Codes 25 and 26

[suprabee]

I have continued codes 25 and 26 that first started back in June. I'm currently looking at the fuel pressure with gauge connected on fuel rail. I have previously monitored the fp while at idle and parked and the numbers were in spec:

7mgte engine, 92 auto tranny - all stock, except bosch bov
38 psi with fpump jumpered on
38psi, engine running at idle, fpr vac removed
30psi with vac on fpr.



some notes: brand new denso o2 sensor soldered to chassis wiring, all other EFI sensors/signals have been checked at ECU using a meter or oscilloscope, new OE TPS, timing correct at 10 D BTDC, have checked coolant for combustion gases, vac leaks searched for and none found, etc..

Some observations:
I've observed repeatedly the occurrence of the codes/check light and it ALWAYS occurs when stopped at idle after some period of driving, only when engine is warm of course.

symptom #1 Sometimes the idle is a bit shaky, or rough, but does not seem to be missing. It feels like the idle "wants" to increase, but because I have the brakes on, it holds the rpms down and it causes the shaking. Feels similar to when you open the throttle a bit and hold the brakes at the same time, the way the car will shake a bit.
symptom #2 Other times, the idle just is a tad bit below 650, but does not feel rough like explained above in #1, nor will it increase if I move to neutral or park.

If, during symptom #1, I move the tranny into neutral or park (from drive) the idle jumps to 800 or 900, the shaking subsides, and eventually the idle will return down to 650. It is at this point, when it returns to 650 (note: idle is now smooth), that codes are thrown, not every time, but enough that I see a pattern.

During symptom #2, It seems that I can "cause" the check light to come on if, while stopped and idling just below 650 I start turning on all the accessories,...head lights, fan, fog lights, A/C, and the rear-defogger which seems to have the most effect. When I turn on the rear-defogger idle drops a bit more, and on several occasions when I did this a few seconds later the codes were thrown.

Today, I decided to drive with the fp gauge installed and see what was happening with fp when the codes are thrown during two symptoms I've described above. I can't see the gauge while driving, but when the codes came, I pulled over and checked the gauge.
I had symptom #2 occur and when I turned on the accessories and the check light came on, pulled over and opened the hood...
fp was 31psi (while in drive with e-brake on), pulled vac on fpr, fp goes up to 40psi, and a couple seconds later the check light goes out. (codes stored of course).

Drove home (1 mile) turn on/off accessories, and after a couple key cycles and turning off and on the accessories (the fault needs to occur in 2 ignition key cycles for the o2 codes to be thrown), I got the codes again. This time I was watching the fuel pressure gauge while I turned things on and off.

When I started turning things on, the idle would drop, and fuel pressure increased. At one point it was up to 32 psi...which is too high, spec calls for 30 psi....However...

I know that the ECU is suppose to increase idle (using the ISCV) when engine is under load, but I don't see this happening. It makes sense to me that when I turn things on, it puts a load on the electrical system, and idle drops, thus vacuum drops on the fpr vac line, causing the fuel pressure to increase just a little. That's what I'm seeing, but I'm waiting for the idle to increase due to the increased load, but it never happens, not even when the A/C is turned on. It also makes sense to me, that if the idle goes too low, the 02 sensor will stop cross counting, or get really slow and thus send the EFI out of fuel control and set off a code.

and finally, the ISCV has previously been bench tested/ cleaned and the bearing lubricated and operates as described on many other threads, it clicks, and when disconnected car starts/runs way up around 1500 rpms.


ok enough for now,
any suggestions?

Rick

 

[Nick M]

Is the oxygen sensor in the stock housing, or further downstream? 25 and 26 together can be a lazy sensor. You said it is new, so I wonder if it is heating up. I was going to say check the fuse on the heater, but the diagram shows direct control. Does your scope scale down to one second so you can see at least 8 counts? Is your car a California model? EGR temp sensor and the post cat sensor will tell you. The rough idle can set 25 and 26 together with a change in engine speed. Sort of like a precursor to P0300.

