JP84Z430HP

89, I'll let you know WHEN I get it working, that's right, I'm very confident that I can get it to work. When I get all the parts I need (Module, coils, connectors, and crank sensor) and I get a trigger wheel worked out, I'll be ready to start on it. Unfortunately, right now, I don't have anything to run it on! My ccar still has the old LG4 harness and carb on it! I plan to remedy that ASAP. I guess I just need to pur in more hours to buy make extra money to buy more parts!

Grumpy

And things can change over night.
Things went for YEARS, and YEARS before Programming 101, and little was public domained about GM ECMS. Once the silence was broken then things were able to flourish.

In theory, yes, you could use the DIS code and run the Nstar stuff now. But, again the link was only rementioned a few days about it having been done.

If you really have an idea you want to pursue, then by golly just do it. There is always new stuff going on, and it takes pushing the envelope to discover things.

JP84Z430HP

Well put Grumpy. That's my entire point of this idea. I really like to do things that haven't been done, or things that not many people know of. I find I learn a bunch more when taking on a project like this! And I love to do things that turn out to be a learning experience!!!!

JP84Z430HP

Also, if there is much interest after I get this stuff worked out (I'm looking at several months, as time and $$$ allow!) would some people be interested in getting together on a quantity of the fabricated parts? I'll do the work of getting it right! I have a couple machine shops that would take on the project. I'm sure I could get a deal on a larger number of them! In true DIY fashion, I would be doing it for very little $$$ above cost. Just enough to cover my time.....

I will make the drawings available once I get it working, also, that way people can have them mave on their own.

Right now, the parts I'm thinking I'll need to have made are:

Trigger wheel
Dual Crank sensor mount
Mounting plate for module

Let me know!

Grumpy

Some day I'll write up a list of stuff the looked like it ran on the bench but puked when actually run in a car. Or stuff that ran for hours on end in a car, and then for no reason choked.

JP84Z430HP

No offense, but I bet that's quite a list! I wish I could remember all the stuff in general I've tried that just wouldn't quite work. I'm sure my list will get much longer as I get into this stuff!

funstick

NAH im sure it would work fine on the car. it wasnt until i made the changes i was just talking about that it actually started working. mind yu i didnt invent or figuree out those changes but i do undersatdn why they work. i did figure out the issue with the 6x signal VS the tach output however and substituing the tach output helped the ecm TPU code work a bit better. not nearly a scetchy.

as for my honda its running $8f with DIS. much easier to start from a working DIS tpu code that actually accepts the incomming signal properly. plus it has alot of added functionality the the $58 doent have like highway mode AFR etc etc etc the list goes on. it also has better AE and PE modes and i feel a more refined idle stragey. but im just getting this up and running. im switching from the dizzy based $58 to the DIS based $8f. still some teething problems. ill post all my info soon enough along with pics and hurdles i had to overcome for any of those interested in using $8f on a 4cylinder honda engine or similar.

JP84Z430HP

I'm gonna be interested to see how the GM DIS and ECM will do at the RPM's that Honda's regularly turn! That is also a very interesting project, as is any GMECM on another make engine!

The_Raven

I'm just curious as to why there needs to be 2 crank triggers?

JP8, can you post some of the info you have found or maybe write some of it out?

JP84Z430HP

The two crank triggers aren't required, but the way the system works, it can determine when to fire within the first 180 degrees if both are used. Otherwise, it would take one full crankshaft revolution to start firing, and sending a signal to the ECM. I'm thinking that with the dsign I have in mind, it won't be more than just another hole, and 1 more tapped hole for it, along with the cost of a second sensor.

I'm posting a pic of the CAD drawing I'm working on for the trigger wheel now. If I get my scanner uncovered, I can scan in the pages I have printed out!

Does anyone know where #1 TDC should be on this wheel? I labeled it as in the info provided to me from the manual I have.

funstick

its an a20a3 from a 88 accord. its a loafer a 6500rpm.

JP84Z430HP

That means no V-TEC. It would be interesting to do something with the V-tec based on more than just RPM's! Maybe I can talk my friend into doing a GMECM setup for his project car! He wants to build a fast Civic. Got my wheels turning even more now! Anything to get rid of the failure prone Honda distributor!

funstick

exactly. the 4cyl module should run upto 7500rpm no problems. might need deeper more defined notchs.but it should

JP84Z430HP

Disclaimer::::::

Don't try to use that .jpg I posted to try to make your trigger wheel! That was the design I was thinking of for under the cover with the vortec crank sensor, but I don't have it at the correct angle!!!!!

