Freshly rebuilt EJ22 seems anemic

[frag]

Ok, so I got the Fuel pressure gauge and tested last night it's right on at 32lbs. I assume that even thought the pump is capable of over 60 lbs that the pressure regulator on the fuel rail is releasing the excess hence the results of the test.

 

If you still have the pressure gauge, briefly pinch the fuel line downstream of the gauge (befor or after -return line - the regulator, taking the regulator out of the equation) and you will see if your pump can give its full output, which it probably can.

[hankosolder2]

To go somewhat OT- what kind of a kit car is this? VW based, or not?

At 2100lbs, it should be far from anemic.

[pvenuti]

The engine is in a replica of a Porsche 356. No VW frame it's a custom steel frame purpose built for this application.

 

Before I got the engine in this is how the car got around.

 

Engine in it's new home. Since this photo I have moved the overflow and haved it mounted properly.

 

 

I'm curently waiting on a replacment Idle air controler as it's tough starting when the engine is cold. But I dont know if this will have any effect on my low end power issue.

Edited September 6, 2009 by pvenuti

[hankosolder2]

Cool. That should be MUCH faster than any stock 356 once you get it running right.

 

 

Nathan

[ericem]

Nice car and very clean ej22!! Sounds more like a Ecu coolant temp sensor issue to me especially with the difficult cold starting issue.

[pvenuti]

Thats what I thought originaly but I replaced the sensor and still the problem persists. I'm expecting a new idler tomorrow, hopefuly that will fix the cold start problem and maybe more.

[pvenuti]

Problem recap and troubleshooting summery to date:

 

Just a quick refresh on the problems I seem to be having with this newly rebuild EJ22:

 

Low power below 3k RPM.

 

Hard to start when the engine is cold. QAT can be 70 and it takes at least 4 tries and fiddling with the peddle. After its warm the engine starts respectably.

 

Something I tuned into today while troubleshooting slow to return to idle. When the engine is revved up it slowly comes down to 2k then 1500 and then back to idle at 900 over a 10 second period maybe longer.

 

Troubleshooting began with cam timing as I was initially only looking for the low end power. Found that the right cam was off by one tooth and was told that this could be cause if I mistakenly rotated the engine backwards. Fixed that and there was a slight increase in power at the top end.

 

Thought I had a bad CHT sensor feeding the ECU bad data so I replaced that with no improvement anywhere including cold start as we had hoped. Today I retested the new CHT with the Ohm meter it’s fine and was able to confirm 5v at the connector from the ECU. With a warm engine I pulled the connector and started it and it idles at 2500 rpm, I guess that it is doing something. I may try this when it’s cold as well just to see what symptom it produces.

 

Fuel pressure was confirmed at 32lbs while running at idle. If I have time tomorrow I may test again and pull the vacuum line from the pressure regulator to see if that is working and how much pressure the pump really will make.

 

Next on the list of possible culprits for cold start was the IAC(idle air controller). I received 2 used in the mail tried them both no change. I'm however able to replace one in about 30 minutes now down from my original time of an hour.

 

Today I tested engine vacuum and confirmed that it’s nice and steady at 21 inches. A quick snap of the throttle and it drops to 13-15 inches and when the throttle is released it shoots up to 25 inches. So I feel pretty good that it’s not sucking or leaking air where it shouldn’t be.

 

While I was at it I checked the ignition timing and found it to be about on speck. I say about because the manual says it should be 18-22 degrees BTDC at 700 RPM. My engine was idling around 900 per the tach and I had 20 degrees BTDC. As you increase engine speed the degrees increase BTDC so I assume if I had lowered the idle it would have been at 18.

 

At this point I’m fresh out of ideas of what to test so I think I’m going to order a cable that will allow me to connect to the ECU and see what’s going on with that Select Monitor software. I don’t know if this will be of any use but I don’t know what else to try.

 

The last piece of information I can offer on this engine is that I asked for a set of Delta cams to be installed so that should modify engine behavior as well. If anyone has any experience with this I would love to hear about it.

[89Ru]

Well, now you have a couple of spare IAC's.

 

still not great on the low end torque. For a smooth start in first I need to bring the rpm’s up to 1500. Power really doesn’t develop until 3k and then it starts to pull strong(relatively speaking).

