'90 Honda accord quit running

Sorry... I mangled that first part of the sentence - rearranged it and left
the word "shaft" as the 3rd word by mistake.

It seems similar to the Honda system but from the actual picture I can't
make out the corresponding parts. The Honda multi-tooth reluctor is well
portrayed here:
<http://www.markl.f9.co.uk/howto/electrical/igniter/ign9.jpg> and the
pickup is right behind it to the left. I think I can just make out one of
the lobes of another reluctor under the base plate. There's also the fact
that since that early HEI system, with mechanical and vacuum advance, we
now have the variable ignition curve profile(s) in the main ECU, so the
reluctor/pickup sensor signal gets sent there first and the igniter is
"driven" from there.

There *is* a bigger picture, here:

The big (power) and small transistor are kinda obvious - no idea what the
IC might be.

Yep - that's a mystery. I wonder if it could be some kind of electrical
insulating/thermal conducting material which got broken down with the heat
cycles. The igniter does have an external heatsink plate so the heat has
to get to it somehow.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

Good intution there, George. The dark (but faint) lines are laser
trimming
scars which bring the resistor to within a very close tolerance of its
design
value by measuring its ohmic resistance while laser-cutting!

The high temperature baking that happened previous to laser-trimming
is the same process that carbon film resistors of the conventional type
undergo.

This one became reluctant to start in cold weather. It wasn't the
original
part, however. I'm a pragmatist! "It's dead, Jim." I haven't got a test
rig
to make it work and the new one gets me running again.

The small IC is a Telefunken U2226B but I don't know what is in it and I
haven't seen my TFK databook for 15 years! Vishay acquired them
in 1998.

Later today, I'll put a drawing of the distributor components and wiring
(which I got hold of) on my website for download. I'll put a link in the
text.
This shows a how-to for testing. I'll add a note to this thread when I've
done it.

 

I've placed a link to 'ign.zip' about a third of the way down in the text.
It
contains the drawing referred to above, a photo of the LED probe and
a text file with further notes. Look for "download this ZIP" hilighted.

Rgds

 

"George Macdonald" wrote

Yeahbut I got the idea.

So there's no ECU in the General Motors HEI system? And this is one of the
big differences between the Honda and GM systems (as well as the drawings I
have cited)?

Is the ECU located under the dash in 90 91 or so Hondas?

Sorry I know these are basic questions.

Also, as Gordon pointed out to me, we are well off the topic of just getting
someone's car to run.

I for one am in this strictly for the education (self-taught and taught by
others, but encouraged to research based on what's been said here) at this
point.

Thanks. I've tried several times in the last few hours and the danged thing
won't come up. Will keep trying.

I have been looking at Gordon's schematic (he emailed me one), too.
Hopefully it's up at his site now.

snip

The two electronics newsgroups where I asked about this yielded only one
interesting response but not especially on this.

I'm sure the following is way more than most people wanted to know, but if
anyone wants to comment, I'm reading.

The interesting part to me was to do a multi-meter diode check of the
ignitor's four terminals to try to identify which connect to transistors. As
I mentioned, I happen to have a spare ignitor from my car. It is old but it
was working when I took it out. Unfortunately, I can't open it up. Anyway,
hopefully all the following indications are what a new, properly functioning
ignitor would yield. It has four terminals. I will call them T1, T2, T3, and
T4. The following "checked out as" diodes. Anodes (= + sides of diodes) are
listed first. That is, for T4T1, the voltage appeared on the multimeter when
the red lead was connected to T4 but not when connected to T1.

T4T1: diode, 1.5 volts
T3T1: diode, 1.3 volts
T4T2: diode, 1.0 volts
T3T2: diode, 0.6 volts

Dunno if the voltages are all that meaningful. Seems like they might give
clues as to which is the base, which is the emitter, etc. But I don't think
one can come to any conclusion without knowing more about how and whether
the small transistor and Power transistor connect. From some other
commentary the author has added (and which I believe he posted to his site
yesterday), I have some ideas, but I don't want to throw them down as even
after a lot of thought the configurations I have in mind do not quite all
check out. Yet.

The terminal numbers are consistent with those on Gordon's schematic.
Alternatively, looking at his site's photo, the tab that is by itself on one
edge is T1, the next closest tab is T2, etc.

Are you saying maybe the gunk/gel we see in Gordon's site's upper ignitor
photo might help conduct heat from the electronics in the ignitor to the
heat sink, and then ultimately to ambient? Again, just checking to see I
understand.

I do understand from a lot of chatter on this that the heatsink plate is
important, ensure you apply the heat sink grease carefully, etc.

