H4Inf

have you got a timing light? It would be best to see what's actually happening directly. When I tested i found timing retarded with increased rpm, with R2 in place.
looking forward to hearing if the diode makes a difference. i havent had a chance yet to do any more testing.

devitek

Just got back from the second test run. I got all the diodes in correctly now, including the inline one. I need to pick up a timing light for sure. With the advance all the way forward, the idle was still too low. I had to adjust the idle screw on the carb to keep it where it should be. This setup feels different, not quite as "punchy" as the simple circuit. All of my connections are solid this time, double and triple checked. I'm going to bring the advance down a bit when I go back out, as moving it made no noticeable difference in the idle speed. Wish I had access to more testing equipment like you guys, but I guess feel is everything.

gsl-se addict

iTrader: (3)

It would be interesting to see if the diode made a difference or not. Probably hard to tell without a timing light, though. Another thing we are going to try to look at is the power supply portion. Perhaps we are getting RF interference on the power lines that's causing the circuit to retrigger when it is not supposed to.

When Rob and Paul tested on the bench, they found no delay between the input and output pulses, so I am not sure how timing is getting thrown off. Hopefully we will get it figured out soon. If anyone else want to help in the testing of the new circuit, let me know.

Things we can try:
- check the 5v on the board to make sure it is clean/constant
- possibly add a capacitor (and maybe a diode) on the 12v supply to smooth out any jitters in the supply voltage
- see if the signal diode on the output helps
- test the circuit with a 1kOhm resistor connected on the output instead to the coil to see if the problem is from the additional load on the circuit or interference from the coil itself

I am not sure where else to check. The circuit works well without the coil connected. The only differences between the coil connected and not I can think of:
- extra load on the circuit (5mA or so)
- possible interference from the spark plugs/coil

If anyone has other ideas, I would be glad to hear them. Thanks for being patient, guys. I really wish this had gone smoother. I think we will be able to correct the problem. Thanks to all that are helping test the circuit.

Kent

H4Inf

Hi Kent & everyone,

I have been running on stock ignition the past 2 weeks now, today I switched back to the TT (without T2) - blimey! there is really a huge difference! it just revs easier, more responsive, its fantastic! brings new life to the car

anyway we did some testing with a diode and have some results. for the record we used an IN4009 diode - your average sort.

TT ->|- Coil Trigger Wire
Made no difference at all. Timing was very retarded as we saw with the TT v2.1 (with R2 in place), and it retarded more with increased rpm.

J109 ->|- TT
No signal got through and car would not start

TT ->|- J109
Now inserting the diode here gave us some really really funky results!
The coil fired intermittently, car ran like crap, but it gave around 3-5 pulses at a time, then was dead for a second, a few more pulses - dead - more - dead... on and off, intermittently!
When we aimed the timing light at the timing needle, we found that when it was pulsing, it was pretty close to on target. The leading and trailing were either side of the needle, so it was very close to right...

I'm exhausted now, would be interested to hear how you are going with your TT, devitek!

Cheers

Paul.

dj55b

iTrader: (2)

H4inf,

Would you by anychance be able to post up the output signal of both nand gates with the coil in place and without? The two that i'm refering to are the output from both min and max circuits outputs, before they both enter the nand gates. I'm suspecting that all the noise is coming from where we tapped into the 5V source

H4Inf

i wish i could give you that data - i had to return the oscilloscope to work however. Kent is chasing up some other people to test right now - I may get a chance this week at work - will see how I go.

dj55b

iTrader: (2)

i really wish this college strike that we have going on is over ... then I could do some testing of my own ... well if anyone with access to an occisloscope can do that, that would be great. maybe its time to buy one off ebay.

devitek

With a lack of testing equipment right now, I had to disconnect the TT. My last test run of it was quite bad, all kinds of stumbling at all speeds, stalling out at starts and the like. I have the diode inline with the output, and I didn't notice any difference at all. If anyone has any suggestions for things I could try, I'd be more than happy to pull the TT out and do some more work on it.

dj55b, if you believe the problem to be the 5v tap, do you have any suggestions for what we can do as an alternative? The simple circuit works like a champ, so this problem has to be something we resolve with the circuit itself. I just wish I knew a bit more about this stuff so that I could improvise a bit.

gsl-se addict

iTrader: (3)

Yep. It is hard to check without a scope. I wish I had the equipment here to test. It would be much easier that way. Even if I did, though, I don't have a car or coil pack to connect it too.

One thing you can try: Paul and Rob found that putting the diode on the input line seemed to help the timing issue. The spark was intermittant, though. I believe this is because the diode isn't allowing the pull-up resistor in the TT circuit to return the C terminal on the J-109 to 12v. You can try the diode on the input and put a pull-up resistor between the B and C terminals on the J-109. A 1kOhm, 1/4W would do. The value isn't very critical, though. You could see if it runs better this way. If you can borrow/buy a timing light, that may make it easier to see what is going on.

devitek

Alright. Tomorrow I will have some garage time (mostly to work on the sending unit wiring), but I will also try to run a diode, presumably on the B side and just twist a 1k between B and C on the ignitor. As a note to myself, I'll probably want to remove the diode currently inline with the output. Once I do, I'll post up results.

H4Inf

I did some further testing today with a different diode, as I thought perhaps yesterday the diode we used could not keep up with the frequency.

Device: IN4148/914
Type and Function: 75V 200mA Silicon Small Signal Switch

With this diode in place, the circuit acted exactly as the normal TT with R2 in place. Retarded timing, and increased retarding with rpm.

