Slowing/retarding a signal pulse

[tony b]

Hi guys,

Just found this place after having an idea for modifying an electronic distributor ignition system. The system works with a photosensor detecting a break in a beam of light from an LED, when a fan-like object spins over it. As each blade passes through the beam, the signal is broken, and this tells the ignition system to fire (a good description is here http://www.cars-gb.com/lumenition.htm)

I was wanting to make a simple way of adjusting the advance on the distributor.

I was thinking some sort of variable resistor and maybe capacitor setup, but to be honest, I wouldn't have the foggiest where to start!!! I basically need to be able to "slow down" a signal sent from the distributor to the controller - not sure by how much though, probably less than 1/8 second. This would mean that future adjusting of advance on the distributor would be very easy.

Anyway, figured I may as well see if anyone had any ideas, like I said I'm clueless!!!

Cheers,

Tony

(ps thanks to skip for sending me here!)

[Chris Smith]

There are several ways to do this.

First you can slow down the original signal so that advancing this signal is easy, or you can use the previous pulse as the marker for the next signal and then play with that.

Flip flops can be added to a circuit and then varied in speed response to advance or retard for this project as well as many other types of circuits can be used for this method.

It just depends on how technical you want to get.

Flip flops, timers, multi vibrator, etc can add in a simple “add or subtractâ€

[tony b]

lol, read that 3 times, and it's still predominantly greek!

My electronics knowledge really is abismal - bit of faffing around at school, etc - so I'm afraid I don't understand much of that...

I'll go google flip flops, as they sound promising!

[Chris Smith]

A flip flop multi vibrator like the 555 timer basically is a set of triggers.

They can be set to trigger with an adjustable time delay in between.

When the first signal from a distributor hits the first input of the multi vibrator it then has a preset or adjustable out put [delay] time from this trigger.

So if the Ignition hits at say 12 O’clock, and the delay is set 50 ms later then the out put [second switch] of the multi vibrator will go off at 12 O’clock plus 50 ms later.

This second trigger is fully adjustable from 1/100,000 of a second delay to hours delay between being set off.

You figure out what the slowest and fastest rates are for your design and then design the resistor / capacitor values to fit in between for your timing cycles.

[haklesup]

In general you need to know a bit more about the system you are trying to modify before it would be safe to experiment.

First, the pulse generated by the optical sensor, you just want to deay when that is sent to the car's computer. I think a one shot multivibrator made of a 555 chip will do the trick however you need to look more closely at the polarity of the pulse you seek to delay. Is it a low to high or high to low edge you need to move or is the computer sensitive to the logic level as simple combinational (74XX) chips are. These details will tell you if additioanl components to invert the input ot output are necessary.

There is lots on the 555 and 7555 chips if you search. There are lots of creative circuits that utilize this chip to form a pulse delay or square wave. At least with the 555 you can operate at the car's 12V supply rather than making a 5V supply to power other technologies. I don't know if the logic from this optical sensor is 5V or 12V logic. Another detail you must know ahead of time.

In any case, isn't that what the car's ignition computer does anyway. You may find it more fun to hack the EPROM and make your car programmable from a laptop.

Also you may be able to get a similar effect by physically moving the optical sensor (or the disc encoder) to a different position (out of phase from normal)

[tony b]

Thanks for the quick summation of the 555 timer, seems to be the sort of thing I'm looking for.

With regards to it being a 5V or 12V signal, would be amazed if it wasn't 12V, but will have to ask the question I suppose! Relocating the disc encoder/reader on the dizzy is the way you're supposed to set the advance correctly when setting up, my idea was to enable you to just get it roughly right to start with, then fine tune it electronically.

As for hacking my EPROM - lol - it's a 1976 mini with a 1989 (or thereabouts) metro engine in it!! The dizzy is in effect a physical EPROM...

If you know how to hack EPROM's in general, i know anyone with an MPI mini (the later, fuel injected type) could be very interested, as re-map kits, and programmable stuff for that weighs in at great cost!!

I'll see what I can find out about the signal - low to high or vice versa, and then come pick your brains some more!!

Thanks a bundle for help so far

tony

[rshayes]

A time delay circuit, such as the 555 or a monostable multivibrator, might not work very well. The distributor could be set for full advance and a time delay added electronically. The problem is that the delay is usually expressed it terms of degrees of crankshaft rotation. This would require a circuit whose delay is both a function of an adjustment pot, but also inversely proportional to engine speed.

