TIM 6 ignition

[Charles Lamont]

Thanks, dsage. Don't know what I was thinking. I am casting around because everything is as you suggest and I don't know why I have blown one.

The situation is quite complicated:

At the moment I have the engine mounted on the lathe cross-slide so that I can motor it over with the lathe to try to start it. It is driven via a torque limiter and a sprag-clutch so that if the engine goes tight it can stop, and if it starts it can run away from the lathe. At the other end there is a brake to put some load the engine.

All this means that the engine frame is not only grounded to battery negative, but also, via the lathe, to the household earth. I am getting a problem in that the VFD controlled lathe motor seems to be producing a lot of interference.

The TIM-6 is powered by a buck regulator that is switched on by an opto-isolated relay that is switched on by an Arduino. The TIM-6 LED output is connected to the arduino via an opto-isolator. This is so that the Arduino can provide a tachometer display and can automatically switch the ingition off if it stops with a magnet opposite the sensor. This is to protect the coil primary from overheating. The Arduino responds to an external input from a switch ( with a pull-down resistor). With the lathe motor off this all works fine. Initially I was just using a wire on the breadboard to connect the Arduino input to 5V.

Now, I found that all was OK until I switched the lathe motor on (with the clutch disengaged, so the engine not turning). At this point, with the ignition switch wire disconnected from 5V, the Arduino started oscillating between ingition on and off states. Removing the wire from the input seemed to improve matters, and I seem to have cured that problem by adding a 0.1µF capacitor between the Arduino input pin and ground.

The hall sensor connection to the TIM-6 has the wires tightly twisted together and in a heat-shrunk sleeve, with a finished length of 11".

At some point, I think probably before fitting the capacitor, I tried the whole lot, and the hall sensor was working, at least for a few seconds.

Is it likely that interference from the lathe motor killed the hall sensor?

Would a capacitor and/or diode across the battery terminals help?

Would isolating the engine from the lathe help? (irritating, as I would have to strip everything down to make an insulating ring for the drive).

I have some SS443A Hall sensors, and some Nylon bolts on order, but won't get them till next week.

I will post a picture of the whole kaboodle shortly.

 

[Charles Lamont]

Here is a picture of the set-up. The hall sensor related bits are all missing ATM.

 

[raspii]

Hi,

The buck regulators are great, but i have had many issues solved by removing them from specific circuits with abnormal current draws, is it possible to set up a battery assembly without the buck regulator to rule that out (to supply to the tim 6 directly)?

Best of luck,

From Pat

 

[Tim Wescott]

Oh, the humanity!

So, I haven't hooked up an ignition to a motor, but I have worked on power electronics. Your ground path is torturous. High-frequency pulses (like what happens when you spark a coil) like to go in straight lines. If you give them a bunch of wires that go every which way, they'll try to find a shorter path -- even if it's unofficial.

I've doctored up your picture. The official path that you're asking the pulses to go through is shown in yellow -- note that the ground path goes everywhere. This is bad.

Typically, the two grounding schemes that work well are a "panel ground", where the whole circuit is built on a metal plate or in a metal box, and everything is grounded to the nearest metal surface, and "star ground", where all of the grounding points go to one spot. What you have is a "please get that reviewed by a senior engineer" ground. I couldn't begin to tell you what your circuit is actually doing, but I can tell you that your spark pulse -- either from your primary or your secondary -- is going to show up everywhere.

I've also shown a recommended (star) grounding scheme for the ignition module and the coil in blue -- after I pasted this picture in, I realized that you should probably take the battery ground to the engine, too, or put a honkin' big bypass capacitor right at the TIM-6, so that the pulse current goes into the capacitor. For that connection, it's probably enough to move the blue wire from the engine to your terminal block over to your battery ground right next to the wire for your TIM-6. That'll keep stray high-frequency current away from the ground wires feeding all that sensitive stuff in the lower left-hand corner of your board.

An alternative -- while still bringing battery ground to the engine as above -- would be to splice the coil's ground wire into a "Y" right at the coil; run one wire to the engine block and leave the other going into the TIM-6. For a bit of extra paranoia, twist the power wire (yellow) from the TIM-6 with the ground wire from the coil. Ditto the power and ground for just about anything else, especially the battery leads.

