gr8flyer55

Hi Ed.
It is using the residual charge in the .47uf capacitor. As it recharges from the hv output from the transformer, it draws around 25 mah more. Not too bad for the total draw of 250 at idle speeds. 350 to 375 mah at 7,000 rpm.
I am getting easier starts, even on a partially flooded engine from over choking. It was just a test to see if it would start, and it did very easily. It does kind of diminish the mid range burble too, a side benefit. It seems to be power efficient in the fact it's now using the left over charge in the cap. It involved adding a 3 spark loop in the ASM file and now with the faster speed of program execution, the loop takes very little processing time.
I haven't noticed any extra heating on the oscillator transistor either. Looks like a win- win situation.
My test setup has been running for a solid 3 days on a 5 volt power supply and still firing away at 3,000 rpms with an ignition tester hooked to the input signal replacing the hall sensor. Yesterday I had the new program in my test plane's ignition. Flew great with great transition and smooth idle. So far so good!

John

CH Ignitions

My Feedback: (11)

I have not seen the code, But this was my Idea for the CH RainFire TM ,I think I told you about it.
After intial fire based on the RPM and executed delay from the table,
you can have the code to trigger the SCR one more time before the magnet gets around the sensor.
The CH can spark up to 25000RPM easy.
Now you can create the code that between 1500-4000 to give you 5 (20000/4000=)5 sparks and after that you decrase the # of sparks to 3 ,2 1 ; till
you get the top 10000 and keep 1...
Are couple ways to skin the cat....LOL
get a high performance pic 18fxxx and have all the calculations on the fly;
create an algorithm that even on 12F683 can work, get the code to execute X Sprks equaly spaced on uS right after master spark.

Thanks
Adrian

gr8flyer55

We are still using the 12f683 chip and it is making 3 sparks per trigger. Processing time was increased to around double speed and is working great so far. Flew my plane with it and got great results.
Forum member Jakestew is rewriting the code in C. It'll include some telemetry for rpm and temperature. Since there are unused inputs and output pins, we can take advantage of them for this purpose.

John

bluejets

Theory is fine but I think you would have problems getting enough energy into the capacitor in such a short time.

John,
You mention increased program speed but I'm pretty sure it's still the same, just added the 3 loops from what I can see.

Ed Vollmer

My Feedback: (14)

Hi Adrian,
You may have told me about it but I don't recall. Way back when, we were using .82uF up to 1.2uF Metal Polyethylene Caps on some of the UAV ignitions to get the maximum energy transfer to the output coil. The higher the Cap values, the longer it took to charge them fully and using up any residual charge for a second firing would make that charge time longer yet. At the time we decided it was best to fire one very substantial spark and allow the Cap more charge time between firings.

Multiple spark may work fine with the lower Cap values. It would only be a problem at higher RPMs anyway. MSD Ignitions have used this method for many years.

Any news on the WYK carburetor testing?

Gompy

I am very curious how it's possible this small transformer start charging the capacitors to generate 3 sparks for one ignition.

Muciek

Key to this is good DC converter http://www.youtube.com/watch?v=zNcA0oYLj1s http://www.youtube.com/watch?v=OhG4M2WEM9w
here are my projects. This one used in motorcycle can start from 7v generating huge spark I had at that moment not charged battery (voltage 0,5V yea I know it's dead ) so it works good with voltage straight from rectifier current consumption is like 1,5A at max and 60mA at idle. Small transformer used there  can withstand over 35W of power!! Converter is very high efficient around over 80%.

jpanhalt

Since we don't know the exact circuit used, some of this comment is guesswork. If it uses an SCR for the trigger, as most single-spark designs do, such devices stay open as long as there is current flow. At least, they stay open until the current falls below a keep-open value, which is usually quite small. Thus, once fired, the SCR should stay open until the spark extinguished itself. Remember, the electrical resistance of the spark is quite low compared to the resistance of the fuel-air mixture before it is ignited.

How will "residual" charge on the capacitor have sufficient voltage to re-establish another spark when the SCR is re-triggered? If in fact, the capacitor is re-charging, then Gompy's reservation (post #181) about charge rate needs to be addressed.

Does anyone have oscilloscope tracings of the voltage across the capacitor to show what is happening in these multi-spark systems?

