gsl-se addict

iTrader: (3)

Maybe if you alter the balast a bit, that may be able to burn through that mixture. For instance,an additional 10 Ohm resistor will drop you from 2.5 Ohm to 2 Ohm. Couldn't hurt to try. What plugs are you currently using? I personally like the 2nd gen plugs, but perhaps for your setup, something with a smaller gap would be better.

bad 83

iTrader: (3)

NGK B9ES........I was told by quite a few turbo gurus that these would be best for the blowthrough. They are a single prong that can be gapped alot tighter to confine the spark, and prevent the spark from being blown out at higher rpms. I know that my A/F is rich, but the girl screams through to 8 grand (didn't mean to do that, never goes past 7500). I'll try the extra resistor. I will say the current amount of resistance I am using does burn off alot of heat to the coil. You wouldn't won't to touch them after a long haul. For the SA I went simple direct fire with the second gen plugs. Made a world of difference. Enough that I could use my 600 Holley on the stockport engine.

12acole

I just spent the last few days reading through the entire thread. As an EE, it was an interesting read. I still need to read up more on how exactly the igniter and coils work, but I got a basic understanding through your thread. I had a question though. You state that the coil fires on the trailing edge of the input signal. This was the problem with the max pulse limiting causing the coil to fire too early. Wouldn't this also be a problem with the min pulse limiting causing the coil to retard at high rpms or am I missing something? The results people were experiencing with the early versions of the circuit don't validate this which seems confusing. Sorry to bump this thread without updates, but my post count is too low to send pms.

Thanks,
Cole

gsl-se addict

iTrader: (3)

No problem. Here's the deal. The early version simply inverts the square wave and has it go between 0 and 5v instead of 0 and 12v.

Normally, the stock ignition works like this:
- igniter pulls coil (-) to ground. This causes current to flow and charges the coil. When the igniter releases the pull to ground, the coil discharges. The igniters also have a built-in current limiting mode. See, when the coil is fully charged, if you did not have this, the current continues to flow and all of the energy much be dissipated in the igniter and/or ballast resistor. The current limiting works by lifting the coil (-) a few volts above ground, thereby reducing current (limit is set to about 7.5 amps).

The 2nd gen coil takes a 0-5v input. When the input is high, the igniter pulls the coil (-) to ground (just like the 1st gen ignition). When the signal goes back to 0v, the ignition fires the coil. The 2nd gen igniter has the same kind of current limiting mode that the 1st gen does.

The thought to fix the overheating issue (due to excessive time in current limiting mode), was to shorten/limit the pulse to the 2nd gen coil. That works and all, but since the 2nd gen coil fires off the trailing edge, it causes the spark to occur at the wrong time. Really what you need to do it to predict when the coil needs to fire and start charging some time before that. This could be done with a microcontroller or possibly through a series of flip-flops, counters, and timers.

As to the min pulsewidth limit, yes it could throw off timing some. However, the difference is fairly small and the mechanical advance at high RPM kind of negates the affect.

It appears that the current solution is to add ballast resistors. Not elegant, but does the job for now. Hopefully I'll find some time to make something better, though.

12acole

Alright, that's what I thought. I just found it odd that those using the autoswitching circuit had better top ends compared to the simple version. I guess the longer pulse width provided more benefits even with a slight retarding of the ignition.

Thanks for the quick response and detailed explanation on the ignitors and coils, and I look forward to your future projects.

RotorMotorDriver

iTrader: (1)

So, did all the kinks get worked out on this setup? I seem to remember boxes being made and sent out for testing, but thats about it. Did you ever determine if switching between the pulse widths changes the timing? How has the new circuit progressed? Is it now auto switching?

If it is working, is there any way someone could make an updated post with the current circuit design, installation, etc and have a mod move it to the front page? That way people dont have to read through the whole thread if they dont want to.

Im planning my 1st gen re-re-build, and I had the "old skool" DLIDFIS with 109s, but I was thinking I would go this route if its worked out. Thanks!

~T.J.

EDIT: Oh, and can we just summarize the name to "Super Ignition"?

H4Inf

I assembled a whole lot of information about the setup together a while ago (maybe a year) - the info I put together is about as up to date as it gets with some exceptions:
http://rx7.pw.cx/transistor_trick/index.htm

The exceptions are that the coilpack gets overworked and ends up dying unless you replace the balast resistor in it (I think the one in there fails prematurely from heat). Also if I can remember correctly, the auto switching setup did not work out, so we just settled on fixed and adjusted a resistor for the best timing.

Sorry my memory of it all is quite vague, but basically I think we settled with the fixed circuit and a balast resistor added inside the coilpack in place of the existing one. You should be able to find what people used in the last few pages of the thread. Some people found adding heatsinks to the resistors, or externally mounting them helped too. (I ended up installing a megasquirt efi system so no longer use my TT setup - when I was running it though, there was a definite improvement over stock dizzy seen on the road, although dyno tests showed no real improvement in power, it was more the response and acceleration that improved)

If you have a spare few hours, this thread is a great read - a lot of positive development and testing went on.. I really enjoyed the development process and learned a lot

RotorMotorDriver

iTrader: (1)

Is there a write up for the auto switching version, or did I miss it? Obviously I cant get in on a group buy, lol.

I guess what I was hoping for, was a diagram for each version with materials lists, and maybe known problems for each version and/or RPM limitations if applicable and such. Just something to clarify it all. Like an "end all" thread that gets locked so that technical discussion will only take place in the technical thread and not clutter up the "how to" of it all?

