Hall effect woes!

[creast]

Hi,
I have finally managed to get my Kerzel H&M running but am now plagued with Hall effect failures.
I have used a few circuits but the latest was posted in an earlier thread here which I cannot track back to. The circuit is attached.
This circuit works fine but I have yet to run more than a few minutes before the Hall effect goes T*ts up and dies. I have put protection diode and capacitor as recommended elsewhere but no luck. Any ideas?

 

[Noitoen]

There is no supply to the hall device. You should put a 5v zener diode in parallel with the hall ic to protect it from voltage spikes

 

[jgedde]

If I understand correctly, you're saying you've tried multiple driver circuits and the hall keeps dying?

Is this the circuit you were referring to?

http://www.homemodelenginemachinist.com/f31/new-ignition-circuit-20415/index2.html#post214570



John

 

[stevehuckss396]

I have found that a poor ground to the engine block causes most hall failures in my garage.

 

[canadianhorsepower]

You should be sending you Hall circuit connected to this one
this would allow us to see your source voltage and if you had a current limiter somewhere

 

[creast]

Although the circuit shows no direct supply on the diagram, I am supplying 12v to the Hall device which the spec says it should be ok with. The supply however is the same as that to the coil so maybe thats a problem?

There is no supply to the hall device. You should put a 5v zener diode in parallel with the hall ic to protect it from voltage spikes

 

[creast]

Thanks Steve,
I found that too when I forgot to connect the -ve rail to the engine and it arced from the Hall to the block. However, I have ensured a good connection to the -ve rail now and still I get Hall failures.
A friend has suggested a true earth, ie earth to ground or on appliance, as high static charges can build up and at one stage I was getting mild electric shocks as I spun the flywheel.
I can't think that I would have to have to ground to earth since we don't do that to our cars do we.

I have found that a poor ground to the engine block causes most hall failures in my garage.

 

[creast]

Hi John,
I have not tried that particular circuit.
The one I attached is somewhere on this site but I am having problems finding it again.
I had tried Jan Ridders Blokker circuit modified with a hall effect but had similar problems with the Hall effect dying.

If I understand correctly, you're saying you've tried multiple driver circuits and the hall keeps dying?

Is this the circuit you were referring to?

http://www.homemodelenginemachinist.com/f31/new-ignition-circuit-20415/


John

 

[jgedde]

OK. Try this. Power the hall with a 100 ohm resistor in series with its power lead. Put a 0.1 uF capacitor between the power lead and the ground pin. Put a 12V or 15V zener diode in parallel with the cap between the power and ground pins on the hall.

Using a zener will protect the hall from both positive and negative transients. For positive transients, it'll avalanche at 12V, for negative transients it'll conduct like a regular forward biased diode and clamp the transient to about -0.6V.

If the Halls still blow, then it's the output transistor in the hall that's popping. That transistor is comparatively large in scale, but on the Honeywell die (who make the guts for many manufacturers) the collector lead to the transistor burns up (I was present for a Hall sensor failure analysis and that's what we saw when the part was delidded).

If the resistor, cap and zener diode don't help, put a 100 ohm series resistor between the hall output pin and the drive circuit to prevent excessive currents. If it's this solution that works, we have to dig into this some more. If it does, I would suspect the large value of capacitance in the circuit (4.7 uF) is providing enough surge current to pop the Hall output transistor's collector trace.


In addition, it's IMPERATIVE that the engine block be well grounded. If there's aluminum around, you'd be surprized about how hard it it to get a solid electrical connection through the oxide layer and hardware.
John

 

[canadianhorsepower]

jgedde;
instead of blasting any sugestion"like we are all phone post" are stupid and relating your circuit to close to perfection
do me a favour and everybody else a favour
put your knowledge were your mouth is and draw a circuit that we can all use
NOT ONE THAT YOU CAN ***** ABOUT

cheers

 

[jgedde]

jgedde;
instead of blasting any sugestion"like we are all phone post" are stupid and relating your circuit to close to perfection
do me a favour and everybody else a favour
put your knowledge were your mouth is and draw a circuit that we can all use
NOT ONE THAT YOU CAN ***** ABOUT

cheers

What??? You've got the wrong idea friend... My circuit is irrelevant. He's having problems and I have some ideas on how to fix it.

The parts for my circuit will be here on Monday. If it doesn't work, I'm humble enough to say so, revise it accordingly, and share my results. But, I really have little doubt it will work with no more changes than resistor values.

John

 

[canadianhorsepower]

What??? You've got the wrong idea friend...

ok then prove it, or put it on the table

 

[jgedde]

ok then prove it, or put it on the table

Like I said, the parts for my circuit will be here Monday. Apart from my circuit, if my suggestions on how to cure the hall sensors from dying works, then everyone is happy right?