The accesories are an extra load on the engine.

How many times have you pulled the EFI fuse? 25 and 26 come back every time with it? Your symptoms don't seem too bad. What you are describing seems very close to normal.

Every time you clear the code, fuel trim is relearned as is the idle speed based on age, conditions, etc.

 

[suprabee]

Is the oxygen sensor in the stock housing, or further downstream?

actually I forgot to mention that I have a LIPP elbow, and the turbo exhaust housing only is ceramic coated. I installed a new exhaust manifold gasket, new turbo and LIPP elbow and had the manifold and turbo heat shields professionally ceramic coated back in February and was running the car since them with no codes....so yes, the sensor is in the "stock" location in the downstream. the LIPP sensor location is the same as a true stock elbow.

25 and 26 together can be a lazy sensor. You said it is new, so I wonder if it is heating up.

Actually I've had two problems. One is the codes, and there was a loping idle at warm startup only that would only persist for 20- 30 seconds. I installed the 02 sensor just 10 days ago, and in fact, it resolved my loping idle issue but still leaving me with these codes. I have another thread that I started about the warm loping idle problem.

Question: Does the heater in the sensor only come on at start-up, or can the ECU heat up the sensor anytime based on other parameters?

Does your scope scale down to one second so you can see at least 8 counts?

I was borrowing the scope from a co-worker (who is no longer a co-worker), so I don't have the use of a scope anymore since two weeks ago. However, I did look at the old 02 sensor when I had the scope and at idle it was slow, maybe two or three counts in 10 seconds and the voltage change was jumpy. That's one reason I decided to change the old sensor. When Installed the new sensor, I cleared the ECU memory and they came back right away...well after a three or four key cycles.

Is your car a California model?

Yes

The rough idle can set 25 and 26 together with a change in engine speed. Sort of like a precursor to P0300.

What is P0300? and can you elaborate on "together with a change in engine speed"?

How many times have you pulled the EFI fuse? 25 and 26 come back every time with it?

Pulled the EFI fuse many many since times while trying to troubleshoot to see if the codes return. After I replaced the 02 sensor, like I said above, codes came back after a couple key cycles (normal driving, so it took 2 days), then the codes did not come back for a week, now they are back and occurring regularly the past 3 days.

Your symptoms don't seem too bad. What you are describing seems very close to normal.

I agree, and the idle is not horrible, and I have no other running problems. I guess it is possible that the new sensor is slow, but doubtful. I have not checked the cross-counting of this new sensor, but I can do that with my meter. I know if the sensor is not warm enough it will be slow or not cross-count at all, thus the reason for the heater in it is too heat it up quicker during the first minute(s) of running an engine....or at least that's what I believe its there for.

I was thinking that maybe the ceramic coating I had done on the turbo exhaust housing and heat shields might be keeping the exhaust stream at idle from reaching or maintaining a high enough temp at which the sensor will cross-count. But, I've not heard of anyone else having a sensor problem by just ceramic coating only the turbo exhaust housing, and I was driving daily with this setup for 4 months with no codes, so...

Every time you clear the code, fuel trim is relearned as is the idle speed based on age, conditions, etc.

Yes, I know.

 

[Nick M]

Question: Does the heater in the sensor only come on at start-up, or can the ECU heat up the sensor anytime based on other parameters?

I would have to look it up, but it should come on because response is slow compared to engine temps.

What is P0300? and can you elaborate on "together with a change in engine speed"?

P0300 is OBDII, random cylinder misfire, as opposed to one cylinder that is not keeping up.

 

[suprabee]

I would have to look it up, but it should come on because response is slow compared to engine temps.

Codes had been thrown after driving and the engine is at operating temp. the other day i drove home, 10 miles with no code, finally came to a stop light, and the CE light comes on after @15 seconds. Would the sensor have cooled enough in that amount of time to count slow enough or not at all to throw the codes...its possible i guess.