Yesterdays results of the funky intermittent pulsing was probably a result of the diode not being able to keep up with the frequencies necessary.

devitek

Alright, in a little bit I'm going to hit the garage to do this stuff. Is there any risk to the J109 in putting in a pullup resistor like we mentioned?

gsl-se addict

iTrader: (3)

No. With the stock setup, the coil acts as a pull-up resistor (about 2 Ohm). A 1k will be much less current. The TT already has this pull-up in the circuit (simple version too). The diode would go on the input (stripe on the circuit end). You can just put the resistor between the two terminals of the J-109 (just shove it in the plug). not sure if this will help or not. Seems that Paul tried (without the pull-up) and the same problems occur. More than likely it is something in the power supply causing problems. We will have to poke around the circuit with an o-scope to know for sure. I am trying to add noise on the power supply in the simulation to try to recreate the problem. If I can recreate it, then I can find a way to fix it.

Kent

devitek

Ok, this is a little off-topic, but I just hit the garage to look at the wiring to my sending unit. One of the posts was green and crumbled at the touch, so it looks like no more sending unit. While I was down there, I noticed so many things...this car doesn't have long to live.

I didn't bother with the ignition stuff today, as I'm completely depressed over all the other problems I'm having with this car. I will try some stuff once we get some new ideas, but for now it's going on hold.

Kentetsu

Hey guys. I haven't been watching this thread for some time now, so I figured I check in to see what all the activity was about. Hopefully you guys will get the bugs worked out of the new design...

In the mean time, I'm still running what I believe is the 2nd generation TT unit (autoswitching) with no issues and no failures. Gotta love the power! Keep up the good work. I've got faith in this group...

gsl-se addict

iTrader: (3)

I have had a thought about this. I took a look at some of the scope traces with the noise. It appears that the stray signal looks like the voltage on the coil - during discharge (ringing). You can see what this ringing looks like on page 2 of this thead (posted by renns). Now, the stock system has the condensor that is connected to the coils and ignitors to help with RF and the ringing, I imagine. It turns out the Aussie guys (atleast Paul) doesn't have the condensor connected. Could be the source of the problem.

devitek: Where are you getting power from? I would suggest connecting the circuit power and the 2nd gen coil power to one of the stock coil + wires. This way the condensor will be part of the circuit.

I am going to ship a circuit to teejs (he helped with the simple TT write-up) on Monday. He will help out on the testing. If anyone else want to help test, let me know. We can use all the help we can in the testing department. We could really use people that have access to a scope, a 2nd gen coil pack, and of course an SA/FB.

Kent

devitek

I'm actually getting power from the + coil wire, the coil that isn't used once the circuit is in place. Perhaps I should switch it over to the coil that is active?

gsl-se addict

iTrader: (3)

Shouldn't matter. Either way should work the same. Do you have a timing light yet? Would be interesting to see if you are having the same problems as the Aussies.

devitek

I have a friend that has a timing light, and we are getting the same results. It's so bad on the low-end, it's funny. My mom was driving behind me the other day, and she called me later wondering why I had flames coming out of my exhaust!

So, what about putting a capacitor inline from the power source to the coil pack? Would that make any difference?

gsl-se addict

iTrader: (3)

Is the timing retarding with rpm? If you want, you can add a cap between 12v and ground at the 7805. The datasheet lists a 0.33uF cap for that location. I have seen circuit where people use larger than 0.33uF. On the simulation, I have tried to add noise to the power line to see if I could reproduce the results. Even with very large amplitude noise, I can't get the simulation to reproduce the results. I guess we won't know for sure until we can get someone to explore the circuit with an o-scope.

Basically on paper, the simulation, and on the bench (without coil), the circuit works just like it is suppose to. The problem comes with the coil connected. Basically we have observed:

- false retriggering under certain conditions
- noise on the high state waveform
- retarded timing. Increased retarding with RPM

These may all go back to a single cause. I am going to take a better look at the power supply for the MegaSquirt ECU. renns on here triggered a 2nd gen coil with the MS without issues (see page 2 of this thread). I will see what other ideas I can come up with.

devitek

Hey, just for the record, that whole resistor between B and C on the ignitor just blew my ignitor, I've got no car now. I'm out of this project.

kbeefy

So, is there a functioning TT I can get, or do they all have issues? I need some ignition for my car!

gsl-se addict

iTrader: (3)

That sucks. It couldn't be the resistor that caused it, though. Only if you switched the input wire with the battery on the ignitor would just kill it like that. That resistor is there in the simple TT circuit as well (the 1k right at the input, connects to 12v/B). Thanks for helping out. I really hope you can find a spare ignitor fast so that you can get going again. You could move the trailing ignitor over to leading for now. If you have an 84/85, you would have to bypass the fuel pump relay, though.

kbeefy: The simple TT works well as well as the last batch of TT circuits. The problem with these is that the 2nd gen ignitor gets too hot. You may be able to run an additional ballast resistor to help with the heating. I know we will get this worked out. There isn't many things it coild be. It is just a matter of getting access to an o-scope and spending some time with the circuit.

kbeefy

thanks gsl-se, and thanks for the terrific rundown on the other thread comparing the choices, I'm sure that will help a few people sort it out. I think for now I'll run 2GCDFIS and upgrade to TT when you guys get it sorted out. Good Luck!

dj55b

iTrader: (2)

alright i know i haven't posted in a while .. but I might have some promising news. Our college strike is finally over, we'll be getting back to school on tuesday so hopefully wednesday I can take the car to school install it there and do some testing with an Osciloscope