun. Or real bad sun and oblivious driver may need much more. He'll have to test it, on summer concrete and re-check on snow.

That last diagram shows a 10A alternator, a diode, and a Zener. What is the Zener current? When input is over 12.6V, it wants to be infinite. If the Zener is less than ~150 Watts, Zener self resistance will allow many-many times the maximum safe Zener current/power. You need some resistance in there, so a 16V-18V over-run.

4,700uFd seems over-ample. I suppose the brain-pain to work out tolerable ripple for whatever RPM costs more than this cap costs.

The 555 can be a 3-second delay. Wire a pull-down resistor on RST, wire SW from RST to V+. (I think I have that right.) Fiddle R-C for several seconds.

Note that common electrolytic capacitance drops around freezing. I knew a guy with a Compaq Portable PC, kept it in an Ohio garage (tuned his car with it). He said it would not start right up on bitter cold days. He left it on and the leakage current warmed the caps, 20 minutes later it would start. That might be bad (suboptimal) on an oil light. If it does work, the delay/flash rate is likely to be visibly off from the room temperature condition. Put the prototype in the freezer overnight, and test real fast (or with long leads out the icebox door, or ice-trays on top of the rig).

There is a time-delay relay for home heat oil burners. If the motor starts and runs for 5 seconds and no flame is detected, it kicks-out. Works the same if the motor runs all day and the oil runs out: no flame, 5 seconds later the relay cuts-out. The old ones were real darn simple contraptions, yet exceedingly reliable. (Blow oil in a firebox too long, then get ignition, the furnace blows up.) I dunno very-cold performance, though a church may sit cold for 6 days. Mine was 120VAC, dunno if 12VAC is available or if Hz matters. Dunno if new ones are just chips and caps.

Oh: if thrash is REAL bad, the assumption that there is an "oil level" fails. Consider the crankcase of a radial engine. Oil falls to bottom and hits bottom piston rising at 20Gees. All oil in crankcase is always moving as a mist. There is no "sump", no "sump pickup". If your thrash is of that magnitude (bad), then the float won't float on a 20% oil 80% air mist. It seems unlikely for any sane human to endure that level of thrash... but we are not all entirely sane.

Oh: the radial engines have a few places where oil is not instantly flung-away. These are drilled and plumbed to an oil-air pump. Depending on attitude, most of the lines may be flowing mostly air. The pump feeds an air/oil separator, then the main tank which also lets air come off, before it flows beck to the bearings.

 

[Guest]

One more option: photo flash from a disposable camera will be visible in any conditions, if you are afraid for a LED to run it with higher than max average allowed current.

 

[JohnRoberts]

While it would be useful to know how the float switch works in the margin, why stop at just an indicator light?

Have the light flash if level momentary indicates low while thrashing. If low oil light stays on solidly, let it kill the ignition, or even better crank in a rev limiter so you can still limp home.

It's possible to ignore a light, the rev limiter will surely get your attention.

JR

PS: I used to have an ignition on my old car with built in rev limiter that was programmed with external plug-ins. I liked that I could swap in a rev limit at 2,000 RPM if I ever had to leave it somewhere for service. I hate having my car wrenched on by others. Only slightly less than doing it myself.

Last time I had my current car worked on (starter motor failed) they couldn't find my keys when I came to pick it up. Some guy in the parts dept had the keys to my car... I wonder why?