 

[Nick M]

There is a code for the heater on the oxygensor. The ECU knows if the circuit goes open or short. If you don't have the code, I don't think it is an issue. Besides, I think the sometimes rough warm idle is the cause of 25 and 26 together.

 

[3p141592654]

25/26 on together is not a mixture trim issue, it is instead when a variation in idle rpm (at the cylinder level) is detected that exceeds a threshold value when at idle in closed loop fuel control. It is an early attempt at a P0300 type fault code back when OBD2 did not yet exist. It is not clear to me, but this code may only be present in cali certified cars, maybe JetJock knows for sure.

One might expect that at a low enough idle speed with a heavy load (e.g. rear defroster on) the variation (sag) between cylinder firings might be enough to trigger the variation 'threshold".

Do you have the ability to capture the ISC commands (a logic analyzer would be the best tool for this) during load variation at idle when you say no idle compensation seems to be occurring?

 

[Nick M]

25/26 on together is not a mixture trim issue, it is instead when a variation in idle rpm (at the cylinder level) is detected that exceeds a threshold value when at idle in closed loop fuel control. It is an early attempt at a P0300 type fault code back when OBD2 did not yet exist. It is not clear to me, but this code may only be present in cali certified cars, maybe JetJock knows for sure.

I concur.

I think he should leave the idle alone for several days and see if the ECU can learn a new setting.

 

[suprabee]

Do you have the ability to capture the ISC commands (a logic analyzer would be the best tool for this) during load variation at idle when you say no idle compensation seems to be occurring?

No I do not. Also, to point out that the codes have come up without me turning on any accessories.

And about what Nick M said ^ , Yes, I thought the same, so when the codes came back on the second day after replacing the 02 sensor, I did not clear them. the CE light eventually went off and and stayed off for the next 10-11 days of daily driving. I was starting to think that it the ECU had finally re-learned its values, but then the light came on this past Tuesday morning, then again on Wednesday. Yesterday, was when I started turning on the accessories and got it to come up twice repeatedly, by putting a load on the engine.

Additionally, I did some checking of my records and it looks like the oem plug wires with a date on them of 2000 that have been on the car when I bought it in 2005 could have as much as 90k miles on them! Carfax on my car lists the mileage in Jan 2000 at 47K, now it has 140K. The wires are still in good shape, ie, not cracked and measure the correct resistance which does not change when flexed, However, I've decided to buy some new wires, simply its time.

So, I've got on order a new set of oe wires, probably will arrive next week....couldn't hurt to have new wires on the thing....maybe this is part of the problem. I had no idea there was that many miles on them, and never thought until this afternoon to look at my saved Carfax report for the miles back in 2000. Glad I did.

rick

 

[3p141592654]

I would also do a compression check or better yet a cylinder leak down test to make sure there is nothing mechanical going on that could result in an imbalance in cylinder combustion efficiency. The plugs are reasonably fresh I presume? Also, the injectors need to be in good shape with uniform flow. For whatever reason, your idle is shaky, so follow that line of reasoning.

There is supposedly another cause of 25/26, which is a cross count frequency that is too high, but I'm pretty sure that is not your issue.

 

[suprabee]

I would also do a compression check or better yet a cylinder leak down test to make sure there is nothing mechanical going on that could result in an imbalance in cylinder combustion efficiency. The plugs are reasonably fresh I presume? Also, the injectors need to be in good shape with uniform flow. For whatever reason, your idle is shaky, so follow that line of reasoning.

plugs are new since March, and I've had them out a couple times since them to double check the gap and look at their condition which looked normal. I will remove them again and take another look when my new wires arrive. Maybe at that time, I'll go buy the stuff to do a compression check, as I haven't done it yet.

However, I have no other running issues, pulls hard, boosts, and also to point out that the shaky idle is only 50% of the time when it does happen, other times it idles fine, AND recall that for one week the car ran great with no code or shaky idle. I would expect an internal cylinder problem would be more consistent.

rick

 

[3p141592654]

It seems you are right on the threshold of what the ECU thinks is out of spec. I wouldn't rule out a lazy injector either here, but at some point you need to make a cost-benefit decision about whether this really is worth the trouble to diagnose further.