The_Raven

Hmm, that's odd, there are a lot of trigger notches there.

Is this a copy of what I would thing be a crank trigger wheel (internal to engine), or an external I.E. 24x wheel?

funstick

those sqaure wave signals will in fact run a standard ignition module. the problem is on a production engine actually keeping the magnets in the trigger wheel crank or pulleywas a chore just 10-15 yrs ago. thats why there is such a wide spread use of VR sensors( this is comming from the mouth of sevral OEM engineers i know on a personal level)

do me a favor dont compare a standard ignition module to a DIS module. there like 10 worlds aparts in design and contruction. there is a chip inside the module that decodes the incoming VRS signal into a square wave. as stated hall effects wernt that stable when these units were designed. but the DIS module does do alot of signal conditioning for the ECM thats no secret to anybody in the auto repair industry


i already understand the code changes but expliaing it here would be like pulling teeth with a dull rusty spoon. no body get half of what I say becuase ive gotten pretty far down the rabbit hole with OEM electronics. Nothing to terriably ECM specific but more genral GM design and operating thoery. of course i have engineers i can tap when stuff deosnt quiet click so then i ask them why this way ?? they tell me then i figure out a way to make it work.

Often the problem is not in finding the solution. Its asking the right questions

JP84Z430HP

Or more like having the right contacts to ask the question to!

As for the notches, there are a bunch of them. 24 of them, plus 8 additional notches used for synchronization. The system will work with just one sensor, or with both will know where it is in less than 1/2 crank rotation.

In the OE setup, the crank sensors are 27 degrees apart. Now that one, I have no clue on the reasoning why!

funstick

thats pretty simple really. one wheel track is low res it provides the synce pulse count the actually determine crank position.


the other wheel is the speed wheel. it give it a high resolution count. the sensors are 27* apart most likely due to a mathmatical thing. remeber its not one wheel its two seprate wheels. this i know for sure.ill try to find the lower end guts out of a n* engine and get some photos of the sensor arangment and the trigger wheel itself. i think i may even have the n* pulse charts to look at for comparison. if i find them ill ask you to toss me an email.

funstick

i highlighted the low res track at least what i think it is ill chgeck my pulse charts

funstick

JP84Z430HP

I'm not so sure on that one. According to the documnetation I have:
"The crankshaft reluctor ring has 24 evenly spaced notches plus 8 additional notches used for synchronation"

I'll scan and email it later. (I have to get the scanner working first! LOL)

JP84Z430HP

i found a pic of a crank that someone is selling on Ebay, it looks to me like like just 1 wheel on the crank.

http://cgi.ebay.com/ebaymotors/ws/eB...category=33616

I am keeping in mind that is a crank out of a newer northstar, one that likely has CNP, BUT, I doubt they changed it. To my knowledge, they still use the same trigger wheel in the 2.8/3.1/3.4 family of engines!

The_Raven

The trigger wheel on that N* crank is very different from the 60*V6 crank trigger. In the 60*V6 there are 7 notches, 6 even spaced and one 10*before #1 TDC as a home signal.


Interesting.

JP84Z430HP

Yeah, the inputs to the module are WAY different, but I'm pretty sure the signals between the module and ECM are the same. Some many things have been said in this post, and other outside references, I need to go back and read through and take notes to get it all straight!

Although, I still think my original idea is what will work, even if it involves some code hacking.....

JP84Z430HP

Ok, I got the pages scanned in. You can grab the .zip file of them here:

Northstar.zip

edit: pointed URL to new site.

Doctor J

JP84Z430HP:

Thanks for digging up the N* info, it's very interesting. I had not realized the wheel sequencing was done in the ICM (instead of the PCM). That should make a V8 project do-able, with the right parts.

You mentioned using the Vortec truck pieces for a trigger wheel/sensor. FYI, here are a couple of details of the Vortec parts used on an older small-block retrofit project (non-DIS):

1. Parts Used

Doctor J

1. Parts Used

Doctor J

2. Installed Wheel

Doctor J

3. Cover

JP84Z430HP

The Vortec wheel WILL NOT work. I was talking about doing something based on that wheel. Fortunately, I HAVE a 98 Vortec truck block in my car now! I already have the cover, along with a couple spares with the sensor in them, also.

