Gompy

I can't go back with my Homy to normal......there is no ignition anymore.

COM

Well Rob,

You should not have to. You have a pretty cool setup.... don't you?


Jake and all the rest This circuit has been around for 30 years or more and it is pretty bullet proof. I think that if something is causing the transistor in Johns HV board to draw too much amperage then it has to be in the software. I have 50 timer boards by different people and not one has caused this sort of problem. It all comes down to how you control the pic output. I will say that the commercial timer boards that I have do not work like this software.


Jake I sent you some screen shots of the input vs output from a commercial unit. and they all pretty much work the same and none have ever caused this sort of problem. I will stand by and see what you guys come up with.

bluejets

You are still looking at the circuit operation the wrong way.

If you go to the drawing from the web site, look at the section "differentiator"
(which is essentially the Q2 collector connection in the HV board)
and ask yourself, is the collector going from
1........ low to high state (Q2 turning off)

or is it going from

2........ high to low state (Q2 turning on)

if it is 1, (going from low to hi state) the rising edge of the first square wave input,
then the corresponding waveform pulse is the first (positive) pulse at the resistor cap junction.

if it is 2, (going from hi to low state) the falling edge of the first square wave input,
then the corresponding waveform pulse is the second (negative) pulse at the resistor cap junction.

The SCR needs a positive pulse so the result of the above is No.1
How does the Q2 collector go from a low to high state..?? by turning it off.

Gompy

I don't think the Homy is a problem, I hope befor I reassamble the Homy the "transistor problem" have be solved.
I think it's a small problem, but we don't see it for some reason.

COM

What resistor/Cap junction?

bluejets

SCR gate

gr8flyer55

I'm having my neighbor look at the source code tomorrow evening. He is a programmer for imbedded microchip systems in air conditioning systems. He knows a great deal about electronics also as he is a design engineer. Just showed him the source code printout tonight and he pointed out a problem with an interrupt for the sensor input. Things are looking up already! I'm giving him an ignition test setup so he can hook up to his scope for signal tracing etc. I've seen some of his work around the house, and everything he's built is wireless control, so I have great expectations for him finding a solution to our software problem. He may ask to be in the credits after the problem is solved, and I don't see anything wrong with that at all.

John

jakestew

That's the beauty of open source. If someone else can find the bug before me that would be great.

I ran my engine quite a bit today. I loaded several different curves and it worked great. Then on the last flash the PIC started keeping the LED on for some reason. Still ran good with no transistor overheating. I thought I was getting close to figuring it out but I ran into some problems with my Hall connection and my processor socket giving poor connections.

What I do know is that the simulator isn't showing these problems, and on the test/programming board the LED output is not on. Only when it's in the circuit does it seem to give problems, and then only apparently with some curves or compiles of the firmware.

So, long story short, I'm still not sure what's going on. I had to rework the timer board a little, and the hall sensor mount on my engine, and my processor socket. I should have some more time to test tomorrow.

Whatever is going on is very strange and random. It's not a traditional bug in the software, but something screwy is going on and needs to be fixed somehow.

I know the continuous fire when sensor is activated bug exists since it happened to me that one time. I can't get that to repeat, but I've got a bug where the led stays on happening at the moment and I'm pretty sure the two are related.

In any case, it can't hide for long.


-Jake

bluejets

Jake,
Software simulators are great, they will prove your design operation.
What they cannot account for is real-world interference and the ignition systems would have to be one of the worst.
I think everyone here has come a long way in a short time and credit is due to all of you.
Good stuff.[8D]

COM

Well John,

Make sure he knows how the system works.

Gompy

Hit the prop with two fingers and be sure thoose two fingers are gone befor the engine start

gr8flyer55

I'm feeling confident he'll be ok testing things. About 20 years ago, his dad was a very active member of our airplane club, and Rich was right there with him preparing the models for flight. This guy is ready for anything I can give him to work on. His young son is very interested also in models. We might just have a new forum member from all of this!

We have to install the compiler on his computer because he mentioned he would have to download the one we use, but he will take it from there. Hopefully when he can see this constant sparking on the test setup he can give us some solutions.

Well, my parts order came in yesterday so I can bag up some more kits. I'd like to etch more timer boards, but will have to wait and see if there will be changes made to the circuit for the 12f1840 chip and the new software for it.

John

gr8flyer55

My neighbor has the source file printout now. He was busy this evening with family matters but will be looking over it for Friday. By the next day, we should have some revisions of the sensor input routine and the signal output processing.
Are we having fun yet?

John

Gompy

Houston we have a problem, the cylinder with nitrous oxide is broken......hahahahahaaaaaa.........BOEM

gr8flyer55

Post# 561 is where the problem is! Neighbor says it isn't turning the output off at low rpm routine.

