Hall Sensors dying on Holt 75

[michelko]

Hi Guys,
since i got my Holt running it worked real nice with that cheap chinese CDI Ignition.
But since a couple of days the hall sensor going mad.
I changed 7 or 8 times but the engine will only run a couple of seconds to a couple of minutes fine and suddenly stops.
And another hall sensor is fried. I changed the holder of the Hall sensor from bras to delrin, helped not. Changed to another CDI, same symptoms.
Has anyone a good idea whre to look?
I am a little frustrated now.

Regards Michael

[michelko]

https://www.stirlingkit.com/products/ignition-starter-kit-for-horizontal-combustion-engine-hit-miss-gas-model
this is the first ignition i used with succes for some time.

https://mbi-design.de/wp/produkt/cdi-zuendung-a-01/
that is the second one i testet with the same results.

[michelko]

The Sensor is an Infineon TL 4905l

[Jo]

Have you measured the voltage coming out of the ignition going to the hall sensor?

Jo

[gbritnell]

In all my years of using Hall sensors I have only had two burn out, that is until my flathead engine. I must have gone through five with it. I ultimately traced it to my Hall lead to wire connection.  It was too close to the metal holder and one day I could see an arc jumping to ground. Make sure that you have a good path back to the negative battery connection and that all ground connections are good.

[michelko]

Hi,
here are some pics of the setup.
The ignition only has two connections to the enngine. The red HT lead and the black ground. both were screwed to the engine.
The battery is an 2S Lipo so around 8V.
al the conections to the sensor were covered with shrink tube and the sensor itselv also.
That setup worked for months but suddenly started to fry the sensors.

[Jo]

The battery is used in the CDI to generate different voltages, one of which is the HV for the plug so the battery voltage is no indication of the voltage going to or across the sensor.

The guys have been describing a common problem: If the ground is not earthed to the engine or there is resistance on the return then the Hall sensor may be experiencing the HV as feedback and hence going "Pop".

Time to get your resistance meter out and check the earth is a good dead short all the way to the sensor. Then measure the feed and return from the sensor with the volt function. These devices are pretty good they can normally go over 30V DC but if the earth isn't there or is intermittent the voltage will float much higher.

The fact that you have tried two CDI's tells me it is something about the mounting on the engine.... its all pointing towards the earthing 

Jo

[michelko]

o.k.
7.6V at the red and black lines to the sensor
0.8 Ohm between the black line of the sensor and the black line to the case also between the battery ground.

I agrre to your earth suspect, but canīt find the cause.

[Jo]

0.8 Ohms is a bit high, allow for the engine to warm up its likely to move up or down but importantly it should be a dead short. The quick and dirty way to check if it is /solve the earth problem is to run an additional earth wire up and put it under the screw that mounts the sensor to the engine.

I suspect the other 0.4V from the battery is the float caused by the 0.8 ohms resistance. Have you buzzed out the wire going from the CDI to the sensor? Was it the same one used for the two CDIs?

Jo

[Jasonb]

Is that connector right. looks like red and white are crossed

[Jo]

The TLE4905 has reverse polarity protection.

Jo

[Jasonb]

None of my CDIs need a connection from block to the battery negative. My battery just has a futaba conector and that is how I connect to the CDIs supply.

The CDI ground is connected to the block like the bigger black wire in the photo and the plug lead to the plug.

[Jasonb]

The above was a reply to a post saying that the CDI needed a wire from engine frame to battery -ve which seems to have now been removed just incase anyone was wondering what it was about.

The CDI shown here has exactly the same wires as the popular S/S one from the States and the Minimag from the UK so no need for any additional wires. That has worked ok for me over 18 years since I first bought a S/S one.

2 wires to battery
3 wires to Hall Sensor ( less if just uning contats)
1 wire to spark plug
1 Wire ground to engine frame

[Jasonb]

The TLE4905 has reverse polarity protection.

