mechanical turn signal blinker. 4

[StephenVB]

I'm curious about how a (not so) simple electro-mechanical turn signal auto blinker works.

The simple and incorrect answer, given by the how does it work sites, is a heater on a bi-metallic strip. This doesn't explain the blinker properties.

The blinker is in series with the load. A simple bi-metalic strip would have a shorter on time with more load (more current) and once open it would cool at a fixed rate so the off time would not depend on the load.

This would also delay the turn signal lamps while the bi-metal strip first heated up.

In fact the turn lamps are initially on and have shorter on and off times with less current (a single bulb instead of two).

Does anyone know how this actually works?

Thanks

Stephen Van Buskirk

A human being is a computer's way of making another computer. Yes, we are their sex organs.--Solomon Short (aka David Gerrold, US science-fiction author)

 

[itsmoked]

I'm going to say the flasher units flash faster running fewer bulbs because they heat up slower. This makes them just creep up to the open temperature instead of a massive I²R blast that overshoots the opening point dramatically.

If the flasher normally runs, say, only two lamps and one opens, the heating coil will only be seeing 1/4th the power.

By virtue of them heating up slower and not blasting past the opening temp the coil also cools sooner to below the make-temp.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

I agree Keith. It's a cleverly designed interaction of thermal time constants - which create a lag or delay - and differing power inputs which determine the energy input to the device before the bi-metal 'catches up'. Operation with the full complement of lamps dumps energy in at a high rate and internal temperature significantly overshoots the bi-metal switch point and takes longer to recover; operation with a failed lamp causes an overshoot to a lesser degree and recovery is much faster.

Looking forward to Gunnar's report!


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

There will be some delay before we can start the investigation. Have to have a meal and also entertain grandson before the hard work begins. In the meantime, one could plan the experiment.

Test set up:

Variable DC supply.

Blink relay.

Adjustable power resistor to simulate different lamp loads (this will not simulate inrush, will find and use real bulbs if necessary).

Shunt resistor 1 ohm (lower part of load, actually) to measure load current.

Temperature transducer (miniature NTC thermistor fastened to representative part of heated member) to see temperature vs time.

Recorder.

Recordings will be made with different DC voltages and different loads and the relation between U, P, ON and OFF time, blink frequency plotted with U as the independent variable with one set of curves for each load setting.


If lamps available, the light output from them will also be recorded with an uncalibrated photodiode pick-up.

If lamps available, the UxI product (lamp watts) will also be recorded.

Anything else we should consider? Or any comments as to the feasability? And, above all, who will pay for this investigation? Remember that it is a Sunday, with double hourly rate.






Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Warpspeed]

As this is an altruistic Sunday project for the good of all mankind, a double beer ration might be in order.

 

[Skogsgurra]

That's a VERY GOOD IDEA! But that may delay the project somewhat. The beers are imminent now.

I done the initial measurements on a blinking device (UP-552.12V) and I must say that it doesn't behave at all as I thought or remembered.

First: The lamps are not at all on for the first second. Instead, there is a weak current heating the switch so that the lamps go on after a little more than 1 second.

Second: There doesn't seem to be any big difference in blinking frequency regardless of running with three, two or just one 21 W 12 V lamp (yes, borrowed a complete setup with lamps in reflector housing).

I will make the recordings available as soon as I have finished the beers and put them together in a pdf document.

Stay tuned!

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[MacGyverS2000]

Gunnar,

If you open the packages up, the heating rate will be different due to now-flowing ambient air.

You need to reseal in transparent aluminum casing for a proper test...




Dan - Owner

http://www.Hi-TecDesigns.com

 

[cranky108]

Strangly my car flashers won't flash if one of the main bulbs are burned out (I don't know what happens if side marker light are burned out).
And also when I replaced the rear lights with LED's, they also did not flash (The LED's are now somewhere else).

The emergecy flashers don't act that way. They will flash no matter what.

Strangly all these light bulbs have two filiments, and just the one burns out.

 

[Skogsgurra]

Yes Dan, I noticed that. I can even stop it from working at all by blowing gently on the filament.


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Warpspeed]

Gunnar, is that by any chance an infernal "Lucas" device, designed by the prince of darkness ??

Then any type of behavior is possible, absolutely anything.

 

[itsmoked]

Amen to that!

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MacGyverS2000]

It's Phil, the Prince of Low Light! Darn you to Heck!

Dan - Owner

http://www.Hi-TecDesigns.com

 

[Skogsgurra]

Not Lucas. I really don't know what brand it is.

Result can be found on

http://gke.org/arcus/

Look for document named "MEASUREMENTS ON AUTOMOTIVE..."

Now, hunting for the older type. The one that is ON from start and where blink frequency changes when load changes.

BTW: Any good explanation why heating current stays constant as filament resistance decreases during OFF period??

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[DRWeig]

Awesome, Gunnar!

I'm a little too frazzled (Monday morning headaches) to think straight about the constant-current issue, but your report's comment about the heater resistance rising as the lamp filaments' resistance falls certainly makes intuitive sense.

While you await an older flasher, I'll just throw out that my 1972 Pontiac Bonneville would just about double its frequency when I plugged in dad's boat trailer (approximately doubling the lighting load). However, it would simply stop on a burned-out lamp. This leads me to believe it had a single bimetal element perhaps.

Perhaps someone can countribute a make, model, and year of a vehicle whose blinkers sped up on lamp outage? That would help narrow the search.

Double good on ya,

Goober Dave

 

[btrueblood]

Skogs,

I am surmising that insufficient quantities of beer were consumed to arrive at conclusive answers. Recommend larger quantities be laid in prior to next weekend.

 

[VEBill]

Report: "So the sum of ON and OFF period stays relatively constant when lamps are switched from three to one. This is contrary to common belief and makes us think that there has been a change in technology sometimes during the last decades. This remains unanswered until we have found an older type blinker relay and made the corresponding measurements."

Good work. Mark me down as unsurprised.


Regarding "older"...

It's been my assumption that someone, way back in the depths of time, invented the bimetallic flasher of the perfectly simple sort you just tested. And I assumed that "later" someone figured out how to make a **NEWER** version that includes the 'operator feedback' about lamps being burned out. It would be illogical for this sequence to be reversed.

Obviously not all models would have included the extra feature even if it only cost a penny. My 1970s Chryslers didn't as far as I can recall. But that doesn't imply that it hadn't been invented even in the 1950s.

It should be a simple matter to order up an OEM flasher module, new or used, for any given car (even a 1962 Morris 1000) to check your recollection.

But once you get into the era of electronic modules (early 1970s ? for SOME models), there's no mystery.

 

[StephenVB]

Thank you all.