An LED Ignition Timing Light experiment

[TwoManyXS1Bs]

In this thread, I'm going to try to document the critera, goals, challenges, design and implementation of a permanently mounted (but totally removeable) LED Timing Light for my XS1B, as seen in this video:

I built a timing light for my Vega back in the mid-'80s. Mounted on front of left head, aimed at harmonic balancer degree marks, using hi-brightness red LED, driven by mosfet, triggered by wire wrapped around the #1 plug wire, like this:



This new/experimental version doesn't use the wrapped wire trigger, but reads the points signals instead.

 

[CBJamie]

Very inovative 2M. Now put a clear cover on it and you can check timing at will.

 

[TwoManyXS1Bs]

Y'know, that thought crossed my mind while reading some of the 'clear alternator cover threads'.
Maybe this'll renew interest in those covers...

 

[TwoManyXS1Bs]

Some of the problems encountered with the older 'wrapped wire' trigger are:
Ignition polarity, you'll either get a spike or a dip.
Destructive high voltage exposure with old plug wires and wet conditions, including morning dew.
Inconsistent trigger voltages due to poor ignition, fouling plugs.
Unless hidden well, just looks ugly.
You need a separate power supply wire.

By tapping into the grounding/points side of the coils, the unit could be self-powered and have a strong, reliable trigger signal.

I wanted this new version to:
Be as cheap/simple/small as possible.
Use up some of my 30 year old hobby stock.
Be completely removable. No irreversable alterations or damage to the original systems.
Take advantage of (repurpose) the black grounding wire going to the rotor brush block.
This would avoid removal of the left engine cover.
All LED installation mods can be done with just the rotor cover removed.

The challenge:
The signal from the points wire is hundreds of volts and has a 'ringing' characteristic of about 10,000 Hz.
The trick here will be twofold:
To extract a minimal, non-intrusive amount of current to charge up and power the unit.
To trigger on only the first rising voltage spike, which occurs at points opening.

Here's a trace of the XS1B points signal, at idle rpms, showing an initial voltage spike over 150 volts, followed by the characteristic "kettering ignition" ringing with a period of about 100 microseconds:

The other challenge: How to make this interesting to those members who have little interest in electronics, and would prefer to run away, screaming, arms flailing in the air...

 

[Downeaster]

Subscribed.

(Of course, it helps that I was a Navy Electronics Tech for 20+ years...)

 

[gggGary]

I can be a test dummy without the concrete barrier thingie!

 

[DogBunny]

Very cool project.
Two questions:
Why is your timing retarded? Mark should be at "F". I'm sure you have a good reason...
Second, what about electronic ignitions? I guess those will have to use the wire wrapped around the plug wire system?
I really look forward to seeing the plans/diagram and parts list.

 

[retiredgentleman]

2M .............................looks like an interesting and fun project

 

[TwoManyXS1Bs]

...I was a Navy Electronics Tech for 20+ years...

Great, Downeaster. I could use your input. critiquing, pointing out errors, ...etc.

I can be a test dummy without the concrete barrier thingie!

Okay, we'll dispense with the concrete barrier.

And use airbags filled with popcorn.

Have you done electronics hobbyist projects like this before?

 

[TwoManyXS1Bs]

Very cool project.
Two questions:
Why is your timing retarded? Mark should be at "F". I'm sure you have a good reason...

Haha, noticed that, didja?

Yeah, the timing matches my riding style....................retarded.

It's easier to start, stabler idle (no *cough*, pop, spit, stop).
Took out a bit of the perkiness, tamed it down a bit.
Running temps seem fine.

Second, what about electronic ignitions? I guess those will have to use the wire wrapped around the plug wire system?

Could be. Or, may be able to read the coil's negative lead.
If that's of great interest, maybe we could hang my scope on one of yours, see whut's there...

I really look forward to seeing the plans/diagram and parts list.

Thanx, DB. Coming right up.

2M .............................looks like an interesting and fun project

Thanx, RG. Jump in when you can...

 

[TwoManyXS1Bs]

After multiple iterations, tweaking and tuning, here's the original working schematic:

Update: This is the revised schematic and parts list. The only significant changes were to clarify component pinouts and voltages, and use more easily available components.

 

[TwoManyXS1Bs]

I tried a variety of High-Brightness LEDs, in one, two, then finally three gang.

The LED 'on' duration using 3 LEDs has been optimized to 50 microseconds, which is a fraction of a degree of crank rotation at idle, and 1 degree at 3600 rpm. Longer durations produce a visibly brighter image, but that's because of eyeball response. The LEDs hit full brightness in a matter of nanoseconds. So, a longer duration appears brighter, but begins to blur. Shorter duration is crisper, but dimmer.

The high-brightness white LEDs looked promising, but I couldn't get a crisp image. Found that white LEDs are actually blue LEDs, with a phosphor layer. This phosphorescence comes with a price, latency and persistance, producing fuzziness.

So, I looked into super-bright water-clear green LEDs. From http://www.SuperBrightLeds.com.

Here: https://www.superbrightleds.com/mor...ed-45-degree-viewing-angle-8000-mcd/271/1198/

Recommended as the eyeball has the strongest sensitivity to green. These RL5-G8045 work great:



In this circuit, I hit them hard with an initial current inrush of 40 milliamps, for 50 microseconds, then clip off.