Yet another alternative would be to go with a panel ground. Replace your pressboard with a sheet of aluminum, and just ground everything to that, local to the thing that uses it. If you do that, and keep the wires that carry ignition pulses away from the sensitive wires, you should be OK.

 

[Charles Lamont]

Thanks, Tim. I will think about the best way to implement your recommendations on grounding. Awkward little buggers, these electron thingys. I have plenty of ally sheet I could cover the board with, so that may well be part of the solution, or perhaps a chunk of brass to provide a star arrangement.

Meanwile, do you have anything to suggest on the problem of intereference from the lathe motor/VFD?

Gordon, I am sorry, I seem to have hijacked your thread. I hope some of this stuff helps you too.

 

[Tim Wescott]

Meanwhile, do you have anything to suggest on the problem of interference from the lathe motor/VFD?

There's a good chance that cleaning up your circuit will make it less susceptible to the VFD. Not a guarantee, but a good chance. You don't show the Hall sensor, so I can't tell, but running the wires to that in a bundle (there's 3, if it's like the Hall sensors I know) that's twisted together should reduce coupling to the VFD. The Hall sensor should be electrically isolated from the engine block, and there should only be one point of contact between your circuit ground and house ground.

We probably should move this to another thread -- is the lathe grounded, or is it electrically floating? If it's grounded, then that's your contact to house ground; you shouldn't have another one for your circuit.

 

[dsage]

>>> because everything is as you suggest.
>>> Oh, the humanity! LOL

Charles. Exactly NOTHING is as I suggested. There is so much wrong with this setup I won't make any suggestions.
Charles was brave to have made suggestions for improvements.

There is no magic in using earth ground. As I mentioned GROUND for you purposes is battery negative. In fact having anything touching the lathe bed (including the engine) probably explains why the VFD is interfering with your circuit.

As suggested. Please start another topic and maybe we can make further suggestions.

 

[Gordon]

Thanks, Tim. I will think about the best way to implement your recommendations on grounding. Awkward little buggers, these electron thingys. I have plenty of ally sheet I could cover the board with, so that may well be part of the solution, or perhaps a chunk of brass to provide a star arrangement.

Meanwile, do you have anything to suggest on the problem of intereference from the lathe motor/VFD?

Gordon, I am sorry, I seem to have hijacked your thread. I hope some of this stuff helps you too.

Not a problem but the discussion has gone far beyond my electronic expertise. I am sure that others know what is being discussed. I mostly just blindly follow circuits designed and tested by more qualified folks.

 

[74Sprint]

Wow, Charles there is actually a lot of things wrong here with your setup. I'll start with the Hall-Effect being blown. One thing that is being missed here is that you are using an automotive style ignition coil (transformer) where the negative side of both the primary and secondary are tied together. When an ignition coil fires it creates both a HV positive spike and a Negative spike. It is this negative spike that is killing the Hall-Effect through the ground. This negative spike can also reset your Arduino. The cost to remove this negative spike can reach half the cost of a circuit board in medical equipment. So yes you need to put noise filtering across the battery. I use electrolytic caps of 100uf, 10uf, and a ceramic of .1uf all in parallel across the battery terminals. In some cases I have even had to use a TVS (basically zener diodes) across the battery. If the coil/s are drawing a lot of current then I include an inline inductor on the positive side of the battery. I invented an electronic static bonding clamp and I use 2 different TVS's of different voltages to remove static voltage spikes when connecting the clamp. Go about 1.5-2 volts above your highest battery voltage.

Next do not ground the lathe to your circuitry, unfortunately your setup does not allow for that. VFD drives are very notorious for creating noise and your whole lathe will act as an antenna transmitting noise everywhere. You either have to eliminate the noise at the lathe or provide filtering at the Arduino and/or your buck regulator using the above mentioned filter. I could be more specific about the filtering if I knew the frequency range of the VFD being used. I had a problem with a MegaSquirt once and got lucky using an automotive stereo power supply filter for $10 that cleaned up the power and resets. You can do this on the battery leads going to your Arduino and buck regulator.

Arduino's are low voltage sensing devices not current sensing.
If you had an oscilloscope you could see all the things I'm talking about.
Ideally use an ignition coil that does not have the HV ground/negative tied to the primary negative.
Batteries also work like a capacitor for absorbing voltage spikes but, because of internal resistance they need help.
I hope this helps.