I'm working on a couple options for it. I'm doing one that will be in the Vortec cover (would need a newer block that would take the cover, or seal it up with RTV as DoctorJ has). The other would be mounted between the balancer and the bottom pulley, and would require a bracket for mounting the sensors. The first option would only be able to use one sensor, therefore it would take near a full revolution to get spark, where the second option would use 2 sensors, and it woulspark within the first 180 degrees of crankshaft revolution.

This has been very interesting so far, and I intend to keep updates flowing as they occur!

JP84Z430HP

I guess the vortec timing cover and a modified reluctor wheel is now looking less like it will work. I really would want both sensors anyway. I found out a little while ago that the vortec CKP sensor is a hall effect type sensor. At least it only sends an On/Off signal, instead of an AC wave. I was looking at some stuff on my other project (Vortec 350 in my 88 S-10) when I noticed this. Maybe I could still find a sensor that would fit the cover that would work??? I may go to the parts store and start measuring sensors. I do like the idea of having the sensor in the cover, out of sight, out of mind!

JP84Z430HP

Funstick,
Getting anywhere on that Honda? I have a buddy working on a Honda project, and I'm trying to talk him into using GM electronics, that way, we'd be on the same page, not to mention how inexpensive it is along with the availability of source code info!

To keep this on-topic.....I haven't given up on the Northstart DIS project, I just don't have the parts to work with yet!

funstick

pics




more pics



pics of it in gainsville florida gas station at a 3geez meet


i live near detriot so figure ive run it 1300 miles each way with no efi fialures. cars an animal

LnealZ28

Any more progress?? I'm working on a couple of ideas for a waste-spark DIS as well and have been considering a retrofit of the N* system. I too have looked at all the diagrams for the N* PCM/ICM communication and I think it could be done. Also I have been looking at doing my own ICM using a crank wheel, a reluctor type sensor and THIS computer (click on the NMIN-0021A-BUFFALO link). The computer has limited memory resources, but for the relatively simple approach I'm looking at I think it will be plenty. If not they have others with more memory. Writing the code for this shouldn't be too hard, I've written up a flow chart for it and it doesn't seem to be that complex, but I am taking a simple approach. The computer will get only sync info from the wheel; and the ECM, ('730 running $8D) since it doesn't care what cylinder is firing, will get it's normal "DRP" info from either a cam sensor or a second crank wheel with 4 teeth spaced 90 crank degrees apart. This way the ECM will continue in its duties of calculating timing and instead of sending that pulse stream to a module, it will send it into the ICM which will determine which coil to fire and do it at the timing angle determined by the ECM. FYI I plan to use 4 GM 4 terminal modules, 1 for each coil for the ICM to trigger.

So, go ahead and take potshots at me and poke holes in it. After all it's just an idea................

Grumpy

While not much of an error if you don't but, if you can you'll want to find the latency of both systems, ie the 730 and donor for the DIS module, and change the values.

As it turns out, with the 1227148, in the GN they use a Hall Effect sensor for cam synch, and then with the same code, and module use a reluctor for the cam synch in the FWD'ers.

The North * modules the last I looked were really expensive.

One other thought is using coil near plug setups but firing them in DIS mode, ie two atta time. An MSD dual sych dissy gives you the signals needed, and a home made 1-8 counter, and reset would take care of the rest.

And whatever you do, remember the oem stuff doesn't have an EST at crank, so you need a *module* mode, or research to see if you can set the EST min RPM to some low figure to get an EST signal when cranking.

Jus brain strormin....

LnealZ28

Yea, the N* stuff is pricey, but I have seen a couple of coil packs go for next to nothing on ebay. If you had to get it new though it would be painful.

I'm figuring that by the time you reverse engineer what the factory did with their DIS, one could have a dandy, working system of their own design that will reliably do just what you need it to. Which is why I'm looking at doing it the way I mentioned.

And the EST thing is a good point. I'll be getting an ECM bench shortly and hopefully will be able to get a prototype running sometime before the end of my life expectancy..................

JP84Z430HP

While I haven't given up on this project, it has become a bit stagnent. I don't have some of the dimensions that I need for the trigger wheel (if I get to the point that I really need them, I will get them, but I've been "waiting for an opportunity" to measure what I need), and I still don't have anything to run it on, the Camaro still has a carb, although, if I were to get going on this, I could use the '730 I have sitting here for just timing......

