Coil Packs and Dwell Time

[TrevorS]

For whatever reason, I'm having trouble coming up with a second good J722T power transistor pack. In the interim, I've taken the opportunity to build a new heat sink assembly that provides 17% more mounting surface and 71% more fin area. I figure it can only help. Also, I'm trying to learn what is the usual operating temperature of a J722T with a fully warmed engine, but have yet to get an answer (my car had a J122 and a different mount). This requires a measurement with either a thermocouple meter or an IR gun on the black case (not the connector or surrounding metal). If anyone can help with either of these, it'd be appreciated -- thanks /ubbthreads/images/graemlins/smile.gif!

 

[JNR]

Have you tried Rock Auto? They were having a closeout on the power transistors a little while back, although not sure if they still have them. Don't forget the discount code, too.

 

[TrevorS]

Unfortunately, retail is consistently much more expensive than used /ubbthreads/images/graemlins/frown.gif! Surely there are enough discarded '91 through '99 transistor packs out there that an experiment of this sort shouldn't require an arm and a leg /ubbthreads/images/graemlins/smile.gif! I've seen posts on this very thread where posters were willing to provide parts, however, I'm wondering if that was then and this is now? To what degree has this thread really died?

 

[JNR]

Yeah, I hear ya on the $$, but they were more than half off, iirc and least you now it's good and will have a lot more life left; not so sure on used, although used can probably go for quite some time. For experimenting, I can see the appeal of used thoughm until you get it all ironed out and tested, then maybe use the new one(s).

 

[TrevorS]

Just checked with Rock Auto and I'm seeing over $100 apiece -- far too much for me. Will just have to see how things go.

 

[TrevorS]

Picked up some 14AWG yellow and red wire today with view to possibly upgrading the primary drive from the transistor packs to the coils including the #2 to #3 wire (downside being it blows the OE harness color coding /ubbthreads/images/graemlins/frown.gif). Trouble is, I really don't know where I'm at, only that with a single transistor pack I don't pop my 15A fuse -- plus, I've no measured current info for OE! Still, plan is to insert a .02 ohm resistance inline with the relay 12V. I'm thinking a dual trace scope should allow me to see how the 12V primary current is behaving (would also take a look at the 12V itself). 'Course, it's kind of a slow scan at idle, but hopefully good enough to learn something useful /ubbthreads/images/graemlins/smile.gif!

 

[TrevorS]

Received another J722T, sanded it, and tested it in the car -- it works and I've installed both into a revised heat sink assembly with heat sink compound.

Installed the transistor pack assembly, reconnected the COP and battery, and fired it up -- caught instantly (as in instantly /ubbthreads/images/graemlins/smile.gif)! Took it for a run and it responded as normal -- engine was super smooth both at idle and moving. Got back, and while idling, popped the hood and measured the transistor pack temps. It was evident that stopping had allowed them to start climbing some, but they stabilized with the front at aprox 106F and the rear at aprox 110F (43C). That doesn't worry me at all, far lower temps than a typical PC processor. (Outside temp was 81F.)

Judging by the way the engine started and ran, I consider the mods to provide greater spark energy than OE and an even greater improvement over single transistor pack COP.

 

[JNR]

Thanks for the update and keep us posted on long term reliability and all that...

Did you (or could you) post a wiring diagram earlier in this thread, or maybe there was one up there somewhere for that?

This thread has gotten so long and so much info, I just want to be sure it's the correct way, should I decide to go back COP. Otherwise, I like the dual GM coil idea too, so decisions, decisions /ubbthreads/images/graemlins/grin.gif

 

[TrevorS]

Thanks for posting JNR, I've found the level of participation in this thread discouraging. In fact, my continuance of it is entirely out of respect for "CheekyChimp", "Curtis", and yourself. I surely hope you find it helpful.

The only significant wiring change is that associated with COP Vs OE, different variations for different model years. It seems nobody else bothers, but I strongly suggest installing the standard Intrepid ignition capacitor into the coil +12V, this replaces the one provided by the OE coil pack -- it's there for a reason. The only other differences are that the series coil primaries are fed from a dedicated battery line (not the harness) and the Power Transistor Module grounds have a bolted local ground (again, not harness). Make sure the ignition switch "ON" voltage is delivered to the power transistor pack power supply inputs (as in OE) as well as the relay activation coil that controls the dedicated coil battery line (which is new). If you have a '91+, only use one of the two Power Transistor Module tach outputs, stub the other off (not ground). There's really not much to it, but it does take a little time to put together.

