Purge solenoid INSANITY!

[GeneralDisorder]

I have thought about reverse engineering an ECU.

 

Yeah, I've thought about it too. But I'm guessing that without the code-base for it, it would take months of work just to get anywhere with it's software. Although if it's something simple that an assembler was easy to find for, it might not be too hard to re-write the firmware for it, or locate the sections that reference the EGR and Purge solenoid control lines and just eliminate that section of code. It's probably a sub-routine that could just be NOP'd out and be made to return a valid "I'm ok" when called.

 

In a few years, I could do a megasquirt, since the cars will be old enough to skip emmissions.

 

I think that's going to be my solution as all my rigs are not under test laws anyway. More configurability for my modified cam, etc. Plus the ability to turbo the EA81. The software of the SPFI ecu doesn't handle positive manifold vacuum situations well from what Rguyver has told me.

 

GD

[pisces_0]

I drove mine with a 68 ohm 10 watt resistor for almost two years; the CEL never came on once during that time (at least with an EGR code).

So I hit up the 'Shack yesterday and picked up two five packs of 100 ohm, 1/2 watt resistors. Running three in series, as youareabus mentioned above, brings the resistance to 33.3 ohms, right smack in the middle of the factory spec for the purge and EGR solenoids (32.7-39.9 ohms). The wattage is quite low for the power that's being put through the solenoid, and the resistors do get mighty hot, but nothing a few nice 35 ohm, 10 watt resistors from Digikey can't remedy. So, now on to the good stuff...

 

Like I mentioned above, I've battled an intermittent check engine light for the EGR solenoid for quite some time now. After putting it off for much too long I bucked-up and did a little diagnosis. When the engine is cold the resistance across the solenoid is right in spec at 34.5 ohms. Once the engine, and thus the EGR solenoid, heats up the resistance goes all screwy and triggers the check engine light.

BINGO! Problem diagnosed.

 

So I cleared the codes as per the manual by the diagnostic-plugs-connected, WOT-gas-pedal, 1/2-throttle-gas-pedal, drive, blinking-check-engine-light method. ECU codes cleared and now on to some experimentation.

 

I removed the plug from the faulty EGR solenoid and inserted the in-series resistors into it. Cranked the car over, let it warm up and no check engine light! After a little of a drive to fully test it, the check engine light never came back even after a decently long drive and hot engine bay.

The only downside, the resistors get HOT. The three 1/2 watt resistors just don't seem big enough to handle the power being put through the wires. They may work, but I'm not sure for how long and how safely. Reference the Digikey resistor mentioned above.

 

In the interest of experimentation, I figured I'd try the same thing on the purge control solenoid. It's still in good, working condition and specs out just fine as far as resistance goes, but my curiosity was now piqued and I wanted to see if I could duplicate the results on this solenoid too. With the engine running and the EGR solenoid bypassed with resistors (and no check engine light) I unplugged the purge solenoid. BANG, check engine light! Plug it back in and no check engine light. So I unplugged the solenoid again, the red light of doom pops on, plugged in the resistors and the light went out! Both solenoids were now unplugged, the resistors were in place and no check engine lights for as far as the eye could see!

 

I'm going to pick up some of the 10 watt resistors from Digikey and put some mileage on them to test the longevity, but for initial testing purposes the resistors seem to work perfectly!

 

Cliffs Notes

For those of you who have fought check engine light problems due to faulty solenoids and don't feel like scavenging the boneyards or want to clean up the engine bay a bit (my main reason; no emissions in MN), a 35 ohm, 5-10 watt resistor can be used in place of the botched solenoid(s).

[GeneralDisorder]

So maybe all along I was just giving it too high of a resistance. Perhaps the solenoid I had was also out of range. I do have some 5 watt 33 Ohm units here. Maybe I'll give that a go.

 

That's sounding likely actually - I don't get the code in regular driving, ONLY in D-Check. So the D-Check diagnostic is probably a lot more precise in what it will accept for the resistance of the unit. Sounds like 33 Ohms is the ticket. I'll give it a go and see what I come up with. They have been sitting here on my desk for a month now. I'll get em installed and see what happens.

 

GD

[Gloyale]

My vote: Send the Hitachi gear to hell and go for MegaSquirt.

If the same dudes who designed the carbs programmed the ECU - it could take years for you to solve this one.

 

Doesn't JECS make the ECU's and the TPS, MAF, other stuff? The hitachi carbs are insane to work on. their distributors are good though. they are way better than DENSO!

[GeneralDisorder]

Every single component of the SPFI system has the Hitachi logo. If they farm out stuff..... well who knows, but I would think they would just develop stuff like that in-house being they are big in electronics - or were at that time anyway. There's no way to tell without contacting subaru of Japan, or Hitachi directly... neither of which would be likely to give a crap about finding out anything for you, and likely would be fearful of NDA's so wouldn't say squat.

 

And as far as ND stuff - that depends on the part. A lot of ND things (such as the ND EA81 distributor/coil setup) I actually prefer over the Hitachi units. The original Denso starters are pretty damn rugged too.