 

Yeah well I think that GM system being described at the Web site is an
early HEI, which still had mechanical advance/retard control... for
comparison with the old contact breaker systems

In Civics it was located inside the cabin, just above the front passenger
footwell as it ramps up to the vertical firewall - you pull the carpet back
to get at it. In Accords I believe it may have been under the passenger or
driver seat. You might be able to get a clue from the diagrams at

As long as we're all learning from each other and from poking around the
Web for better info.... it's interesting.

You mean Graham? I meant that possibly it used to do that... before it got
umm, wasted. I guess it's also possible that it was transferred to a fresh
distributor body after the previous one leaked oil all over and that the
"stuff" is just oil gunk.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

Chipdir at <http://www.rabidpenguin.org/chipdir/index.htm> doesn't turn up
anything... nor <www.chipdocs.com>. Apparently Telefunken was first taken
over by Temic, which was later divided between Continental, Atmel and
Vishay. Unfortunately Vishay has apparently abandoned the Telefunken part
numbers. I think it might be an optocoupler... as described by Koji San.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

I have a friend who did his PhD thesis on something to do with integrated
circuit/chip manufacture and continues to work in the field. (He's fairly
"hands on" in this area, as it happens.) I described the gunk we see at
Graham's site, without giving any theory as to what it might do. He
immediately said 'probably encapsulator,' and he'd just been to a conference
that covered gel-type encapsulators. He also immediately observed that the
ignitor sees pretty heavy duty as far as being exposed to harsh
environmental conditions, so such an added encapsulator made sense. (I told
him, 'Yes, well, since the thing still fails so often, it's not enough.
You'd think Honda would locate it better...' As many have noted.) From a
net search, perhaps the gel encapsulator is to be distinguished from the
classic use of only a plastic case as an encapsulator(?). In either case,
the purpose is to keep moisture (and air, dust, etc.?) out. I asked whether
it might also serve to conduct heat away from the electronic components. He
said that was very much possible.

 

San.

You've been all the places I have, George! But it isn't an optocoupler.
It may be a dual Schmidt trigger IC. Or even the TFK equivalent of a 556
dual 555 timer!

I've done a close up photo of the thickfilm and labelled the terminals.
I'll find a place in the article to show it.

Rgds

 

[... big snip...]

No, they aren't since the unit wasn't powered and those circuits which
your tester has picked up are spurious and meaning less. I am
a little surprised that you didn't find that a circuit exists between T1
and T2 and is exactly the same value which ever polarity. Moreover
you ignored the 0v connection!

I think the advice you got in the single reply is more pertinent to
discovering the connections to a darlington pair rather than a
thick film circuit.

The new photo should help.

That 'black stuff' is not any part of the spec of the module! It is what I
said it was - grit and dirt and muck and oil. I have no idea how it got
in there but I don't have a past history of that module and if the dizzy
case can fill up with oil then that may explain a little about its origin.
On the other hand, someone may have steam cleaned an open dizzy
case to make it shiny and the crap got in there as a result. The module's
lid is only a light click fit on four legs.

And who's this Gordon guy? 8¬)

If anyone can find out what a U2226B is, I'll be able to reconcile it
with the components around it.

Rgds

 

Now ask him if that 'encapsulation' washes off in white spirit with a 1/2
paint
brush.

I'm beginning to think you are a troll...

Rgds

 

If I know the basic circuit layout but not which terminal goes with what,
then the voltage drops might tell me which of the paths have the most
resistance, indicating which terminals go with which transistors.

Now I didn't check for a circuit. I checked for "one-way" current flow (a
diode check). When set on the diode symbol, a digital multimeter will pass a
current through the leads and whatever they leads are attached to, and then
measure the voltage drop (as you know, but to put us all on the same page).
No voltage drop, no current flowing. My results above only indicate the
terminals and their orientation indicating "diode action."

When I did the T1T2 and T2T1 diode check, I did indeed get the same voltage
drop for each, about 0.93 volts.

And, performing a bona fide continuity check on each pair of terminals
indicates a very high resistance, regardless of polarity. They're
effectively opens when unpowered.

I would have to read a little more on how transistors work to see if this
makes sense. Maybe you know right off the bat it does...

I checked my multimeter's continuity feature on some lamp cord to see that
it was working. It is. I also quadruple-checked or more all my readings
above.

Oops.

G = ground below.

GT1: diode 1.3 volts
GT2: diode 0.5 volts

Continuity checks:
All opens.

The author of the single reply was interested in helping identify which
terminals might connect to transistors. "Might" is the operative word.

I hear you.

You know about this encapsulator stuff, though? Ever hear of a gel-like
substance applied to keep moisture and dirt out and maybe conduct heat away?

Absolutely no way it's that?

Since ignitors are often said to be the achilles heel of Hondas, then I
doubt this happened to all Honda ignitors.