 

Unless a dyno is involved, power is rather subjective. I'm not trying to mask a real problem, but the ej22 isn't noteworthy at the low end in my experience. Low end torque began with the ej25, and with it the corresponding hassles, like interference. Of course with your low weight, this statement is probably irrelevant. I'm driving wagons. :-\

 

For the hard starting, have you checked compression or done a leak-down test?

[pvenuti]

I have not done the compression test but I can. I just assumed that if it starts once it’s warm that compression isn’t in issue. I'm also under the impression that compression tests should be done one warm engines.

[89Ru]

if it starts once it’s warm that compression isn’t in issue. I'm also under the impression that compression tests should be done one warm engines.

 

True, compression (done warm) isn't a conclusive test, its a data point. Leak down would help diagnose if bad valves are suspected, won't run at all, etc. I was mainly thinking if there is a timing issue, however remote, it would show up in these tests.

[pvenuti]

Did the compression test on a warm engine this evening and here’s what I found.

 

Cylinder #1 150lbs Cylinder #2 205lbs Cylinder #3 160lbs Cylinder #4 203lbs

 

So the right side of the engine is between 40 and 50 lbs stronger than the left. I checked #2 twice because I couldn’t believe it. When #4 came up with a similar reading I was just at a loss.

 

I’m flipping through the Haynes manual and can’t seem to find what standard compression should be. Does anyone know what it should be in a standard engine? I don’t think mine will be standard as I believe the heads where shaved as part of adding the cams.

 

So I guess it’s onto the leak down test to determine if #1 and #3 are leaking or just not as tight? I guess I’m hoping for poorly adjusted valves… I don’t have the tools for a leak down anyone have an idea of what I should expect to pay for something like this?

[89Ru]

Harbor Freight has this one for not too much $$. Can't vouch for accuracy but I'd imagine a relative measurement would be helpful.

 

94190-0RHH Cylinder Leak-Down Tester

Not sure why but the printed catalog sometimes has better prices than the website.

 

Search around on the board for expected compression values.

[OB99W]

I’m flipping through the Haynes manual and can’t seem to find what standard compression should be. Does anyone know what it should be in a standard engine? I don’t think mine will be standard as I believe the heads where shaved as part of adding the cams.

Typical compression pressure (assuming good engine condition and proper procedure) is about 185 psi. Bad rings or valve seating can obviously reduce that. So can the timing being off. If your engine should be closer to 200 psi, then a one-tooth error in the t-belt installation at the cam on the side that's low could reduce it to about the amount you're reporting.

 

I know you've already checked and corrected the timing, but did you go by marks alone or did you do a tooth count? It's usually best to use the marks initially, but verify by counting teeth.

 

As suggested, leakdown test results could be useful. If you didn't try it already, a wet compression test might also reveal something.

[pvenuti]

Ok so I pulled the cam covers and confirmed that timming is still good.

 

Also that a shot of either into the throttle body when it's cold and this engine snaps right to life. Revves high then low almost stalling then recovers.

 

I'm lookinag around localy for someone to do a leak down or a place to purcahse the device tomorrow.

[pvenuti]

Still trying to find a place to get a leak down done. In the meantime found out that one of the injectors isn’t firing. I'm getting a pulse but no gas. When I pull the connector on the other 3 injectors the idle drops but not on this one. I was able to confirm spark here as well.

 

I checked it with the Ohm meter and it has a little resistance where the others that work do not. I assume this means it bad. Anyone know what the proper test is?

[hankosolder2]

All the injectors should measure very close in resistance. (I'm assuming you are measuring the resistance of the injector coil at the injector itself.) Probably +/- 2 ohms or so between the four. If you notice any significant difference, it's almost definitely bad. I don't think they're a high failure item on Subarus, but I have changed a few Bosch injectors on older BMWs over the years.

 

Of course, injectors can also get plugged and fail mechanically.

