Maker Pro
Maker Pro

Hall effect ignition

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Hi, I've been studying electronic ignition circuits for some time because I want to build a circuit for my older motorcycle. Effectively I want to replace points and condenser with a gear tooth sensor on the existing points cam. Then use one of the nice igniter chips optionally running a more powerful coil.

The bike has 12v points ignition, a dual output 4.5Ω coil with wasted spark. On hand I have some BIP373 coil driver chips and 1GT101DC gear tooth sensor. The latter is too big to fit in situ but could be good for testing.

ATS616LSGTN-T looks like it would be a good choice for the hall sensor. I read of reports of success using it.
IRGB14C40L is another good looking choice for coil driver. I see it used IRGB14C40L circuit with apparent success.

I'm aware that I may have to invert the signal from the hall sensor depending if it goes high or low when it sees a tooth.

Is there any recommended circuits that would suit my needs? Keen to hear opinions, tips etc. Happy to supply more details.

Thanks.
 

kellys_eye

Jun 25, 2010
4,880
Joined
Jun 25, 2010
Messages
4,880
The obvious method would be to de-construct existing EI designs - and, over the years, there have been dozens published in various electronics magazines.

Nice as it may be to build your own from scratch you are still re-inventing the wheel and there's (probably) nothing new to be discovered about their operation and/or efficiency.

I'll have a scan through my magazine collection and see if I can post some examples - meanwhile you could do worse than Google 'electronic ignition schematic' and view the 'images' results.
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Thanks for the reply. Agreed, no need to re-invent the wheel. There must be numerous examples of EI circuits that are tested and proven.

Mr.Google has shown me various circuits over the weeks. I'm quietly learning a lot. I posted a link above to one particular circuit that caught my eye.

Thanks for the magazine examples if any turn up.

Further study on ATS616 looks very good re output signal polarity. They are switchable between rising edge or falling edge simply by reversing direction.
 

Bluejets

Oct 5, 2014
5,884
Joined
Oct 5, 2014
Messages
5,884
As Kellys-eye says, re-invent the wheel.
What you are attempting may seem simple at the outset but it will quickly get complex due to detail you never thought about.

However, a bit of diy stuff here , if you must.
https://transmic.net/
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Bluejets, thanks for the thoughts. Yes I know how complex these things can become as I once set up Microsquirt ECU on my dragbike. I adapted a hall effect pickup from another brand of motorcycle, handmade the rotor, then used CNP coils with inbuilt ignitiers.
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Question : Looking at voltage supply for hall effect ATS6xx, it can be up to 26v.
Datasheet for BIP373 states max V(BE)= 14v, while that for IRGB14C40L says Gate-Emitter voltage "clamped". It appears I can run regular battery voltage?? Is this OK? Certainly it would keep things simple.... and simple is good.

Should I need a voltage regulator for the signal, the LM7805 seems like a good choice.
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Here's where I'm up to with the circuit..
upload_2017-9-3_21-50-58.png

ATS627 gear tooth sensor
IRGB14C40L igniter

The hall effect pulls low when it sees a tooth.

I found this circuit, said to be working. I understand everything bar the diode and the cap. Can someone please explain the need/function of these 2 items. Thanks.
 

Bluejets

Oct 5, 2014
5,884
Joined
Oct 5, 2014
Messages
5,884
In my experience with ignition systems and hall effect, the C1 would not need to be anywwhere near the value shown.
It is to keep noise (decoupling) from effecting the hall . Usual value for this would be around 0.1uf.
Essential to mount this right on the terminals of the hall effect.
The diode across the coil is a free-wheeling diode.Voltage may not be high enough and as it is a standard diode, perhaps not fast enough either.
Maybe a schottky diode would be better.
Don't know that R2 is essential but probably one to ground off the gate.

I hope you realise there is a lot more to even a kettering system than you have shown here.
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Bluejets, thanks for the thoughts. You make some excellent points.

Quite right on C1 value at 0.1uf

How would this Schottky do?
Looking at freewheeling diodes, I note it's also called a flyback which reminded me of my study of points and condensers way back in the 70's. As I recall some 400v was induced into primary winding at the spark event.

Yes, I see what you mean about R2. Effectively R1 and R2 are in series. I'm seeing it as R1 is a pullup and a current limiter in one. No need for R2.

Would I add a resistor from gate to ground? I admit I'm unsure on the purpose. Please elaborate.

Any other thoughts before I revise the diagram?

Thanks!
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Update:
Re: resistor gate-to-ground I found this snippet...
a high value resistor to ground is a good idea to avoid capacitive coupling driving the transistor when it is otherwise not connected.
In the example on a mosfet a 1m resistor was used.
 

Bluejets

Oct 5, 2014
5,884
Joined
Oct 5, 2014
Messages
5,884
You are not looking at current limiting perse.....rather a signal voltage as mosfets are voltage operated. 10k to ground should suffice. Note, when the hall is on the mosfet will be off.
 

solarpanels

Sep 4, 2017
1
Joined
Sep 4, 2017
Messages
1
Hi Guys, new here, a noobie ......

What about dwell, coil needs time to build magnetic field, unless hall chip has it built in.

R2 is there should 12v appear on the gate as the gate is protected back to back zeners and depending on the zener voltage they would smoke if taken over voltage without R2 to limit the current. Specs don't say what the zener voltage is?

