gsl-se addict

iTrader: (3)

Sounds good. Will look forward to hearing the results. I still don't know what is up with your circuit. The only thing I can think is that the max pulsewidth is causing a delay of 4ms instead of a limit. I don't see how that is possible at all, though. I figure, with the 20* or so retarded, is like 4ms at 800 rpm. This would also show an apparent retarding of the ignition as the engine turns further within that given time at higher RPM. Doesn't make sense that it could be happening, though. The only way to know if someone tests with an o-scope/function generator. If I had the equipment, I could probably have this thing figured out in a matter of minutes.

Make sure to limit the amout of idle/low RPM use of the TT without R2. I don't want you guys to burn out a coil, especially since they are so much more expensive there. If you get a chance this weekend, try a large resistor in R2 (like 100k) and then try 22k. That may help figure out what is going on.

Kent

H4Inf

Kent,

Around what rpm will the TT's coil pulsewidth output reduce? My idle is at 1200rpm, and about where does the Min limit become active?

Cheers

gsl-se addict

iTrader: (3)

Well, without R2 you have a minimum limit (2.2ms), but no max limit. I am not sure the actual RPM where the 2.2ms takes over. I would guess around 5500-6000 RPM. Without R2, the circuit should behave mostly like the simple circuit in the lower RPM range. The J-109 puts out something like 6.5ms pulses down low. It drops off at higher RPM, but I think at 3 or 4k, it is still like 6ms. I will double check the number. I am going by the numbers Jeff20B posted about the dwell duty cycle.

gsl-se addict

iTrader: (3)

Based on these numbers, I get the following pulsewidth:
1000 RPM, 7.33 ms
2000 RPM, 6.45 ms
3000 RPM, 5.40 ms
4000 RPM, 4.35 ms

Above 4000 RPM, both the dwell % decreasing and decreasing due to higher RPM (less time between ignition events), the pulsewidth will drop faster beyond 4000 PRM. When I get back, I can do some testing on this or anyone with access to an o-scope could check these values. You can see that idle/low RPM is going to cause the most heating/strain on the ignitor.

H4Inf

Alrighty! Rob and I spent a good 3-4 hours tweaking and playing around on the dyno.

We couldn't have done it without Stuie of BA Performance (which I think soon will be named RX Performance) - http://rotaries.baperformance.com.au/
He basically gave us his workshop with tools and dyno and said go for it! Theres this wonderful pipe which runs thru the rear wall behind the dyno, up the wall and thru a straight truck muffler, plug a hose to the exhaust from this = less fumes
So just a big thanks to stuie for helping us out there.

Now getting onto results. First we should all remember that a dyno doesnt measure that many driving attributes - really just power - and then you also have to balance the load with the rpm you want to go to - not real driving condition like.

So the results - from memory.

First ran my 13B with ignition mod.
~100 bhp @ 6850 rpm

Next ran my 13B with stock ignition.
~97-98 bhp @ 6850 rpm

Ran Rob's stock 12A Stuber with stock ignition on same load setting as my car.
~70 bhp @ 6200rpm

Ran Rob's stock 12A Stuber with stock ignition (now reduced the load as we got stuck at 6200rpm (have to find good balance between load and rpm you want to reach).
~73-74 @ 7000 rpm

Ran Rob's stock 12A Stuber with Ignition Mod
~75 bhp @ 7000 rpm

Ran Rob's stock 12A Stuber with Ignition Mod and NO Air Filter
~80 bhp @ 7000 rpm

So conclusions are that it may add 1 or 2 hp, however there are other things I have noticed that a dyno doesnt measure I guess.. the feeling behidn driving it.

-Sometimes my stock ignition breaks up at high rpm, whereas the TT mod does not.
- It feels like you get more bang for buck - a better response - ie going up hill in 4th gear, it feels like you get a better response on TT vs the stock ignition - see Rob now has the loom we made up and I had to drive home on stock ignition (been driving TT for a week) and it didnt feel as responsive under load conditions.

The only remaining problem is the coil pack heating up - apparently it shouldnt get too hot to touch - so we do need to find out what the problem is with the latest circuit design in relation to R2.

Didnt get any photos/video etc. 37.5 degrees C (100F) in the workshop (not ideal dyno conditions.. well I guess the tyres will stick at least lol)

So that's todays findings. Aims for next weekend are to find a function generator and CRO and see what we can figure out about the latest circuit.

Cheers

Paul.

gsl-se addict

iTrader: (3)

Cool, Paul. Now we know what kinds of gains to expect. I figured it wouldn't be too much. But as you said, the behavior of the car is much better with the TT. Idle seems smoother, engine revs faster, and so on. A few guys reported an increase in fuel economy and I bet emissions are better. The dyno doesn't tell the whole story, but it looks like there are some gains.

Once we figure what is going on with the circuit, I will be able to start shipping the kits/assembled units.

Kent

devitek

Once I get the board and a new coil pack, I'll be happy to do some testing. The short, one-week experience I had with the simple circuit was enough to sell me on this mod. I've got several people around here that really want to know what it's all about.