Ray

 

[Charles Lamont]

Ray - Just to clarify, the coil is a Minimag Novus 2 coil designed specifically for 2-cylinder model engine applications using a wasted spark. The HT leads are connected one to each plug. The HT is not connected to the LT.
I can isolate the circuitry from the lathe. Nylon bolts (on order), and fibre washer under the mouting plates will insulate the engine from the cross-slide. The brake is already insulated by its mounting block, and I can make a delrin bush to insulate the motoring drive at the chuck.

Dsage - following yours and Tim Wescott's observations, I now have much more idea what 'DIRECTLY' means, the wiring will be tidied up accordingly.
The Hall sensor leads were already tightly twisted together, and insulated at the engine end. The question about grounding at the engine end was just brain fade.

Is it OK that the hall sensor is mounted inside an aluminium timing cover shown here?
https://www.homemodelenginemachinist.com/threads/hall-sensor-mount.32290/post-341725This cover is mounted on the timing case, so it is well connected to the block.

I think we are on topic, as it is still about the practicalities of using a TIM-6 ignition, blowing Hall sensors, and other heffalump traps I have fallen into.

 

[dsage]

Charles:
Have you had this engine running yet?
If you are in the stages of just getting it running then I would suggest eliminating all of your fancy electronics and wiring and getting back to basics. until you have it running. Then you can add it back a bit at a time.
To that end:
Mount the engine on a board, clamp the board to the bench and simply use the batteries, coil, TIM6 and hall sensor.
Do not use any wires longer than about 3" ANYWHERE.
When I mention HEAVY wire here then #18 or so would be recommended.
SEE IMPORTANT EDIT BELOW
Bolt a #18 wire securely from the engine block to a terminal strip right beside the engine. Connect your battery negative to that same point.
Mount the coil right there beside the engine and terminal block so that all wiring to the coil is also SHORT and #18.
Use another terminal strip screw to connect your battery positive to. You can mount a switch on a bracket right there beside the positive terminal and bring the output to another terminal. SHORT and HEAVY.
I've forgotten the exact connections for the TIM6 but you get the idea. SHORT and DIRECT and heavy wires for all ignition wiring.
For your hall sensor you can / should twist the wires but all those wires should go DIRECT to the TIM6 module for power ground and signal. Of course keeping them well away from the HV wires of the ignition coil.
The hall sensor should be well insulated from the engine bock however it's mounted.
The TIM6 board should be connected to the power and ground terminal strips as required. with short #18 wires.

Spin the engine over with an electric drill and appropriate mechanical connections. I think you mentioned you have a sprague clutch available. Make an adapter to go in the chuck of your drill.
You could continue using the lathe arrangement with this simple ignition but it will become an issue later when you start adding the Arduino etc. The VFD will play havoc with it and it will be a nightmare to try to eliminate those issues. Best to eliminate the lathe issues now. If the engine is willing it's not going to be hard to start so there's no need for the elaborate lathe drive setup. Your going to want to mount it in a compact setup anyway so start now.

Also:
I believe you might have been using a buck converter to supply power to your ignition circit. It is unlikely that will work. The buck converter will not be able to supply the high pulse current required.
I would suggest a basic ballast resistor arrangement. I'm not familiar with the coil you have so I can't suggest a value for the resistor but the resistor in series with the coil power connection enough to drop the voltage from battery voltage to the coil recommended voltage when it's energized is a much simpler solution and it can be implemented -
You guessed it - with short heavy wiring.
KISS.


IMPORTANT EDIT:
Did you mention you are using a waste spark coil?
In that case I don't think you NEED to connect the engine block to battery negative. The block can be floating.
The spark energy is trying to travel from one HV lead to the other and uses the bock as a path. It's not looking for battery negative to complete the HV path.
You can make changes as required depending on this.

 

[Charles Lamont]

Guys, thanks very much for all the suggestions. I have a lot to think about, and quite a bit of work to do. I will be back when there is something to report.

 

[Charles Lamont]

Soo. Listening to the advice above, I have isolated everything from the lathe, and that seems to have eliminated that noise problem for now. I also took away the buck regulator power supply for the ignition and wired it directly to the 4.8V battery pack. I fitted a big aluminium bar to ground everything to, with short, fat wires. Meanwhile new SS443A Hall sensors arrived.