 

One thing to note- some Subarus (perhaps not going as far back as the donor vehicle for your car) have software that shuts off the drive pulses to one injector if the idle RPM becomes uncontrolled. (i.e. if the computer senses the car is revving at over 2500 rpm (don't recall exact RPM #) with the throttle shut, it disables one cylinder to bring the idle down. I ran into this issue with my '98 Outback...my replacement 2.2 had a sticking idle air control valve and on starting, the car would run at high RPM for a while, throw a check engine light and begin chugging on 3 cyl until the battery cable was disconnected. So, assuming your injectors check out OK, verify that the cylinder which doesn't seem to be working is getting an injector pulse.

 

Good luck,

Nathan

[pvenuti]

So testing it on the 20k ohms scale on the meter 3 of them test too zero while the bad one is at 40. So I'm going to buy a new one today and cross my fingers.

[hankosolder2]

Is that a typo? The 20K ohm range on your meter is not the right one to use. An injector should be perhaps in the range of 5 to 30 ohms...so your meter should be set on a low resistance range.

It does sound like the bad injector is essentially open though, so I think your conclusion is right even if your methods are wrong! ;-) If you're using the 20 K ohm scale, you're doing a simple continuity test and any good injector will read zero.

[pvenuti]

Not a type O that’s user error on my part with the meter. Where should I set it?

 

Ordered the injector this morning for a 92 and it's totally different from what I have. Looking at the photos at the parts store it would appear that the injectors I have are for a 91. Guess I will have to wait until late next week before I can get an injector.

[hankosolder2]

I don't know the specific scales/ranges for your particular ohmmeter, but as I said, the resistance pretty much any injector ever made should be somewhere between 5 to 30 ohms, so select a range/scale which is slightly higher than the resistance you expect to be measuring- perhaps a 50 ohm range would be a good choice. Basically, you want select a range so the meter is not pegged all the way at zero or infinity so you can get a valid reading. Some inexpensive meters do not have a good scale for measuring resistances below 1K ohm (1000 ohms.)

 

You might be better off with a modern, auto ranging digital multi meter. The Fluke brand is really nice, and you can get a model 115 or 73 for less than $150. Just set it to the ohms scale, probe your injector and you'll have an accurate readout. They also make a model specifically for auto diagnostics, but it's more expensive and probably not that much more useful...IMHO.

 

I wouldn't worry so much about the absolute reading of your injectors- just make sure that they all measure within a couple of ohms of each other. A Haynes manual probably has a specific spec- I'll try to look it up tomorrow.

 

Nathan

[hankosolder2]

Per the Haynes manual, the fuel injector resistance for a '90-'94 Legacy should be 11 to 12 Ohms. Hope this helps.

 

Nathan

[pvenuti]

Thanks for the specs on from the manual. I just got a used one in the mail and will be puttin git in on Saturday. I did mesure the replament at it checks out at 12.5.

 

Should have an update on Saturday.

[pvenuti]

Got the injector installed now it seems that I'm troughing codes all over the place(or maybe not able to clear them). With the new injector installed it seems to be running better. I can pull the connector on the injector and her the engine idle down and back up when re connecting so I feel they are all working now. Some injectors make a more noticeable drop than others. Anyway now I have a few codes that I can’t seem to clear. I have 21,22,14,15,16,17. 14-17 are the injectors I'm guessing I have these because of my test with unplugging the connectors. 21 is coolant sensor that I have already replaced and confirmed that the connector is getting 5v. The last is the Knock sensor(replaced that today with a new one form AutoZone). In the past I thought I have been able to clear codes by connecting the D-Check and Mem-Check wires at the same time. I do this then start the car and the codes are already there. I read in the Hayes manual that I need to disconnect the negative battery cable for 30 seconds and this should clear the codes as well but I seem to get the same results after disconnecting it for over a minute. Does anyone see what I might be doing wrong here clearing these codes? Could these be a harder code to clear requiring another process?

[Fairtax4me]

You need to probe the connection at the ECU to see what it is getting back from the Knock and temp sensors. Testing for voltage at the harness connector, and resistance across the sensor, doesn't tell you anything about what is happening between there and the ECU.

 

There's a factory service manual in here for the 1992 model year called "92 legacy full". It may have some hints for diagnosis. http://cid-4ca3c3459aaa7f7f.skydrive.live.com/browse.aspx/.Public

Edited October 3, 2009 by Fairtax4me

[johnceggleston]

disconnect the battery cable overnight to clear and reset the codes. if you get them back after that, then you still have a problem.