Then you could say "Heres the diagram for the fixed pulse width and parts list", or "Heres the diagram for the auto switching pulse width and parts list", and then we could all pick the one we want to run based on our own decision and application requirements. For instance, someone with an engine that will run mostly in the upper RPMs for drag racing or whatever might be more willing to run the fixed pulse version since low RPMs is of no real consequence, but someone who say does AutoX might be more worried about throttle response and not so much upper RPM and might want the variable pulse width and/or auto switching. Not to mention street guys like me would be more interested in the auto switching version.

~T.J.

gsl-se addict

iTrader: (3)

T.J.,

Paul (H4Inf) has it pretty much covered. The variable PW didn't work out. It does limit the PW like it was supposed to, but the coil fires on the trailing edge of the signal. When you cut the signal short, you also screw up timing. In order to make that work, you would have to predict when the coil needs to fire and then start changing some time before that. I've messed with some diagrams to do it, but it is very complex without using a microcontroller. If you did go through the trouble of using the microcontroller, you might as well go all the way and have it run the full 2nd gen ignition.

So, as of now, the fixed pulse version is what works (and it works well). You do need to change the ballast resistor in the 2nd gen pack or add an additional resistor on the outside of the pack. This limits power to the coil and helps with the overheating. There is a "how-to" in the archive. Search "simplified TT" and you should find it. I can also give you a hand with parts or any info you might need. Just shoot me a pm.

I've been wanting to take this further, but the house, my car, work, etc. has taken all my free time. Maybe this winter I'll have some time to improve it some more..

RotorMotorDriver

iTrader: (1)

That simplified thread is the one I got that quote out of, so Ive read it. I guess I was confused about the variable pulse versus the auto switching? I was under the impression you had a version that regulates the pulse width independent of the 109 (fixed pulse), another version where the 109 controls the pulse width (variable), and a version where you had built a circuit to transition from one to the other pulse width as needed (auto switching)? Is that what you are saying never worked out?

I can see how that could be what youre talking not working though, because if the circuit cuts the signal short as youre revving up (auto switching), that changes the timing halfway through like you said. I guess the terms just have me confused?

If it worked out using the fixed pulse width I can go that route easy enough, which is what the tutorial is for. My other thought was the "basic" DLIDFIS, but I was thinking about doing it with GM ignitors and MSD SS coils instead of the typical cylindrical Blaster coils this time, instead of the 109s and MSD Blaster coils I used last time.

~T.J.

gsl-se addict

iTrader: (3)

I actually mis-spoke earlier. It is the variable that the simplified does. That is what works the best for now. The fixed and autoswitching have the same problem. The auto-switching unit is basically both the fixed and variable in one unit. When the J-109 is asking for a long charge (this overheats the coil pack), the idea of the autoswitching was to limit this. So if the J-109 was saying to charge for 10ms, the TT box would say "that's too long" and would clip it down to 2.2ms. The problem is it cuts the tail of the pulse off..you would need to cut the front part of the pulse to make it work.

So fixed and autoswitching both mess with timing to some degree. The fixed isn't as bad in that regard. Kentetsu loves the fixed. For me, it seemed a little weak. But I think that it depends on how you set timing. Mine was stock (set at idle). My guess is that Kentetsu was setting his at full mechanical advance (4000 RPM or so).

RotorMotorDriver

iTrader: (1)

Ok, got it. Thanks. Its interesting how such a simple concept becomes such a problem, haha.

~T.J.

H4Inf

I have updated the webpage with some of these details to hopefully help people who come across this thread:

Transistor Trick Overview
http://rx7.pw.cx/transistor_trick/

It is late here so if you notice any mistakes or can think of anything I could add, please let me know

Paul.

RotorMotorDriver

iTrader: (1)

Makes much more sense now, thanks for the updates and info everyone!

~T.J.

GSLSEforme

Just read the complete 49 pg. thread,very interesting,educational. Not trying to bring thread back from the dead,where was this left,what was the final outcome? Thanks

t_g_farrell

iTrader: (1)

I think the last post by gsl-se addict sums it up pretty well. The original simple implementation
works fairly well for most folks and the fancier auto-switched or fixed flavors do not.

No one makes this for you, you have to wire and solder your own solution up and its all documented in the earlier posts.

GSLSEforme

I've seen a pic or two of your car with an FC leading coil,TFI trailing coil using HEI igniters,how much stronger is the spark from the FC coil compared to oem 1st gen with DFI? Do you use FC style plugs in leading position on engine? How long have you had the FC coil setup installed/ Did you ever run any of the TT versions in your car. I know20 questions.... Going with some form of DFI soon in my SE,trying to gather as much 1st hand info as possible. Thanks.

t_g_farrell

iTrader: (1)

1. I never did the DFI using the OEM coli so I couldn't give a good comparison. Just
using a direct fire solution with any coil is a win/win. I used the FC because I wanted
to a) reuse most of my wiring for the 2 HEIs I already had, b) I wanted to
mount the FC coil in my modified OEM coil mount in the stock location and c) I just
rewired one HEI to be trailing and gave the single HEI left firing the leading FC
coils.

2. I use rx8 plugs, they seem to work really well. I've had the FC setup in place
now for about a year.

3. I built a version of the TT but never actually tried it out.

You can use OEM coils but will want 3, as 2 of them will be dedicated to direct
fire th leading. I know Jeff20B uses OEM coils in his DFIS setups and seems happy
with the performance.

I like solid state coils because they have no oil to leak and they handle heat a
little better plus they usually are a smaller package.

GSLSEforme

Thanks for info.