Like that, I was getting a reliable tacho reading with the Arduino from the motored engine - so long as the coil was not connected. Connect the coil, and the electronics did not like it. So, get rid of all that for now.

Down to a simple arrangement of the TIM-6 ignition, coil and battery. I got the engine to cough a few times before the second hall sensor died. With a new sensor in place and lighting the ignition LED - nothing. Took the plugs out and fitted them in an earthed plate. No spark. I thought I might have knackered the coil.

An email discussion with Julian at Minimag revealed that the TIM-6 is OK for small coils such as the Modelectric which might produce 90 volts on the primary, but higher output coils like the Novus 2 can produce 300 volt pulses on the primary which over-stresses the 100 volt rated TIP42C power transistor on the TIM-6. He advised that a Minimag MIC-1A ignition unit would be able to handle the coil.

Well, the new ignition unit arrived in the post yesterday, and produced a spark on test. So the coil was fine and the TIM-6 buggered. The engine was running by early afternoon. The third hall sensor has survived so far.

It would seem that much of the problem may well have been that a TIM-6 is not up to driving a Novus coil.

 

[dsage]

Sounds like you are making some headway with the suggestions.
When you say "Earthed" what do you mean?
"Earth" as in a connection back to AC house ground or anything similar should not be any part of your circuit.
Everything should be connected back to the battery negative post with as short and heavy a wire as possible if not bolted directly to it. Especially the engine block. Using the term "grounded" is not great either but in most cases it is understood to be CIRCUIT ground or the battery negative.
Please don't use the term "Earthed" it is misleading.

Your sensor is likely blowing because (as mentioned earlier) the spark energy is finding an easier path from the engine block back to the battery negative through your sensor wiring rather than through the engine block. Your sensor wiring should be well insulated from the block and should be as short as possible to the ignition module. Long wires could also pick up electromagnetic interference taking out the sensor.
You likely blew the transistor on the TIM-6 module. As mentioned the kick back from the coil probably exceeds the rating on the transistor.

If you want a good ignition circuit please look up the Sage / Gedde ignition here on HMEM.
https://www.homemodelenginemachinist.com/threads/a-new-ignition-circuit.20415/Our coil driver driver uses an IGBT transistor that is specifically made for automotive ignitions. You'd have to build the circuit yourself. The complete circuit was detailed in Model Engine Builder magazine with build instructions and circuit board layouts and parts ordering list. The circuit is pretty much blow-out proof and will drive any coil you want up to a full sized car coil. I gave the circuit board files to Mike Rehmus the editor of the magazine . I'm not sure if he had some made that might be available.

 

[Charles Lamont]

Earthed, grounded = battery negative - my sloppy language. I will try to be more precise in future!

I think it was probably the TIM-6 that blew the two hall sensors. The sensor always has been well insulated at the engine, and the engine connected to battery negative (the latter less important anyway, as you pointed out, with a twin-spark coil). The sensor has been fine since fitting the MIC-1A ignition - touch wood.

 

[Tim Wescott]

At least among British car people, "earthed" seems to mean the same thing as "grounded" means to an American. Could just be a regional dialect thing going on here.

What matters most is what you do, not whether folks in your part of the world describe it differently than they do here.

 

[Tim Wescott]

You could find a different output transistor for your TIM-6. It's getting hard to find good high-voltage bipolar transistors out there (but there's probably some). There's plenty of FETs that are good up to 600V, though.

 

[dsage]

Ah yes. I forgot about the waste spark coil. In that case the conection to the block is probably not required per se. BUT if it sparks ok with the connection then leave it on. It might help with shielding everything from EMI when you get to hooking up all the other stuff. You'll have to play with that if you have issues. Certainly on cars the block is connected to ground for waste spark systems but things are a little more organized in a car.

 

[dsage]

You could find a different output transistor for your TIM-6. It's getting hard to find good high-voltage bipolar transistors out there (but there's probably some). There's plenty of FETs that are good up to 600V, though.

My circuit uses an IGBT transistor. That's what they use in automobile ignitions.

 

[Charles Lamont]

Thanks for the massive help. I don't think I need to go any further in terms of the ignition at the moment. Getting back to the electronics will have to wait until I have a coolant pump, and possibly the starter motor. See my post at Westbury 'Seagull'