Kent, is there any kind of postal tracking number available on that shipment? I'm fairly excited to get this all put together, provided I can find a coil pack before then.

gsl-se addict

iTrader: (3)

Dave,

No there isn't any tracking. It is hard to do any international tracking unless you use like FedEx, UPS, and so on. FedEx charges like 62+Euro to ship to the US for packages under 1 pound. Yours will come through USPS. I send it using laposte (France's postal service), they then hand it off to the USPS. It takes about 5 days to ship to the US. Sometimes it is longer depending on customs and such.

Kent

devitek

No rush really. Still gotta track down another coil pack first. I was just curious about the tracking thing. I've ordered things from Canada before, shipped postal, and there were tracking numbers. Never hurts to try!

H4Inf

Alrighty, Rob and I did a little work today, however we found that the function generator I had was insufficient for our needs - I will have a go this week at work with a more complicated one for better results.

So basically today we ran the circuit on the car and recorded some data.

First thing to note. This following image ive made up from 2 stills I took of the CRO today. I had the CRO connected to earth and the output to the 2nd gen ignitor/coil pack. This was happening erratically, switching between these 2 states. I think this is the effect we see of the ignition timing jumping around when we look at the pulley with the timing light. The below images were recorded with a steady idle at 950 rpm on the latest TT circuit without R2 connected.
- Each major horizontal division = 5ms in this image.
- Each major vertical division = 5volts. This jumping back and fourth was about 3ms.

It bounced around, kinda looking like this - best i can re-produce on pc:


Now I've attached an image of observed data: (all times in ms)


Basically without R2, the circuit just seemed to invert the + and -..

The idle was not as smooth with R2 reconnected. We didnt have a chance to check timing/advance situation as neighbours had complained multiple times by this stage(trying to sell their house)- thus why the R2 data is a bit sparse.

Didn't get a chance to look inside the circuit and probe in there, hopefully will this week @ work.

Just wondering, does R2, which basically runs between the connector out to coil and ground... do you think that could be adversely affecting the 2nd gen ignitor module? Does the data above make any sense?

Thoughts would be great!

Cheers

Paul.

gsl-se addict

iTrader: (3)

Hey Paul,

One thing is we need to make sure we are talking about the correct resistor. The schematic I posted for dj55b does not have the came numbering of parts as on the board. I just quickly made that layout for simulation purposes. The R2 is marked on the board. It is near R1, R6, and the 2 caps on the top/right side of the board.

Was that jumping on the input from J-109 as well? Sometimes scope traces will move around like that unless you do something about the triggering on the scope.

Those numbers from the J-109 seem higher than I would expect. No wonder the 2nd gen pack gets so hot without the pulse limiting. I never had the change to run the scope on my ignition to see what the J-109 was doing. I don't understand why the pulse is still so wide at higher RPM.

Does your o-scope have 2 channels? It would be helpful to compare the input from the J-109 and the output from the TT. What is wrong with the function generator? You only need a very simple one for testing this circuit. I even did testing on mine before with just a 555 timer chip and a couple resistors and a cap. You just need to put out a square wave between like 20Hz and 350Hz. Ideally it would output pulses 0 to 12v, but that is not totally needed.

The J-109 takes a 12v signal and pulls to ground when firing. I imagine that for your data, you were measuring the time low for the J-109 (or did you measure time high).

If you have a function generator, I can guide you through testing. No need to run the car and bother the neighbors that way. It will also allow us to check the pulse limits. Maybe you can send me a pic of the circuit as well so I can make sure everything is in the correct spot.

Hopefully we can get this figured out so I can get the problem corrected and get the other circuits shipped out. If anyone else can help in the testing of the circuit, I would really appreciate it.

Thanks for the help.

Kent

H4Inf

Hi Kent,

To answer above questions:
1) R2 to me is the resistor which is the middle one in a row of 3, marked R2, near 2 caps.

2) There was no jumping from the J109 input, it was smooth.

3) I am pretty sure I read the CRO correctly and those numbers are accuarate.

4) OScope does have 2 channels - I just need a 2nd BNC connector + probe - can do this week.

5) Function generator does not have an offset setting. I connected it to the trigger input for the TT circuit but i could not get it to register a pulse at all - the only settings for the func generator are sine/square and then you set the MHz. Thats it. Rob said he can build up a proper func gen - I have a better one at work which I can't take home so will try to get testing this week. ( holidays :-) )

6) Correct, I was measuring the low time, the high mark was very small.


7) Pic of the circuit: (R2 is bent up (disconnected))



Please add me to msn: paulw9_1@hotmail.com so we can make progress faster

Cheers,

Paul.

gsl-se addict

iTrader: (3)

Will do, Paul. I need to run down to the store right now. I should be back in like an hour. I will hit up your MSN then. Hopefully we can trace this down.

Kent

H4Inf

Sounds good, Rob and I will be available then..