This could be done by generating a voltage proportional to the engine speed using the frequency of the input pulses. This could be used to set the charging current to a capacitor which is reset to zero between ignition pulses. The time for the capacitor to charge to a threshold voltage would then be inversely proportional to engine speed and a fixed number of degrees after the trigger pulse.

[MrAl]

Hi there,

One thing you should keep in mind i think is that when the engine
speed increases the delay should decrease. The reason for this
is so that the firing angle stays the same. If you dont do this too
then say at 1000rpm and you have the delay set such that
it gives another 2 degrees, at 2000rpm it will look like 4 degrees.
In a normal car system the only time this changes (at least with
the older model cars) is when the engine is gunned, where the
timing may advance more somewhat to give the engine more pickup.
With modern engines, i wouldnt doubt it if it is computer controlled and
the design tied in with polution control, so you could cause the exhaust
polutants to increase without knowing it unless you have a tester.
It's even possible the car wont pass inspection with too much of a change
like this.

BTW, have to tried rotating the distributer, which is the way a normal
car (at least the ones i've seen) gets it's timing adjusted?

[tony b]

The problem is that the delay is usually expressed it terms of degrees of crankshaft rotation. This would require a circuit whose delay is both a function of an adjustment pot, but also inversely proportional to engine speed.

...oh...

One thing you should keep in mind i think is that when the engine speed increases the delay should decrease. The reason for this is so that the firing angle stays the same. If you dont do this too then say at 1000rpm and you have the delay set such that it gives another 2 degrees, at 2000rpm it will look like 4 degrees.

...bugger...

lol, looks like I left my brain at the door when I envisioned this being simple...

Thanks, may have to rethink as this is getting well beyond my station!!!

(will revisit the idea if I get the system off ebay though... )

[haklesup]

Aside from the apparent flaw in the application, the 555 might also not be suitable due to its sensitivity to temperature. You would not have a very calibrated delay as you warm up an engine on a cold day.

Sounds like what you really want is to invent your own linited ignition controller. Probably possible but you need to better understand what inputs you can get and what outputs you can control. Then the complexity of the circuit will become apparent.

Once you go through that and choose a uController and develop your own firmware to burn on an Eprom, you'll understand why the kits are very expensive. I wouldn't give that effort away for free either.

However if you do develop something useful, you can write an article for N&V and sell the parts kits on eBay.

[philba]

The uC approach to providing the delay should be pretty easy to create. For some one familiar with microcontrollers, that is. I don't think you need to create a whole new controller module but just an independent delay unit. Basically you just have to read the input pulse, determine the RPM and map that onto the desired delay. Determining the delay is the part I don't understand but that's because I don't mess with ignition systems. This could be done with a very small/cheap micro. You'll probably need to map the in/out voltage level to/from 5V but this is not much in the way of parts, probably well less than $5 worth.

[MrAl]

Hi again,

philba i see what you mean. I would include a pot that can
be read from another AD port and depending on the setting this
could increase or decrease the delay percentage.
The delay percentage calculation is of course done based on
the frequency perhaps? I dont know for sure how well this would
work though, because rotor position is not really determinable from
frequency, unless the speed is constant, so i have no idea what
would happen when the first cycle of the frequency was suddenly
shortened (stepping hard on the gas petal). Perhaps the frequency
ramps up slow enough to make measuring the frequency (or rather
period) of the previous cycle and calculating the new angle from
that accurate enough.

One thing i still have to wonder though because my question was
not answered, is what is the main purpose for doing this in the first
place...after all automobiles do have an adjustment for this already
built in, and they even have automatic advance for extra pick up.

[Chris Smith]

None of this is simple, and then don’t forget the engine vacuum, rpms, road speed, Load, etc.

I would only use such an advance for a fixed engine, not a comprehensive engine under all conditions, unless you have the knowledge of all the feed back systems.

[tony b]

Well, I didn't win the set up off ebay anyway! Oh well, probably a good thing! Instead I picked up an electronic distributor for a fraction of the cost.

There's some very interesting sounding stuff flying around here, makes me realise just how complicated a project it would of been! Thanks for the thoughts