John

Gompy

Why do you want turn the output of at low rpm, and wat will be low rpm ?
If the output is turn off at low rpm, you can';t start the engine anymore.

BTW, I hope you like my (dry) humor and don't get angry of it.

gr8flyer55

Rob,
The ignition I sent you for your Homie has Nyemi's good software in the timer board. Just plug the ignition and time the ignition and run it. The ignition #2 I sent also has Nyemi's software in it for the Aprilia. Hookup and run, no problems.
Extra boards and parts are for you to build and test jake's software on. Plenty of parts to build 2 more. Hope it makes it thru customs inspection ok with all the wires in the box.

John

Gompy

Mmmmm, I can start my own webshop with all the stuf you send me.
I think this can be a problem with customs, I hope you write on the pakkage "Sample, no value"

gr8flyer55

The problem as seen in post #561 by Nyemi's scope trace clearly shows that the signal is not being turned off, causing the high current drain on the HV board, heating the transistor like a dead short. The low speed pulse signal, below 200 rpms is the trouble spot in the sensor input routine. There are about 10 lines or so, according to my neighbor that are suspect to cause the condition. The signal must be turned off after every trigger in low rpm mode or this problem, although a random occurance, is holding the SCR gate connection and draining the charge on the capacitor a cross the coil leads. There is a positive way to correct this and a solution to the latch up. Could be fixed by the weekend, as soon as he gets time for his scope hookup to test it.

John

gr8flyer55

Hi Rob.
One of the HV boards has all the parts on the board, but I just bent the leads over to hold the parts in place. Be sure to check parts placement since I did it quickly to pack up the rest of the box.
I like how you trimmed the excess crankcase material from the Homie !! A first class job as always, looking good!
I just use a small grinder, but you used a lathe or milling machine the way the finish looks. My engines are always done in a hurry so I can test, I never take the time to dress them up like you have done.
My hall sensor gets mounted with a worm gear hose clamp so I can adjust the sensor timing, but it holds up very well for me. Other engines have places to use a bolt on mount which I will try to post a picture of for you. Again, not only way to do it, but it works.

John

Gompy

Becaus I have all the time untill ariving the parts, I can spent more time to the Homie.
The big work is now done........and everthing is now stopped for a 3mm (?) magnet I missing.
The smallest magnet I can get is a neodeen 5mm super magnet I can't use.
I have to find out where I can get a normal 3mm (?) magnet, maybe little train shop ?

It's a small lathe, 40cm (16") between the centers and the mill is a cheap Chinese one.
Nothing special, but I can do wat I want to do with the toys

BTW John, our biggest problem is te get the engines.
We use allmost electric engines for jigsaws / weeteaters / and other engines where you use petrol engines.

gr8flyer55

I think I sent you 2 of the 3/16" diameter x 1/16" thick rare earth magnets in the box. I buy them 2 at a time from Radio Shack. Not sure if you have Radio Shack stores in The Netherlands. Either way, you will have 2 of them when you get the goodies!!

Nice equipment you have there Rob! I have a small Harbor Freight 7 inch lathe, and a Taig Micro lathe which I use as a milling machine.



John

Gompy

I think I buy for 30 years ago the last parts from Radio Shack.....when he was closing.
For 5 years ago we have 4 electric stores into town, now only one.
Last year the last RC-models store is closed and we have a town with 95,000 people.

BTW, the machines are second hand, I can't permit me new.
The CDI's cost me all my money, I can't permit me even a cruise to Hawaii

I start also with a Harbor Freight 7 inch lathe, but it was to small to modify my 50cc oldtimer bike.
I start with a 1970 Kreidler who never see a repairman and rebuild it to new.
http://www.kreidler.gompy.net/Kreidler-TRS/index.htm
When I finish the bike I buy myself a littlebit bigger lathe and now I modify smaller engines

nyemi

Hi Gompy
"You can't make spikes with this kind of transformer"
Yes.
The ignition signal all dirty.
Take precautions against it, and you'll have no problems.

"The SCR can only shutdown if the Voltage is going true zero."

Do you think 'ZVS CDI "SCR is not turned off?

Gompy

HI Nyemi

Thats wat I mean, the flyback of John don't have any of it.
I think we need first clean the signal, more capacitors, resistors and diodes.

The SCR have to be turn off, but it will not shutdown if the Voltgae is not going true zero.
Put some Voltage over A and K, open the gate, release the gate Voltage and the SCR will stay open.
Only if you make the A / K zero the SCR will be closed.

Using a CDI for RC, I like the one from John.
It's simple, less components, small and low Voltage.
Using a CDI for my bike, I like your designe ZVS DC CDI.
It looks to me reliable, but it have also more components......and John say it by his self, your software is at the moment better.