Oh and just to add, I have found more fuel escaping unburnt down the exhaust pipe since I have changed back to stock ignition from the TT circuit - increased frequency and intensity of popping out the back. This verifies that the TT circuit is helping to clean up emissions (well at least emission of unburnt fuel).

gsl-se addict

iTrader: (3)

Okay, guys. Paul, Rob, and myself discussed the circuit and did some testing. It appears that the TT circuit works just like it is supposed to on the bench. There are a couple things we are going to check on the bench and then go back to on car testing. We will be doing some more testing Sunday morning Aus time. We will post up the results of the testing once we have all the info.

Arockrx7

Cant wait. Good testing and good luck

H4Inf

well we did some testing and preliminary results show that the ignitor/coil is somehow causing interference in the TT circuit. Further testing will take place this week. If anyone at all has detailed knowledge of the workings of the 2nd gen ignitor/coil or in general please contribute any info that may be of use!!

Cheers,

Paul.

Whanrow

I'm the farthest thing from an electronics expert, but when you say interference, I presume you mean RF interference? If so, isn't there some way to put a filter in there much like you would for a stereo? Personally, this may sound stupid to you, just trying to help with an idea....

H4Inf

We are not sure right now if its RF interference, or whether the ignitor/coil is giving back electrical inteference in reverse through the connecting wires. It would be interesting to see on a 2nd gen ECU where the output to the coil connects and see if there were any inline diodes/components that may stop this sort of interference.

I am going to try a diode inline with the output to the coil to ensure that it is only a one way transaction. Below I have a few images recorded, the first with the coil disconnected, the 2nd with the coil connected.





The divisions are 2ms. The probe is connected to the coil output from the TT.
When the coil is attached, the pulse seems to last an extra 2ms - this additional 2ms looks a little different on the scope though, and it does drop to ground for a minute amount of time before jumping back up for this additional 2ms.

I think to avoid rf interference, wiring and the circuit would need to be shielded with foil, and the foil grounded.

I will be conducting more testing this week, but if anyone has any ideas at all - definitely contribute! They may spark some discussion leading to a resolution

gsl-se addict

iTrader: (3)

In the pics Paul has shown, you see the circuit is being retriggered with the coil connected. The first pic shows a pulse of about 4ms (max pulse limit of the circuit). With the coil connected, some feedback is causing the circuit to retrigger after this 4ms, producing an additional minimum pulse limit (2.2ms).

In some of the other tests, Paul and Rob found that there is a strange signal riding on top of the high pulse under certain frequencies with the coil connected. This behavior shows up on both the input and output of the circuit. This may be due to the "ringing" the coil produces on the coil - terminal during discharge of the coil. This signal is somehow going back through the circuit or the ground. Maybe testing with the o-scope connections further from the discharging spark plugs would yield different results.

We are going to 1st try a diode inline on the output of the TT. This should keep the coil from producing feedback into the circuit. If the diode works, it can simply be added to the output wire without changing the circuit itself.

Thanks for being patient, guys. I know many of you are wanting your circuit. I am wanting to get you the best working circuit as possible.

Thanks.

Kent

Frankenrex

iTrader: (1)

It looks like maybe you're getting some "ringing" from the coil, basically inductive kickback from when the coil field collapses. Just a guess based on some of the discussions from the MegaSquirt forums - I'm far from an electronics whiz. I know that there's sometimes a problem getting a good tach reading from the (-) terminal of the coil because of this. The MegaSquirt folks use a capacitor to filter it, known as the "Dave" cap. see here: http://www.megasquirt.info/v22manual/mwire.htm#ignition
I'm not sure that'll help, as this is dealing with a tach reading rather than triggering the coil, but it sounds like a similar problem.

devitek

I got my board on Friday, and have all the components minus the LED's and the wires soldered (need to pick up a box for it still). I should have a coil pack today or tomorrow. So one idea is to put a diode inline, another to put a capacitor inline. The diode makes more sense, what type of diode should we go for?

gsl-se addict

iTrader: (3)

I think the diode would be better. A capacitor would round off the waveform. Probably a 1N4148 signal diode would do. The only thing is that the diode will drop the voltage ~0.7v. This means our high output will be like 3.5v. That should still be plenty to trigger the 2nd gen ignitor.

devitek

I'll try to pick some up tonight after work, when I get a different project box. Hopefully my coil pack is in tonight as well so I can set it up.

devitek

Ok, aside from the problems (backwards mounted LEDs), here's what has happened so far.

After installation, I started the car with the trailing ignitor disconnected, just to be sure it was working. Turned off the car, plugged in trailing, started again. Idle was pretty low, so I advanced the timing a bit. Didn't make a lot of difference. Went out for a test spin. Backing out of the driveway, the car died about 3 times. Once I got it on the road, it seemed to have very little punch at lower RPMs but would pick up, but it wasn't steady. On deceleration, you could really hear the missing again. I checked my connections again just to be sure, but it was dark and late. After work today, I'm going to try playing around with it some more, and I'll also install the diode inline with the output to see if that helps.