My remaining interests are my COP module temps, and measured coil primary currents and durations. I strongly believe it works, but my ability to take measurements is just a tad limited /ubbthreads/images/graemlins/frown.gif!

Lastly, I recommend against inverting the polarity of any of the COP modules (and hence the spark plugs). Yes, to feed -/+/+/- does to some degree emulate the OE coil pack, but as usual, the devil is in the details. To start with, the coil secondary currents are opposite, but not matched since the primary provides a shunt for the resulting differences. Second, when the lower ionization voltage results in spark in the OE system, the voltage drop is immediately shifted to the other plug helping it fire. This is not the case for inverted COP where the voltage change at spark is electrically isolated from the twinned plug. I strongly recommend maintaining standard plug polarity.

 

[TrevorS]

Quoting JNR:
Thanks for the update and keep us posted on long term reliability and all that...


The main reliability question for me was whether there might be issues associated with doubling up the transistor packs -- either at the packs or COP modules. However, the successful long term usage earlier suggests not. I've moved coil power completely out of the OE harness and with the heavier shorter wiring, there ought to be greater spark energy, so my main concern is operating temps. The heat sink assembly is clearly keeping transistor temps well down, but I still have to verify the COP module temps are also fine. With the aid of the small air gap around the COP plate, I'm not expecting an issue.

That leaves two things that are new -- the COP capacitor and the relay. The capacitor is located under the COP plate adjacent the timing cover. This puts it fairly near the rear cable opening and there's an air gap around the edge of the COP plate. Plus, the mounting tab serves as a heat sink and is bolted to the valve cover. In addition, it's designed for automotive use and as I recall, it's rated for 120C. I don't think it should be a problem.

So, that leaves the relay and it has six things going for it.
*It's rated for 40A and is probably carrying under 15A peak (line fuse is 15A)
*The only load it sees on closing is a .47uF capacitor (safe for the contacts)
*It takes about 14msec to open and coil load should have ceased (safe for the contacts)
*It's located in a fairly cool area of the engine bay -- low behind the plenum
*Its temperature rating is 40-125C, and its cycle rating is 100,000 at 40A
*I sealed its case to protect it from vapors or gases in the engine bay

So, the only downside I currently see is there's a new load on the battery with the ignition ON. The relay coil draws around 140mA at 12V -- well, we're not really supposed to leave the ignition ON without the engine running /ubbthreads/images/graemlins/wink.gif! If I can get a scope on the coil battery line I should be able to learn more about load behavior.

 

[broxma]

I am amazed this topic is still going. I never got around to posting any results using the dual ignitor setup so here you go.

I got rid of/parted out my Galant. Too many frame issues for the power levels it was looking to make led me to the conclusion it wouldn't have been safe at the speeds it was capable of getting up to. That being said, we did dyno the car up to 362 WHP using the dual ignitor setup. The car was boost limited as in it had a massive crack in the intake manifold which prevented the car from seeing more than 18 pounds, but the ignition system performed flawlessly.

Onto my current project. I'm building a 240sx now, using the KA motor, mainly because I ain't no punk who thinks "SR20 yo", when the car came with a perfectly decent closed deck KA motor. The interesting aspect of my build is that it will be running on the exact same Evo 8 AEM EMS that my Galant was running on. Incidentally, the ignition system on the KA, although it normally uses an old style distributor, the was made by non other than Mitsubishi. If you remove the cap, rotor and inner cover, you find an optical cam angle sensor. It happens to be the exact same cam angle sensor as a 1st gen green top CAS. You can simply replace the optical wheel with either a 24 tooth or stock 1G CAS wheel, and any Evo 8 designed ECU will work just like factory, provided you do some wiring. This means the same ignition principles we worked on here, will transfer over directly to the Nissan. That car should be a most excellent test of the ignition as it will be running an S372 turbo at around 35-40 pounds hopefully pushing north of 750 HP at the wheels. The car won't be running for a long time. It is currently in the shop getting the 10 point cage and parachute mount fabricated. However I do not believe the common wisdom that a 1200.00 stand alone ignition is needed for a car at such levels. I believe people simply aren't ingenious enough to figure out how to do the same thing using factory components.