 

GD

[daeron]

It seems to me that most major Japanese manufacturing firms are pretty reliable in general. All have products that aren't their best, and all have bulletproof products. Alot of them have changed over the years, too. Hitachi makes fantastic alternators, starters, and electrical equipment. With carburetors, you have to consider that ALOT of Japanese carbs were just ripoffs of European models anyhow... knockoffs. What do you expect? The FI system in general is liable to be a grey area.

 

My point is, I try not to make any rule about a japanese company. They change too much, and in the long run their ideas and philsophies seem to work well for me to ever pass judgement on them.

 

Except for honda. F*** honda.

[pisces_0]

Just to wrap this thread up. I put off this project for much too long, but have good news.

 

I removed both the EGR and purge control solenoids and clipped off the plugs on them, leaving 1"-2" of wire. Then soldered the 3 parallel 100 ohm resistors between the two wires and shrink-tubed the solder joints. I plugged in the two contraptions, cleared the error codes and started the car up. The check engine light has been nicely unlit for a week now and I couldn't be happier.

 

For those looking to clean up the engine a little bit and/or remedy a nagging CEL, this may be the ticket!

[hooziewhatsit]

awesome, glad it worked for you.

 

I have 40ohms in place of my EGR solenoid, and still have a CEL I think it's something inside the ECU causing the error.

[GeneralDisorder]

awesome, glad it worked for you.

 

I have 40ohms in place of my EGR solenoid, and still have a CEL I think it's something inside the ECU causing the error.

 

40 Ohms is too high. Needs to be about 33.

 

GD

[hooziewhatsit]

hmm, I tried 30 before that (only have 10ohm, 10W resistors) and that didn't work. Guess I'll get 3 100ohm'ers next time I'm at Rat Shack.

 

-Dave

[pisces_0]

hmm, I tried 30 before that (only have 10ohm, 10W resistors) and that didn't work. Guess I'll get 3 100ohm'ers next time I'm at Rat Shack.

The resistance for the solenoids needs to be between 32.7 and 39.9 ohms. Looks like you hit either end of that spectrum so far! :-p

[Skip]

Hoosie, why not grab a 0-100 ohm pot and dial it in??

[hooziewhatsit]

Hoosie, why not grab a 0-100 ohm pot and dial it in??

 

a pot of that wattage I'd think would be pretty spendy. I also just realized I have several 100ohm 10w resistors, and a dead solenoid I can grab the connector from. Wish me luck

 

-Dave

[Skip]

Dave, wattage in this case is very small.

 

cost about 3~ 4 dollars, resistors only cost pennies

 

=> therefore you are correct pots are spendy.

 

rheostat is a high wattage pot, this is not called for here

 

but they are about the same cost - read expensive.

 

You have the resistors mix and match and

maybe you will find the right resistance.

 

I like keeping my systems intact and use

a Honda vesion of the valve in many of my cars.

[hooziewhatsit]

last night I found two 100's, and two 50's, which I can make work at 33.3ohms.

 

If this works, I'll find yet another solenoid and try to hook the system back up. I've already tried a couple 'good' solenoids (that didn't work), and never got the CEL to go away.

 

-Dave

[msteel]

Dave, wattage in this case is very small.

 

Great...I have to either contradict GD's resistance number or contradict Skip's wattage claim. Well, here goes:

 

If 33 ohms is the correct value for a resistor, then there will be significant power dissipated in the resistors:

 

12 volts divided by 33 ohms is 0.364 amps of current - not at all out of line for a solenoid. 0.364 amps times 12 volts is 4.36 watts. That kind of wattage would certainly heat up, and possibly fry the three recommended 100 ohm quarter-watt resistors.

 

A couple of years ago I had my purge solenoid start generating a CEL. Using a vacuum pump/gauge it seemed to work fine, too - when it was cold. But the coil became an open circuit after the engine got up to temperature. I took it off the car hot and verified that it didn't work. After it cooled down a bit it worked fine again. I have no idea if this is common with these solenoids. But for me a replacement solenoid stopped generating CELs.

 

As a side note, before I replaced the solenoid, I took a 1000 mile trip without the solenoid. I decided to connect vacuum straight to the cannister valve (the equivalent of the solenoid always on). My mileage certainly didn't suffer and in fact I wonder if I didn't get better than normal fuel economy that trip.

[Skip]

Props on the math msteel

but the pots I was talking were

all 10 watt pots.

 

I guess in the world of Einsteinian reality the

statement

"Everything is relative"

rings loud.

 

To me, 4 watts is very small -

to a PC board

builder it is not so true.

 

BTW P also can be found directly => as E ^2 / R

[GeneralDisorder]

I used 5 watt 33 Ohm ceramic flame-proof resistors. Seems to have worked fine for me. The package of two was a whopping 89 cents.