John

gr8flyer55

A scope trace will be posted showing the spark count and intensity. Charlie Moore (COM), has the new MSD hex file for testing at the moment and will be able to do a show and tell of the operation. System requirements are a strong inverter circuit and charge capacitor, and a heavy duty ignition coil. The CH ignition board is one of the designs I have been testing and has the required inverter circuit. Preliminary tests on that ignition have proven very successful in my plane. Don't go expecting over 10k rpms with the same spark you get at lower rpms. Most model ignitions will have diminished spark at that rpm because of the small ignition coils used.
Possibly I can get Nyemi to comment on this in a forum post.

John

Gompy

Multi-spark with a SCR is impossible.

captinjohn

Nothing is impossible with electronics.

nyemi

Hello to everyone
Very possible. The key: the thyristor gate time.
Good high-efficiency inverter.
The data from:gorbunov.narod.ru/CDI_ST.pdf

Watch Video:www.youtube.com/watch

jpanhalt

That's helpful. It appears that rather than using residual charge on the capacitor, you are capturing the reverse EMF from the transformer with the added diode.

I don't recall that diode in the earlier circuits.

Can you post the circuit you are using for this? What part numbers are the SCR and diode that you are using?

John

Edit: Here's a direct link to a more recent version of that application note (AN819): http://www.st.com/internet/com/TECHN...CD00003947.pdf

Gompy

This schematic / CDI is not the same John have made !
Thats why I say with a SCR, in Johns schemattic, multi-spark is impossible.

nyemi

Hi Gompy

John cdi is even better multi-spark.
Because there is a capacitor charge-back.
(Capacitor receives energy from the ignition coil primary.)
He sparks up time: maximum 200uS
But now working on: 3*200uS spark .

Gompy

I don't understand the changes you have make, sorry.

captinjohn

Gompy....Maybe you need too do a different hobby~

Gompy

Is this a compliment for all the work I have done the last 8 years ?
Nyemi, now you know wat they thinking about you if there will be a "new" guy in town.

Goodby to all, I quit.

nyemi

No no! Send a personal message.

gr8flyer55

here are the scope traces at various rpms..first is 1000 rpms then 2580 then 6900 rpm.

jpanhalt

Where is probe 2 attached? Is that the firing signal from the PIC? If so, that does not mean there is a spark for every trigger signal.

Do you have tracings at the capacitor perhaps across the capacitor? Or, either end relative to ground/common/negative battery would help. Can you put a small coil pickup on the HV lead to the plug like an old fashioned strobe timing light to see when sparks actually occur?

John

gr8flyer55

John, I'll have to let Charlie answer that question. He is the person with the scope testing everything. I know I am not losing any performance with it, so far I have 10 flights with no problem at any rpm. If I hook up my tach to the led output on my timer, I get an almost tripled reading but I don't think the optical tach is fast enough to register all the firing taking place. I'll leave that testing and results to Charlie on his scope. I am curious also since I didn't write the routine for it.

John

jpanhalt

I don't believe the LED monitor is anything more than a trigger pulse monitor. That is, it really doesn't measure the spark.

John

gr8flyer55

This is true John. That is why I said it's not a valid test. Charlie will have to test it for us, although Nyemi has already done this and says it is producing the multi sparks. I trust this will all turn out to be true. Time will tell !!

John

ngopaul

Very intersting with all your post. Say again, I think we just focus to multi spark (at least two) when engine start in cold weather condition only, that means if engine good start then when it ready go to advance (180-970RPM)
multi spark will be shut down, this time it running with single spark. Why we need multi sparks for cold start? I converted Ryobi SS30, using Nyemi' schematic, build PCB, loaded Rexel'Hex, engine just run after one flip,
but winter last year, I has flip one hundred time to make engine start in 35 deg.F, after that I explore if I connect signal out put from Hall direct to gate'SCR it will produce two spark: one direct from Hall and one from
12F683 out put, then engine start very easy, but it lose advance until I disconnect direct signal. I'm not coder, but I think it can be done with software, and why it needs to be shutdown multi sparks when engine go to
advance mode:
-Small engine'chamber don't need multi sparks to completetly burn-off fuel mixed.
-Multi sparks reduced spark plug life and wasted more battery current in RC engine.
-We don't need to replace power invert circuitry because in slow speed (180-970RPM) the capacitor enough time to full charge before discharge multi sparks.
Bottom line, we need multi sparks for start engine only.