 

[TwoManyXS1Bs]

To drive three of these LEDs in series, with 40 milliamps, takes close to 12 volts. This scope trace shot shows the 12 volt drive output to the LEDs, at a 50 microsecond duration. Once the drive voltage drops below about 6-7 volts, they are essentially turned 'off'. The trailing voltage on this trace simply bleeds off.

Yes, that's how I breadboard projects. Things sticking up in the air all over the place. Easier to swap parts, and hang test clips onto.

 

[TwoManyXS1Bs]

To select 'Right' and 'Left' ignitions, a mini SPDT on-off-on toggle switch is used.
Economical and available on eBay.



(Pausing for awhile for dinner, and to assemble more posts)...

 

[homer]

That is pretty damn cool!

 

[TwoManyXS1Bs]

That is pretty damn cool!

Thanx, homer. I wonder if anyone wants to make one of these things.

A big part of this challenge is the assembly of the project board. Since I was mostly experimenting, clanking around with multiple versions, I didn't take enough pictures to fully describe the final working version (since I was expecting it to be problematic, too).

The final version of the breadboarded project looked like this:

I wanted this thing to be as tiny as possible, with minimum wiring. So, I'm using some unorthodox tricks to compress it into a small form factor, like vertical/standing discrete components (resistors and diodes), and connecting (soldering) at the tops as well as the bottom, sort of a 3-D layout. The ground is also provided by a thin sheet of brass that gets sandwiched between the switch body and the grounded mount bracket, eliminating an external ground wire. The assembled/crunched version looks like this:

Enlarged to show detail.

 

[TwoManyXS1Bs]

The LED display plate is mounted to the stator housing, using already existing 4mm screw mounts.
The first two experimental plates used one, then two LEDs.

This is the final, 3-LED mount plate. 3 short pieces of 7/32" hobby brass tubing are cut to length, with 45° bologna cuts, and attached to the 0.016" sheet brass plate, aimed downward towards the sweep zone of the rotor's timing mark, one centered on the idle mark area, one on the full advance area, and one in between. The wider 45° illumination angle of these LEDs gives no dark zones. These short brass tubes hold the LEDs, and prevent side flashing into the eyes.

I had to tape over the emission side of the LEDs, then paint the topside with flat black paint to eliminate the blinding flash from the backside of the LEDs.

The insulated standoff is a 0.400" long piece of 3/8" phenolic/plastic, threaded completely through with a 4mm x 0.70 tap. A 4mm screw is cut as a short stud, epoxied into the standoff (about 4mm deep), then cross-drilled and staked with a 0.020" (1/2 mm) pin. Flats are filed on the top to accept a 5/16" (8mm) wrench.

 

[TwoManyXS1Bs]

This shows the 2-LED plate, mounted to the stator. It's thin enough to simply slip under the existing stuff without disturbing anything.

The black ground wire for the inner brush of the early alternator is repurposed as the flash/power wire for the LEDs. It's simply disconnected, bent over, and its lug fits over the insulated standoff. The screw that held this lug is returned to its original location in the brush block.

The (+) positive lead of the leftmost LED in the gang has a small lug soldered on it, and fits over the black wire's lug. The 4mm screw that was replaced by the standoff is now used to connect these two lugs. Its length needs to be checked, shortened if necessary, to clamp these two lugs, without bottoming out first.

The (-) negative lead of the rightmost LED is soldered to a small turned-up corner of the plate.

This shows the 2-LED plate after painting flat black.
(Just imagine the third LED sandwiched between the two.)

 

[TwoManyXS1Bs]

The other end of the alternator brush's black wire originally goes to the regulator connector, and (at least on my XS1B) doesn't connect to anything else. Best to confirm this on later models. It simply supplies the only ground connection to the regulator. So as not to damage the original harness, it's removed from the connector shell, and a short piece of shrink wrap is put over its lug.

A new black grounding wire is made to replace the original, one end with matching type connector fitted into the connector block, and a ring lug on the other end, which goes to the frame ground nearby.

The whole idea behind repurposing this black wire is to avoid removing a bunch of stuff and threading a new wire to the stator housing. Cleverness or laziness, your call. The idea came from the "three nylon screws" mod.

The picture shows a large diode in that position, which was another project explained in here:

http://www.xs650.com/threads/alternator-rotor-inductive-kickback.42605/

 

[TwoManyXS1Bs]

A right angle Z-bracket holds the unit. Two 6mm holes, and a rise of 0.8" positions the unit with plenty of clearance, above to the seat pan, and below for the unit and its wiring. Grounds also go to the mount bolt.

I wrapped the unit in electrical tape, for now, intending to make a better enclosure later.

The repurposed black wire plugs into the unit's output line.

I just used what wiring colors were immediately handy, declaring that bLack and bLue would be for Left, and Red and gReen would be for Right.

The two left/right input lines will plug into the harness connectors where the condensers originally plugged. Then, the condensers are plugged back into their respective pigtails.

Finished assembly. Ready to release smoke.

Later: More and better details on building the unit...