I also have other projects in the works (LT1 in the S-10 will be coming soon, likely to run a 730 on it, and that may be where I get this project into operation, so that I don't have to run the opti-scrap ignition)

If I get any further on it, I'll put updates on my website, and link it here, prolly in this thread....

LnealZ28

Ya know JP, I don't think it would be too hard to set up my little idea for timing calcs with a carb. Just a MAP sensor for the vacuum advance and a couple of one column tables, one for vac advance and one for centrifugal advance. Of course the tables could have as much resolution as you wanted. You'll already have the RPM data going in there, just write a little code to interpret it. Might work, I don't know. I do know that other DIY'ers have done this for carbed apps.

JP84Z430HP

THe plan is to get away from the carb ASAP, but for now, I'm dealing with it, and learning as much as possible about the ECM stuff, and I'm even trying to figure out some of the source code....

I was doing some checking into pricing on the N* module, and I found one on gmpartsdirect.com for $256. Not all that bad (our shop cost from Carquest is Over $100 more than that!) The coils are the same as the other GM DIS systems (Other than the Quad 4 and the older Buick stuff) at least they seem to be..... I wouldn't think it would be much cheaper to build a system in reality, taking time into consideration....

Just some more thinking aloud here..... Feel free to yell at me all ya want!!

JP84Z430HP

No new news here, other than this project has NOT died! I am working on some new ideas for the crank trigger, and another member is getting a N* crank, and sending me the info needed to complete the trigger wheel. I may have a new, better idea for a trigger wheel, other than one that would sandwich between the pulley and balancer (thanks to 89 Iroc Z)

I'd prefer to use the LT5 ignition system (boat pieces that were mentioned earlier in this thread) but the module is more expensive, and far less plentiful.


Can someone find me a few more hours in each day?

TheGreatJ

Personally, I don't see the purpose in going to the extra trouble of running 2 sensors to get spark in the first 180* of cranking. It takes 360* of crank rotation to get a fresh charge of air and fuel in and compressed anyway.

180* of crank rotation only gets you one piston stroke. In order for a cylinder to fire, that has to be the compression stroke, which means the cylinder which happens to be the one you're compressing must already have a fuel/air charge in it, and it must be relatively fresh (the fuel can't have settled out of the air yet because the heads most of us run don't have enough quench to adequately mix the fuel and air, especially in a cold cylinder.)

By my thinking, the only way spark in the first 180* is going to help is with a hot-start, and to me at least, it's just not worth it.

JP84Z430HP

The main reason I'm going to keep both is that's the way the system is designed, and just because I can.

As for the starting quicker, ask Grumpy about converting his Buick to the "Fast Start" type system. He seemed pretty impressed with how it started. I'm also pretty happy with the way my 350 with a carb starts (most of the time). I bet it's actually firing within the first couple rev or 2.

I also like the redundancy of 2 sensors, in case one fails (More like WHEN).

TheGreatJ

Now redundancy I can understand, and I agree that it's easier to leave it as-is than to eliminate one of the sensors. I just don't understand why GM did it in the first place. It's not like the General to stick in an extra sensor that (it seems to me) didn't really need to be there. Makes me wonder if it doesn't have another purpose.

But I've only worked on a couple of NorthStars, and never really gotten deep into their ignition setup, so I dunno. I'll read up on it a bit at work......not much else to do (business is dead slow lately.)

JP84Z430HP

From everything I've read, the second CKP sensor is there to help figure out which cylinder to fire, faster.... Notice that they only put a faster starting system on their higher lines (Cadillac and Buick)?

The system will run on just one sensor though, from my understanding.

JP84Z430HP

I feel ya there! Luckily, some old projects are coming together to keep me busy right now!

rattle

the system will definately run on just one sensor, I work in chevy dealership, when one sensor quits the owner complains car takes to long to start, 1.7 seconds instead of .63 seconds, (i timed it on more than one)

JP84Z430HP

Thanks for the verification!

I still think I will design around using both, and use them on mine, but document that the secon sensor is totally optional when I post the info.

As I stated before, I'll make all my info and drawings avail for $0 after I get it all working, and maybe provide parts and possibly package deals for those that don't want to machine or source the machining and parts themselves....

Z69

One sensor...

My diesel has one cam sensor only. Guess what happens when it fails......Least they last for 90-100k, got a new one in the console already.

eric305TPI

Let me guess....A Powerstroke

JP84Z430HP

I've NEVER changed one of those........

Hehe......