Yes PetePan, use pull down resistor to keep the gate grounded as input impedance is very high and can be turned on by stray capacitive voltage. Gate capacitance will hold a charge so the device can be held on (or off) without a pull down resistor to turn it off. Pull down resistor is then the input impedance.

I have a preference for CDI for some reason, hall device just replaces the points. No dwel required with CDI.

More info, gate emitter breakdown voltage is 10 to 12v on spec sheet.
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Bluejets, OK thanks, sounds good, will run with 10k to ground.
Re: hall is on the mosfet will be off. Yes, that's the way I see it. When the hall sensor sees lobe of cam, it grounds the 'out' pin which turns the mosfet off. This mimics the cam opening a set of points.

Solarpanels, thanks for chiming in with some great food for thought.
Re:dwell. My thought is the existing points cam has dwell already built into the shape. The size of lobe and distance between lobes dictate how long the circuit will remain open/closed in the same way points running on the very same cam would. Clearly this is not nearly as ideal as a modern system which uses milliseconds for dwell.

RE: R2. Isn't R1 going to limit the current in any case?
The zeners appear to be the method it uses to "clamp" voltage? Looks like overvoltage shorts to ground through source. How to understand if it's safe to 14v?

A pull down resistor on the gate is a very good idea as any wrongly timed, random spark could spell disaster.

Yes, there's lots to like about CDI. In my case I'm keen to get rid of points as a 1st step.

Re: gate emitter breakdown voltage . Is it pushing the limits expecting it to be happy with 14v (running max volts)?
Would a 5v regulator be safer or is 14v safe enough?

Thanks!
 

Bluejets

Oct 5, 2014
5,884
Joined
Oct 5, 2014
Messages
5,884
Originally you spoke of "replacing" the points cam with a geartooth ( don't know how or why) Your drawing shows using a mosfet yet the number is a different element. Also your description jumps between transistor and mosfet as is your circuit operation description. I feel if you want to get this thing right with any inkling of success you will need to rectify these items. Overall I fail to see where any improvement would come about.
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Bluejets, sorry for the confusion. Please allow me to clarify..

The plan is to replace points and condenser with a "Gear Tooth Sensor". They differ from a conventional hall effect because they have an inbuilt magnet. Therefore there is no need to have a rotating magnet. The existing points cam will be retained to trigger the hall effect. Each lobe will be as a tooth.
Crane™ have made such a hall effect "points replacement" for V8's
http://www.cranecams.com/403.pdf
Sadly the unit is too big to fit my tiny distributor housing.

Researching the IRGB14C40L
Insulated Gate Bipolar Transistor (IGBT) is a minority-carrier device with high input impedance and large bipolar current-carrying capability. Many designers view IGBT as a device with MOS input characteristics and bipolar output characteristic that is a voltage-controlled bipolar device.
http://www.irf.com/product-info/datasheets/data/irgs14c40l.pdf

The circuit diagram provided above was copied as found. I believe the G-S-D terminals on the IGBT should be Gate, Collector and Emitter as the datasheet shows.

Hope this helps. Thanks.
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Continuing the study I find my choice of hall effect gear tooth sensor isn't ideal according to comments on various forums. Better is the ATS675 which has TPOS (True power on state). As I see it, these can be purchased either switching high or switching low.
ATS675 datasheet

A question: where in the circuit can I safely put a simple LED to help with static ignition timing? Effectively the LED should illumine when the IGBT is on.
 

NMNeil

Oct 3, 2014
110
Joined
Oct 3, 2014
Messages
110
Check out the ISL9V3040D3S from Fairchild, it's designed specifically for vehicle ignitions and has all the associated protection circuitry built in. The other advantage is that it is logic level so you can use a 5 volt microcontroller at a later date for timing and dwell control etc.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
The diode across the coil is a free-wheeling diode.Voltage may not be high enough and as it is a standard diode, perhaps not fast enough either.
D1 is not a freewheeling diode, it is the wrong way round. It may be there to reclaim unused energy.
A freewheeling diode stops the drain voltage rising much above the supply voltage, this would kill the spark. A better arrangement would be to put a 500V zener across the fet to protect it from overvoltage. Is the 1N4007doing this job?
 

PetePan

Sep 2, 2017
27
Joined
Sep 2, 2017
Messages
27
Thanks for the thoughts. WIll study the ISL9V3040D3S

I believe that (1N4007) diode deals with flyback voltage from the coil. I upgraded to a 6A of which I had plenty.
I removed R2. I fitted a 10k resistor to ground off the gate but it keeps the transistor off. I tried values up to 3 times but result same. Without the resistor connected circuit was working.
 

WHONOES

May 20, 2017
1,215
Joined
May 20, 2017
Messages
1,215
Look up RCEXL ignition systems. They are spark ignition systems intended for model aircraft engines. They use a Hall effect device to trigger the unit. Ignition timing is adjusted automatically over a range of 28deg (if I remember correctly) up to 8000rpm. They are used on some fairly large model motors of 20HP or so. They are not expensive at about £40 to 50 UK and would save you lot of time and heartache. Go to www.rcexl.com. I am sure they are available in NZ. If not they definitely are in Aus'.
 
Top