I'm glad to see this thread is still alive. Several of us put a lot of effort into testing and designing this stuff, and I hope in the long run it helps out more than just the few of us who started looking into it. That appears to be the case so I'm glad to have contributed.

Side note, I have whittled down the wiring on the factory ECU. It can run on only 25 wires. I know they are numbered over 100, and you need less than 1/4 of them to get the car running properly.

/brox

 

[TrevorS]

Quoting TrevorS:
If I can get a scope on the coil battery line I should be able to learn more about load behavior.


Installed my curent sampling resistance at the relay today and idled the engine while trying to take measurements (the mosquietoes loved it /ubbthreads/images/graemlins/frown.gif).

I tried two different scopes and ran into anomalies with both, but finally came up with what appear good and reasonably consistent data. From the waveform, it was clearly not reaching coil saturation, though the charge time was aprox 8ms with aprox 96msec between individual coil charge cycles. That translates into 625rpm which roughly agrees with my dash tach. I measured a peak current of about 12.5A using the above .02 ohm sampling resistance, which translates into 3W peak on a wire-wound resistance rated at 25W and operating at a roughly 8% duty cycle -- don't think resistance heating should be a factor. Looking at the battery side of the resistance at same sensitivity (.005V/div) I was seeing only transients, no noticeable level shifts. By the time I completed these measurements, the engine had been idling for quite awhile and I doubt the battery voltage was higher than 12V, 14V would make a difference.

Since the coils didn't reach saturation, I can't be sure what the true peak current would be, but with 14V the current would increase by about 2.5A and judging by the shape of the charge waveform, the true peak is likely on the order of 18A. Perhaps that would achieve saturation, but I doubt 4ms could do it. Thinking of fusing, at 6Krpm individual coil charge cycles occur every 10ms and coil duty cycle could be similar to 60%, but there won't be time for the current to climb (suppose a worst case of perhaps 15A @14V, 60% would be under 10A average) and so I expect my 15A fuse is fine. Regarding COP module heating, looks like they're unlikely to actually saturate and so resistive heating should be minimized, but there's still laminate heating (particularly at higher duty cycles). However, given the Intrepid operates the COP modules independently, they each see battery voltage (with at least .4 ohm less series resistance) whereas ours share it, and so I'm guessing their higher current heating is likely similar to our waste spark double duty-cycle heating -- probably not an issue.

Quoting broxma:
we did dyno the car up to 362 WHP using the dual ignitor setup. The car was boost limited as in it had a massive crack in the intake manifold which prevented the car from seeing more than 18 pounds, but the ignition system performed flawlessly.



I'm glad to see this thread is still alive. Several of us put a lot of effort into testing and designing this stuff, and I hope in the long run it helps out more than just the few of us who started looking into it. That appears to be the case so I'm glad to have contributed.

/brox


Thanks for dropping by again Brox and that's too bad about the Galant, but it's pretty clear you have lofty targets /ubbthreads/images/graemlins/smile.gif! Very interested in how things go with the Nissan and the dual ignitor COP ignition, keep in touch! Hopefully others will jump in here too.

 

[broxma]

To ride along with one of the things you mentioned in the last post, about the way the Intrepid coils each see battery voltage...

One of the changes I am going to do to the setup I run on the Nissan is run the power wire off a relay. Since the positive wire at the harness is always hot, I plan on running it to a 30A relay instead of running the coils off power supplied by the ECU. The ECU is getting all new wires, all new pins, so I'm increasing the wire size on both the positive and negative of the ignition.

I also may not go with intrepid coils.

/brox

 

[TrevorS]

Quoting Broxma:

To ride along with one of the things you mentioned in the last post, about the way the Intrepid coils each see battery voltage...

One of the changes I am going to do to the setup I run on the Nissan is run the power wire off a relay. Since the positive wire at the harness is always hot, I plan on running it to a 30A relay instead of running the coils off power supplied by the ECU. The ECU is getting all new wires, all new pins, so I'm increasing the wire size on both the positive and negative of the ignition.

I also may not go with intrepid coils.