John

 

[GailInNM]

Play nice Luc.
Everyone here is just trying to help. No need to fight.
Gail in NM

 

[creast]

jgedde,
Thanks for the suggestions. I will give them a try and keep fingers crossed.
Not sure why Canadianhorsepower gave that outburst when you are being helpful.
Thanks anyway
Rich

OK. Try this. Power the hall with a 100 ohm resistor in series with its power lead. Put a 0.1 uF capacitor between the power lead and the ground pin. Put a 12V or 15V zener diode in parallel with the cap between the power and ground pins on the hall.

Using a zener will protect the hall from both positive and negative transients. For positive transients, it'll avalanche at 12V, for negative transients it'll conduct like a regular forward biased diode and clamp the transient to about -0.6V.

If the Halls still blow, then it's the output transistor in the hall that's popping. That transistor is comparatively large in scale, but on the Honeywell die (who make the guts for many manufacturers) the collector lead to the transistor burns up (I was present for a Hall sensor failure analysis and that's what we saw when the part was delidded).

If the resistor, cap and zener diode don't help, put a 100 ohm series resistor between the hall output pin and the drive circuit to prevent excessive currents. If it's this solution that works, we have to dig into this some more. If it does, I would suspect the large value of capacitance in the circuit (4.7 uF) is providing enough surge current to pop the Hall output transistor's collector trace.


In addition, it's IMPERATIVE that the engine block be well grounded. If there's aluminum around, you'd be surprized about how hard it it to get a solid electrical connection through the oxide layer and hardware.
John

 

[jerrybilt]

The first thing I would do with this circuit is place a diode ( 1N4004 ) across the primary winding of the ignition transformer. The anode of the diode is connected to pin 2 and the cathode is connected to pin 1 ( + 12V). The other thing that I would do is to redesign the final drive to the ignition transformer. Notice that the transformer is left floating when the final transistor (IGBT) is turned off so an emitter follower, where transformer will be earthed, will be more suitable.

Jerry.

 

[jgedde]

The first thing I would do with this circuit is place a diode ( 1N4004 ) across the primary winding of the ignition transformer. The anode of the diode is connected to pin 2 and the cathode is connected to pin 1 ( + 12V). The other thing that I would do is to redesign the final drive to the ignition transformer. Notice that the transformer is left floating when the final transistor (IGBT) is turned off so an emitter follower, where transformer will be earthed, will be more suitable.

Jerry.

Jerry, I believe you are mostly correct. However, a diode across the primary clamps the kickback voltage (since it spikes positive by a few hundred volts at the collector of the transistor when it turns off). While this would give a path for the current as you state, but it also affects the output voltage... Instead of transforming 300V or so by a ratio of 100, you'd only be transforming the diode drop by a factor of 100.

That's why I maintain the age old method of putting a "condenser" across the breaker points (or output transistor) is necessary. Not only do you give the oscillating current somewhere to go without clamping the kickback, but you get a LC resonant circuit which, if used with the right value cap, will give you even more voltage...

John

 

[jerrybilt]

Only one problem John, when the drive transistor turns off the transformer is effectively removed from the circuit as the common connection is left floating and the only thing completing the circuit is some parasitic capacitance that helps complete the circuit. The capacitor across the points will stop the arching due to the high current being interrupted. This arching causes the current to decay more slowly hence the induced voltage is much lower in a system with points. But using a transistor, as opposed to points, as a switch there is no arching and the rate of change of current is limited by the primary inductance and the circuit resistance.

 

[jgedde]

Only one problem John, when the drive transistor turns off the transformer is effectively removed from the circuit as the common connection is left floating and the only thing completing the circuit is some parasitic capacitance that helps complete the circuit. The capacitor across the points will stop the arching due to the high current being interrupted. This arching causes the current to decay more slowly hence the induced voltage is much lower in a system with points. But using a transistor, as opposed to points, as a switch there is no arching and the rate of change of current is limited by the primary inductance and the circuit resistance.

Again, the capacitor does more than prevent arcing. It provides a bidirection path for the flyback current, thus the circuit is not open... In any case, even if there is no capacitor, the circuit is not floating when the transistor turns off. An ignition coil is an autotransformer and secondary current flows through the spark plug thus completing the circuit.

 

[canadianhorsepower]

Play nice Luc.
Everyone here is just trying to help. No need to fight.
Gail in NM

I don't have problem with playing nice and I understand that we are all trying to help.
What pinches me is when your giving wrong information to someone just to look smart.
Not having the circuit diagram or parts number any parts replacement
suggestion can make things worst .
And nobody on earth can guess what circuit he's using