I figure they either all come with the gunk or they all come without it and
fail other ways.

I've tried to open mine up using your directions but remain resigned to the
Koji hammer approach. (Which I'm not going to do, as my ignitor supposedly
still works.)

Slightly different module, perhaps?

Someone who definitely does not call distributors "dizzys".

;-)

Worth a post to my electronics newsgroups, where this sort of question seems
more typical. You post it. I'll read.

sci.electronics.basics
sci.electronics.components

 

Mine reads 2,200 Ohms and is the black patch on the right in photo.

This is where a circuit resistance range is more useful than the diode
range on the test meter. That way, you'll get a reading on one range or
other on the Ohms, kOhms, megOhms scales.

After 45 years of this electronic engineering stuff, I'm starting to get
the hang of it! 8¬)

[...]

As far as the beeper goes, but what does the resistance range say?

I've already indicated which do earlier, perhaps you missed it or
just didn't believe me.

Yes. Potting compound is another descriptor. I've never seen
an encapsulation product that readily washes out with a little
white spirit. It's muck and oil in there.

It is what I said it was. Why can you not accept it?

Why should it happen to all? There is more than one component
in there and any of them could fail.

There are more than one maker of these things and each has his
own ideas as to what potting compound to use, if any. I haven't
seen a genuine Honda module since my car is a Rover.,

OK, I'll post in there to see if anyone has an old catalogue to hand.

Rgds

 

The reliable NEC uses no gel. If it serves to conduct heat it is not
used here.

The NEC and older units use a visible diode (likely a clamping zener)
placed between collector and in series with a resister to ground. The
resistor is there so that the diode does not melt! Igniter failures
are mostly due to the zener and resister absorbs most of the power
caused by plug wires fallen off, damaged coil, etc.

Newer units use a clamping zener inside the NPN package as what you
refer as to the larger black box.

T1 = Tachometer (output)
T2 = Collector (-output)
T3 = Voltage input (+12V input)
T4 = ECU (digital input)
G6 = Emitter (Ground input, not shown)

The ECU triggers the igniter*s IC, the IC triggers the base of the
NPN, the NPN triggers the coil, etc.

The positive crosses at one point. In fact, the igniter*s ground
connection is not drawn in.

The small upside down triangle on the upper-right of drawing is the
PGMFI-ECU. The bottom right with six triagle arrows also go to
PGMFI-ECU.
http://www.markl.f9.co.uk/howto/electrical/igniter/igniter.htm

Usually under the driver seat or right passenger kick panel, as George
noted.

 

Try charging the coil for 3ms only and then switching off to get
spark. Then try varying the charging time base on rpm. This is how the
IC does it. The charging time required is coil and voltage specific so
the micro controller should apply corrections to the charging time for
most efficient operation. A coil on time can range from 2.8 to 14ms
base on rpm assist by the reference buffer.

The typical reluctor controlled IC should consist of a comparator,
reference buffer, power supply and a protected output stage such as
the Motorola High Energy IC. It is not an optocoupler. I`d
over-stated.

 

I replaced the igniter and tried to start the engine with a few coughs
and sputters but never quite taking off. How would a person rule out
a distributor or timing belt problem?

 

I repaced the ignitor module with little sucess. Trying to start it
will pop
cough maybe sputter but never takes off. Is there a way to rule out
timing belt or distributor?

Thanks.

 

That's the condition mine did on a warm start. If I did get it running
it would pop and misfire and only when I got the revs up did it the
firing smooth ouit. It was only just drivable. As everything reached
running temperature, the effect went away.

When I changed the coil for a second-hand one I was reasonably
sure it fixed that misfire warm start. However, I've since had a
single instance where I had to let it cool off for half an hour.

Have you read through my article under
Miscellanea -> Rover 216 GSi on my website below? If the above
doesn't make sense on iys own, it will after reading the article.

HTH

 

If you have sparks on all plugs then the distributor is fine. If you
have sparks, your timing belt is fine. Straying from your question,
try this:

1. Assuming a pop, cough, sputter indicate an incomplete combustion on
at least three.
2. Try flushing out standing fuel or what*s left in there.
A. Disconnect ECU fuse and fuel pump fuse and crank for 7
sec. This will clean out the cylinders because no new fuel is
introduced. (At least ECU fuse.)

3. Disconnect TW coolant sensor if this warrants. This will aid with a
richer mixture if it*s not already. (Do not run longer than 5 min. or
plug in while running.)

4. Try starting again.

If you believe the timing is disturbed then reset it. A slightly off
timing should start okay.

You might also want to check each plugs once in a while to see if they
spark, since they may die under some stress. At the same time look for
a smell of fuel on the plugs.

Are all four injectors working?

The #1 key factor is good fuel and sparks.
Koiji