/brox


In case this info is of any use to you:

1) I ran 10 AWG (starting with a fuse holder) from the battery to the relay following the existing firewall harness.
2) I ran 12 AWG from the relay to the coil connector.
3) I ran 14 AWG from the coil connector to the COP module +12V terminals.
4) I also ran 14 AWG from the #1 to #4 COP modules (a relatively long run).
5) Remaining COP wiring between ignitors and coils are OE 16 AWG and similar total length. (Charge waveform is also similar height.)
6) Ignitor ground is 16 AWG each to a stubby 14 AWG to a nearby ground bolt.

Given the peak primary current might reach 18A, I'm considering changing the 16 AWG dual ignitor to COP coil module wiring to 14 AWG, but that may be going a little over the top for my purposes -- having difficulty deciding /ubbthreads/images/graemlins/frown.gif! Would love to know what you come up with, and whatever measurements you make /ubbthreads/images/graemlins/smile.gif! (Including the COP modules!)

 

[TrevorS]

Quoting TrevorS:
I've moved coil power completely out of the OE harness and with the heavier shorter wiring, there ought to be greater spark energy, so my main concern is operating temps. The heat sink assembly is clearly keeping transistor temps well down, but I still have to verify the COP module temps are also fine. With the aid of the small air gap around the COP plate, I'm not expecting an issue.



Took the car for a run and then measured the COP module temps. Coil 120F, COP plate 140F, valve cover 170F, outside ambient 85F. Typical COP temp rating seems to be 120C (248F), so I think the COP margin should be fine. However, I notice the battery is discharging faster than normal at idle. Thinking of backing off the primary current a little by returning the transistor pack ground to the harness.

 

[TrevorS]

Quoting TrevorS:
Took the car for a run and then measured the COP module temps. Coil 120F, COP plate 140F, valve cover 170F, outside ambient 85F. Typical COP temp rating seems to be 120C (248F), so I think the COP margin should be fine. However, I notice the battery is discharging faster than normal at idle. Thinking of backing off the primary current a little by returning the transistor pack ground to the harness.


Replaced the cracked rear transistor pack connector and returned the ignition transistor pack ground to the harness. Still draws more current at idle than OE (guess not a huge surprise), but not as bad as before. Got to scope it to see where I'm at!

 

[curtis]

/ubbthreads/images/graemlins/applause.gif Liking what you've been up to with this /ubbthreads/images/graemlins/jawdrop.gif

 

[boostx]

Well i total missed this discussion. On the cooling problem for the cop you might want to add holes to the plate so that the heat will not get trap between the valve cover and cop. This is the reason why i add the logo and or the words to my cop's.



On the cooling the transistor pack have why not try using old computer processor heat sink. Some are pretty big. I just yank this out of a old pc i have at work.

 

[TrevorS]

Well I totally missed this discussion. On the cooling problem for the cop, you might want to add holes to the plate so that the heat will not get trapped between the valve cover and cop. This is the reason why I add the logo and or the words to my cop's.


Perhaps there are PC heat sinks that would be good (mine are appropriate only for CPUs and graphics cards). Given 110F measured and a since reduction in peak current, I doubt I've a problem, but perhaps there are other possibilities that will help others /ubbthreads/images/graemlins/smile.gif! As far as COP module cooling, those cut outs should surely help, though given a 30F difference between my COP plate and valve cover, I think my grommet solution works pretty well /ubbthreads/images/graemlins/smile.gif!

 

[TrevorS]

Quoting TrevorS:
Replaced the cracked rear transistor pack connector and returned the ignition transistor pack ground to the harness. Still draws more current at idle than OE (guess not a huge surprise), but not as bad as before. Got to scope it to see where I'm at!


My neighbor was kind enough to recharge my A/C today (had to open the system 4-1/2 yrs ago to replace the heater core) and since the engine had to be run, it seemed a good time to hook up the scope. Had the engine running with a low current charger to get the battery up to a more normal road voltage. Measured 10.5A ignition draw with 13.8V at the battery. That's roughly 1/3 less load than with the transistor modules hard grounded -- I'm satisfied.

Separately -- I recently extended the 10 AWG battery wire to the left rear of the car and mounted a relay for the fuel pump. I ran 12 AWG to the fuel pump connector, the original power wire to the relay coil, and re-wrapped the connector wires. Could probably have gotten away fine with a 25A fuse, but it's now 30A, same as the OE circuit. I think that leaves the original ignition circuit with only the instrument cluster.