LED Timing Light


#21

OK with that working, now this should get the trigger working correctly I think:

The 560K/ 330K (R3/R4) voltage divider should provide around 4.4 volts with no voltage from the coil. The 555 trigger voltage should be 1/3 of 12V or about 3.6V so if the coil manages to produce at least -1 V (and the larger value resistors should let it do that I hope, if not we can increase the size of R3 and R4) that should trigger the 555 and you should be able to stretch the pulse so you get both enough light without overloading the LEDs and enough time on the output pulse to ignore the second trigger you were seeing. You will need to adjust the R2/C3 values to get the appropriate length of pulse for what you need,

Peter


#22

I’m confused about the junctions - I know you don’t use FZ much -, but is the link from R4 to C1 jumping over the link from pin 2 to the pickup wire L1, or connected to it, ie a 4 point junction.


#23

You are right there should be a dot at the r4 /L1/C1 junction as R4 should connect to the trigger line. I missed that when I modified the drawing. The idea is when the coil is not active (due to no spark) the R3/R4 junction (and therefore the trigger) will be about a volt higher than the trigger level. The spark will hopefully cause at least a volt or 2 of change in the coil which will drive the trigger below the threshold and trigger the 555 to extend the pulse.

Peter


#24

Junctions are a bit tricky but you make them by holding down the ALT on the keyboard and clicking on a bendpoint, and that starts a trace that you drag to the pin. As usual grab the junction and move it and all the traces should move - this test doesn’t work for 4 junctions because only 3 move at a time -.

I corrected all the junction and made this circuit, but it’s inconsistent, ie sometimes it latches and needs a reset, sometimes it’s instantaneous. I still have the switch connected to GND pin2, but practically every button press is different. If I cut the wire where the label is, every button press is as expected.

Timing_light_new5.fzz (11.9 KB)


#25

What happens if you short the coil (i.e. remove L1/C1) leaving R3 connected to R4 and pin2? That should work as normal (if the resistor values are correct) as pin 2 should be about 4V which should be just greater than the trigger value. Shorting R4 with a jumper should then trigger the circuit as now. If it is inconsistent without the coil in there then the R3 value is likely too low and needs to be 680K instead of 560K (to raise the voltage at pin 2 above the trigger voltage). The latch up is strange though, I’d expect it to always trigger or never trigger not latch up. It may be worthwhile to put a capacitor between ground and VCC to make sure current spikes aren’t causing VCC to drop, but I don’t think we are drawing enough current to make that likely.

Peter


#26

That’s pretty much the same as when I cut L1, because I have a button that shorts R4 to GND. The only difference is C1 is replaced with a jumper.
Replacing C1 with a jumper makes the LED latch on button press.
Changing R3 to 680K makes the LED stay on even before the button press. But putting C1 back makes it somewhat normal again.


#27

Can you measure the voltage at pin 2? In theory the trigger voltage should be 1/3 of 12V or about 3.6V. What I am trying to do is get the rest voltage (with no pulses from the coil) 1 V or so higher than the trigger voltage so it should trigger with 1 or 2 V less from the coil. Its possible that the trigger voltage isn’t doing that for some reason (such as the trigger pin is drawing current and lowering the voltage which I don’t think it should be).

Peter


#28

Damn, I had my pin 2 at the junction of R3 and C1, which worked somewhat, but now that I move it down to C1 R4 the LED stay on all the time. With that config, LED always on, it’s 180mv from pin 2 to GND. I don’t know if this is right but the other side of C1, the C1 R3 junction, it’s 11.4V.
Replacing C1 with a link shows 4V at R4 R3 pin 2 junction.

PS : Is this annoying yet :slight_smile: Because i bet if you were here you could fix it in 1 min.


#29

Two problems here (one of them mine :slight_smile: ). The resistor values are a bit wrong, it looks like the 555 draws current on the trigger lead so R3 should be 560K and R4 should be 680K. For me that makes pin 2 6.3 V and grounding pin 2 then reliably triggers the 555 (I dug one up and tried it!) Initially I saw the same problem as you: the 555 would trigger once and then hang. It turned out my C3 (a tantalum cap) was defective. With pin 6/7 disconnected the capacitor was only getting up to about 4V (likely due to leakage in the cap) which isn’t enough to trip the high side comparitor reliably. When I replaced the cap with one that gets close to 12V with pn6/7 disconnected the circuit started working. It would be worth disconnecting pins 6/7 and making sure the R2/C3 junction gets close to 12 V as it should. From the voltages you quote (.2V at pin 2 which is low and thus continuously triggered and 11V at R3) your problem is the coil is disconnected between R3 and R4. There needs to be a DC path (either the coil or a jumper across where the coil should be) so the voltage divider functions. The coil (if it isn’t seeing any sparks) will be basically 0 ohms and not affect the voltage divider. When a spark happens the voltage induced in the coil is supposed to drive pin 2 lower than 3.6 volts and trigger the 555. Hopefully the 560K/680K resistors are large enough that the coil will produce enough voltage to cause the trigger. As noted above you should be seeing about 6.3V on pin 2 of the 555 with it not triggered. If I reduce R4 to 390K (which produces 4.3 v on pin2) the circuit triggers once and then hangs, so it looks like there is some slop in the supposed 3.6V trigger threshold. Hope some part of this helps! Who knew 555 circuits were this complex :slight_smile: , they usually just work. Once this circuit works with grounding pin 2 reliably then the excitement will be if the present coil provides enough voltage to reliably trigger the 555. If not you may need to increase the number of turns on the coil to get more voltage.

Peter


#30

Yay, the LED blinks when the spark plug is firing.

I went back to 10 ohm for R1 - it was 1K - and it’s brighter, but I might increase that a bit. With my cheap Ebay scope I get 1V across the LED with 1K but 7V across the LED with 10 ohm.

EDIT - 50 ohm makes 3V.


#31

Good news! With that working now you can adjust the duty cycle and LED resistor to get as much light as possible without burning out the LEDs. Even a cheap scope is a huge step up from no scope and while I don’t have one (I paid a couple of grand for a real Tek scope 25 years ago or so, excellent investment!) the 138 type Ebay scope looks pretty good for the price.

Peter


#32

I wouldn’t mind a real scope, because the DSO138 certainly comes in handy, it’s just that I would never use it enough to justify the price. Oddly enough I don’t use the waves that much, more the displayed values like frequency and voltages.

The 10mm LEDs came in so I just have to wait for the COB and I will test in on the car all at once, because it too hard doing multiple setups.

Thank for your help, because I’m definitely not smart enough to design circuits - I’m more monkey see monkey do :slight_smile: -.


#33

Preliminary brightness testing only, no accuracy testing.

With old LED timing light circuit :-
1 - 4 off 5mm LEDs
2 - 8 off 5mm LEDs - resistor value wasn’t reduced for larger current draw, so light intensity could be improved
3 - 8 off 10mm LEDs - resistor value wasn’t reduced for larger current draw, so light intensity could be improved
4 - Peter’s(Vanepp) new improved LED timing light circuit using 8 off 5mm LEDs . The circuit filters stray after-triggers and increases light by delaying LED switch off. and that produces a longer “block shape” mark. To use this circuit accurately you must only use the leading edge of the timing mark.

I stuffed up with the COB because it was 30V, and that made it barely turn on even without a resistor.


#34

If anyone is still interested.

I checked the accuracy and they all look good, even Peter’s.