 

GD

[pisces_0]

I spoke too soon when I said three 100 ohm, 1/2 watt resistors did the trick. Was under the hood last night with the engine running and noticed a slight burnt smell coming from the engine. Immediately I looked at the resistors in place of the EGR & purge solenoids and sure enough, one was fried to a crisp!

If 33 ohms is the correct value for a resistor, then there will be significant power dissipated in the resistors:

 

12 volts divided by 33 ohms is 0.364 amps of current - not at all out of line for a solenoid. 0.364 amps times 12 volts is 4.36 watts. That kind of wattage would certainly heat up, and possibly fry the three recommended 100 ohm quarter-watt resistors.

I think you're right in saying that three 1/4 or 1/2 watt 100 ohm resistors won't cut it. Looks like I'll be ordering some higher wattage 33-35 ohm resistors from Digikey!

 

For the archives, using the correct resistor(s) in place of the solenoids works just fine, you simply need to use higher wattage resistors than are commonly available at Radio Shack. Just my $.02...

[msteel]

the pots I was talking were

all 10 watt pots.

 

To me, 4 watts is very small -

to a PC board

builder it is not so true.

 

BTW P also can be found directly => as E ^2 / R

 

Aha...10 watt pots yes, they would not have a problem. That would be about the minimum I'd recommend if you want a reasonable safety factor. And yes, there are lots of forms of Ohm's Law and the Power Law.

[psylosyfer]

I think that you guys are too hard on Hitachi, one of the foremost engineering and production companies in the world; I would take them over Bosch any day. AND...

 

...I also think that you are placing too much complexity in the ECU. It was designed during, and using components from, a relatively primative era in electronics, processors and support parts. I doubt that there is any provision in these early ECUs for anything more than simple on/off monitoring (using a simple voltage comparator); almost certainly not an ADC circuit needed to properly monitor whether it had activated normally.

 

I am mildly surprised that the 10W resistor is warming up. Power dissipation should be around 3 watts. My surprise isn't so much that 3 watts would raise the temp noticably, but that the little plastic package of the original would be subjected to that much warming. Is it possible that we are misjudging what is needed for the solenoid's fake-out? Is it possible that the ECU is basing the solenoid's functioning off of an inductive EMF "buck"? Admittedly, this would not explain why the code is happening with assumed-working solenoids.

 

Sorry for rambling... sleep deprived. :-\

Nobody likes their carbs but I can say in the world of consumer electronics(tv,vcr etc.) they were the longest lasting, best built easiest to repair brand of all, Matsu************a next, and contrary to most laypersons opinionof sony they are about the worst.

[GeneralDisorder]

Nobody likes their carbs but I can say in the world of consumer electronics(tv,vcr etc.) they were the longest lasting, best built easiest to repair brand of all, Matsu************a next, and contrary to most laypersons opinionof sony they are about the worst.

 

Yeah - the SPFI isn't so bad once you understand it's needs. It's not the system that has a problem, but rather the badly translated FSM's that make me scratch my head in wonder when trying to follow their troubleshooting. They are badly organized, with pertinant information often being located no where near the section it needs to be in. Basically you have to read the thing cover to cover, and then read between the lines to infer the information you actually need.

 

Engrish sucks!

 

GD

[daeron]

Yeah - the SPFI isn't so bad once you understand it's needs. It's not the system that has a problem, but rather the badly translated FSM's that make me scratch my head in wonder when trying to follow their troubleshooting. They are badly organized, with pertinant information often being located no where near the section it needs to be in. Basically you have to read the thing cover to cover, and then read between the lines to infer the information you actually need.

 

Engrish sucks!

 

GD

 

Well, I highly doubt GD would come away from it with much more knowledge than he's already got.. but ANYone else who would like a good thorough read on how fuel injection works, to help supplement their knowledge from the FSM, can click here for a pdf of the datsun 280Z's "fuel injection bible," Its a thorough factory shop manual for the Bosch L-Jetronic FI system installed on the Z.. its different from ours, because it uses a flapper door AFM (not a proper MAF) and an ANTIQUATED TPS... (three positions.. idle, run, and WOT) but the ideas are all VERY well explained there.

[hooziewhatsit]

well, been driving for a while now with (measured) 33.3 ohms in place of the EGR solenoid, and still have the EGR code (three 100ohm, 10W, in parallel, soldered to EGR connector)

 

Yes, I've cleared the codes through disconnecting the battery, driving with the greens connected, etc, it's still there. Just put 500 miles on it, so if it was fixed it should've gone away.

 

Someday when I have time to dig into it, I'll make sure again the wires from the solenoid make it to the ECU.

 

I have an ECU from an 87 DL SPFI. If the numbers match, I may try swapping into mine (88 GL).

 

-Dave

[stephenw22]

For me, a 47 ohm, 4-watt resistor made the CEL light go away and never come back. Maybe the wiring harness condition has something to do with it as well.

 

When I originally did it, I just bent the resistor leads into little loops, and then plugged the resistor into the plug on the wiring harness. I was planning on wrapping it with some tape, but I never got around to it. That was about 2 years ago, and still no problems.