Jo

But the wires that look to be crossed in the connector are not the +Ve and -Ne (ground) as the central negative or ground is black and right in the connector. so signal and positive look to be crossed

[Jo]

Reverse polarity protection means if the voltage is connected to the wrong pin the device won't instantly die and will still work if you realise and connect it correctly.

The initial condition was the engine was running happily from the connected CDI: This tells us the wiring was correct. Something has changed which has caused the device to blow. The logic is the likely thing to change is either wires have broken/come loose or in combination the warming of the engine is moving the wires causing a poor connection.

Jo

[michelko]

Hi,
i tried an old self made Transistor ignitiion with the same effect, so its sure the problem is not located in the ignition itself.

But i think i found a solution.

i moved the triggerwheel 0,4 mm away from the sensor and also used some heatglue to isolate the sensor. Donīt realy know why but that worked so far.
Ran the engine a couple of times for several minutes without frying the sensor.

Michael

[petertha]

Hi Michael. I've been following this thread. Glad you are making progress. Can you show some pictures or sketch of your trigger wheel & hall sensor proximity?
I'm curious about this stuff as I'm nearly ready to make a distributer based on design ideas from others once some more materials arrive. I want to try both a shutter wheel & magnet wheel on the bench first.

[michelko]

Hi Peter, of course. After my Ski Trip i will Post the Information

[michelko]

Here is a cut view of the Ignition system
You can see the triggerwheel with the small windows. The magnet is glued into the dizzy body

[petertha]

Thank for sketch to help me understand. Hopefully your issue is becoming resolved. I cannot offer anything constructive or useful as I'm just about to start on a bench test distributer myself. Both a magnet wheel style & shutter wheel style just for personal interest. In both cases, heavily leveraging on the work & help of others who graciously shared some details. I have a method to hold the hall effect sensor in what will probably be a plastic part, maybe 3DP. But that has more to do with it being a fiddly little thing to make in metal. I'm not really certain if plastic vs metal is better from electric/magnetic perspective. I did evaluate the V8 distributer dimensions provided by George Britnell in another post, hopefully he doesn't mind me including attached CAD renditions of his 2D PDF & of course he would be the expert to consult on that design. (Assuming I drew it right) this is his shutter window relative to magnet size. And I think I measured about 1.2mm between magnet & hall effect faces. His shutter wheel is very thin. A lot of 'stuff' packaged into that ~26mm OD distributer body to fire 8 cylinders.

Anyway, I didn't want to drift off your troubleshooting topic which is maybe completely unrelated to 'design', especially since you mention it has already been running. Thanks again! PT

[michelko]

After the small adjustments it is running fine. no more fried sensors so far.
Here are some pics from the trigger setup.

[petertha]

Thanks for the pictures. I've been working on a spreadsheet that calculates some ignition related parameters. I think I understand your shutter type disc/wheel configuration. I have a few questions:

- what is the shutter open window size & magnet diameter it is exposing. Just as an example, 4mm wide window exposing 3mm diameter magnet, 22mm wheel/disc diameter....
- do you have an estimate of your typical idle rpm & max rpm range?
- presumably the trigger wheel is made of steel? What is plate thickness?
- you mentioned you moved the trigger wheel to 0.4mm from sensor, what is the distance between magnet to sensor face?
- when you had difficulties before, do you recall if the engine was at sustained low speed? And now you are basically repeating the run & no further issues?
- does either the Chinese or the MBI module instructions/documents mention anything about a minimum or maximum dwell time?

I don't want to speculate on anything because I am complete novice to this ignition stuff, but just wondering about some things...

[michelko]

Hi Peter,
i attached a drawing of the trigger wheel made of mild steel
Idle rpm is about 900rpm, max not measured yet. I believe it would be around 4-5000 rpm.
Distance between Magnet and sensor is aprox 2,2mm, the magnet is a 3mm dia neodym.
No dokumentaion with the chinese ignition.  But they are firing at opening of the window while the transistor ignition is firing at closing.
Just like a kettering Ignition with points and capacitor

Greetings