One thing I do notice about my car is that the idle speed adjustment is slow, so that if I jump on the freeway before the engine is warm when I get off the idle is way high, and I can watch the ECU slowly reduce it to the target speed over about 20 seconds while idling. I think that is why throwing the defrost load on it cause the idle to sag, because the ECU really doesn't attempt to adjust the idle speed quickly to compensate for a sudden load. Over 30 seconds or so you should see it come back to target.

 

[suprabee]

I wouldn't rule out a lazy injector either here, but at some point you need to make a cost-benefit decision about whether this really is worth the trouble to diagnose further.

In January, when I had the intake off to clean, and exhaust off replaced the mani gasket and rebuild the turbo, I contemplated sending out the injectors for servicing, but opted not to to save money, since I was spending so much on the turbo. In hindsight, I wish I would have, to remove that possibility. I'll recheck the plugs when I put on new wires, and do a compression check hopefully this weekend. If that does not turn up something or resolve it, I will then pull the injectors and service them.

Rick

 

[3p141592654]

Servicing the injectors is a pain. One thing you can do is just listen to the exhaust. The 7M idle is very uniform sounding if all is well. If you hear a rhythmic sound then that would indicate a cylinder or two is out of spec.

 

[Nick M]

You can try an at home injector service. You remove the fuel line and try to pump in a lot of injector cleaner. That is how an on car injection service works. The return is blocked and the supply is cut off. Then the fuel rail is fed with injector cleaner.

 

[suprabee]

tested compression today,
two tests/cylinder
this was a cold test
let the battery charger charge battery between each cylinder test.

#1 145, 145
#2 145, 145
#3 140, 140
#4 70, 70
#4 80, 80 <=== higher with oil, but is this enough to call it the rings?...
#5 135, 140
#6 120, 120

Although #6 is below spec, this was a cold test, of course #3 looks a problem.
I've read some info saying cold tests are not valid, that numbers can be all over the place and misleading.

I guess I should warm it up and test at least the #3 again
Rick

 

[suprabee]

Just warmed engine and tested #4, #6, and #1

#1 150
#4 80
#6 150

So, it does look like there is something to testing warm, which makes sense to me.
#3 still same though...damn...guess I know finally know where the problem lies.

My next step is have a friend come over with compressor to fill up the cylinder with air and see if we can locate where the leak is. Hopefully in the valves. IF so, I'm thinking the next steps in easiest to hardest would be:

1. check the shims on any suspect leaking valves
2. pull cams and check valve spring(s), feel if valve(s) are sticking in their guide


If that does not turn up anything, then I guess I'm left with pulling the head, yes?

suggestions please.

rick

 

[3p141592654]

Sorry to hear this, but good validation that the cylinder balance software works, even from 20 years ago!

Most likely scenario with the 7M is the head gasket has gone between cylinders, hence the low reading on both. I think you are looking at a rebuild. If is has never been done, then also a good time to refresh the water pump, oil seals, tensioner pulley, T belt etc.

 

[suprabee]

Most likely scenario with the 7M is the head gasket has gone between cylinders, hence the low reading on both.

No, I think you misread my data, or rather the way I listed it is misleading and I had a typo which is now corrected, only one cylinder is low #4, maybe this is better way to show the numbers:

COLD TEST
#1 145, 145
#2 145, 145
#3 140, 140
#4 70, 70 ==> with oil added 80, 80
#5 135, 140
#6 120, 120

WARM TEST:

#1 150
#4 80
#6 150


and as a reminder ...there is no mixing of oil and coolant, no combustion gasses in coolant either, no smoke out exhaust .

 

[3p141592654]

My bad, I thought at one point you showed #3 and #4 low. Any gurgling noises on a cold start? My HG took seven years from the time it started gurgling until it blew completely.

Usually a stuck valve would make a big ticking noise, wouldn't be hard to miss. Leak down tester should help isolate the problem.