Estima/Previa 4WD feels slow in overdrive

[eastman51]

Hello all,

I acquired a 1994 Estima S/C 4WD last year, and its a blast to have two drastically different vans to compare and of course drive/work on. This is a rhd van, but for all other intents and purposes it is pretty close to identical to an Alltrac S/C sold in America.

Oddly enough, this van has given me way more problems than my 2wd with 3 times the mileage. However, there are a few larger issues that are plaguing me and being difficult.

The current problem on the chopping block is something I find very odd. I noticed a few days ago that when I tip into the throttle on the highway to avoid losing speed going uphill, that the van seems to fall flat on its face. I can feel it slow down or lose power, and feel subsequent jolts through the seat and floor; almost like it is misfiring or something similar. The van only has 52k kms on it (roughly 31k miles), but due to what I know about the van's history I elected to change the spark plugs, wires, etc. The van is pretty much all original, besides a few obvious things. From what I was told by the previous owner, it was primarily used for short trips back in Japan; followed by use as a mail delivery truck by the previous owner himself. Lots of low speed, short trips, and tons of idle time; nothing like the balanced or primarily highway mileage I'm used to in used cars.
I found that at one point, someone had replaced the plugs with Denso Iridium (likely when the van was imported to the US), which were visually worn out a bit. The rest of the ignition was 100% original.

I swapped the plugs out for NGK Platinum, and replaced the distributor cap/rotor and wires with NGK as well. Same stuff I put in my 2wd Previa that has 230k miles.
This does not seem to have fixed the problem, which I was kind of expecting; but I am glad I know the ignition is sorted anyhow as it seems to run a bit stronger and idle smoother. I did not see any signs of the valve cover or distributor o-ring leak, so its good there.

I did verify that this issue seems to occur only in 4th gear. It is not limited to hills, nor is it limited to speed; the two key factors are adding throttle and being in overdrive.
My two leading theories are that the van is 4WD, perhaps the engine doesn't particularly like being loaded up in overdrive with this drivetrain. My 2WD van has absolutely no qualms with cruising on the highway in 4th, especially when using cruise control; this is my first all-wheel drive vehicle, especially being full-time, so I could just have unrealistic expectations of what is reasonable for the drivetrain. My other theory is that there could be some kind of issue with the supercharger system; perhaps when I place this load on the engine, it tries to engage the s/c clutch but the intake stepper motor does not open?
It is not a primary suspect of mine currently, but I would not be surprised if this fuel related.

I will be digging into my US FSM for some supercharger related things to test, but figured I'd post here for other suggestions and thoughts.

[Jan-Willem]

The previa 4wd is center locked right? Are you sure front and rear tires are the exact same size? And neither of the diffs/axles could have been replaced with another? (Jdm’s may have a different ratio to us models)

[eastman51]

I don't believe there is any sort of locking on the Previa 4WD, its just got some sort of transfer case/center diff and runs as full time 4WD. Front and rear tires are the same size, I just recently swapped the steelies from my 2wd van because the alloys that were on the Estima had mismatched tires (still all the same size); the wheel specs are also identical. It acted this way on both sets of wheels/tires though.

I've never seen a US model Alltrac, do they have live axles like the 2wd? My JDM has IRS, but I don't know if that is just 4WD exclusive; my general understanding is that JDM (and possibly AUDM) had IRS regardless of trim, but I could be wrong.
I don't really have any reason to suspect that a diff and/or axle was swapped at some point, so I believe they are original (aka correct).

[John Kaufmann]

... I've never seen a US model Alltrac, do they have live axles like the 2wd? ...
Minor point, so that the thread does not wander off: I believe all US models (which would mean Previas, not Estimas), both AWD and RWD, have solid rear axles.

[eastman51]

Minor point, so that the thread does not wander off: I believe all US models (which would mean Previas, not Estimas), both AWD and RWD, have solid rear axles.

Ok, thank you for confirming.

This does mean that the rear end on the Estima is most likely good. Unless someone swapped the axles from side to side (if that even matters?).

[eastman51]

Well, I have a slight update.

I checked the FSM, and it seems like to check the supercharger bypass valve you need to remove it (and possibly have some kind of bench power supply). So most likely the only things I could realistically check myself are the resistance of the clutch stator and the supercharger boost pressure.

I also drove the van today, and noticed something interesting. On the way to and from work I did not notice the weird feeling, nor any jolts. However, when I drove to lunch I did. The difference is the highest speed limit on my commute is 55mph, while on the way to lunch I was able to go 70mph. So there DOES seem to be some relation to speed, not just being in overdrive. I also felt like it gained power back as I tipped back off the throttle after driving a distance with it feeling odd.
On my way home from work I tried to see if it was possibly related to the speed at which I tipped into the throttle, but even around 95kmh going uphill I didn't feel the power loss thing and it just shifted down to 3rd instead. So I'm thinking it is more likely to be fuel related than air related. I doubt the bypass valve fails very often anyway.

[tbuyan]

Your van is right hand drive, so it's not Canadian. What was its original destination country?

Probably unrelated to your symptoms, but here's an FYI: US spec SCs will not shift into 4th gear (or OD) until the engine is warmed up a bit (temp needle 1/4" or so above C) and speed is at least 50 mph.

[eastman51]

It is an Estima, original destination was Japan as I believe I have mentioned earlier.

I've noticed the shift speed for O/D is 45mph in my 2wd US spec, and the 4wd JP spec seems about the same.
If anything, the only difference in transmission behavior is that the JP model seems more willing to use 3rd gear before the temp needle reaches at least 1/4"

[cb97alltrac]

Check the timing, esp with acceleration. Some of the sc engines were overly sensitive to knock sensor signal, which would retard the timing with accel, would almost act like a bad SC or sc relay. On the old yahoo group they would overcome the knock sensor oversensitivity by placing a small resister at the ks signal at the ecu.

[eastman51]

How would I check the timing? With the engine placement I'm not sure where I'd use a timing light. With the JDM ECU, I don't think my code reader works.

[cb97alltrac]

Can use a scangage ii. There is a timing setting that changes realtime with
driving

[John Kaufmann]

If not a scangage, just do it the old-fashioned way (except of course that you're looking at the timing pulley from below); clip the sync line onto the #1 plug cable. [Before starting, jumper Check Connector terminals TE1-E1 to disable spark advance.]

[eastman51]

Yeah, I can't use a scangage at all due to a lack of OBD2. Its a 94, after all. I'm fairly certain the scan tool I have, while it can read Toyota OBD1 in my Corolla and US Previa; when I tried it on the Estima it didn't work. Something with the Japanese ECU I'd assume, unless I'm doing something wrong. But to be honest, the tool is pretty basic and I'm not sure it reads ignition timing anyway.

I'll have to make time to grab a timing light and check it sometime, just not sure when since the garage is currently full with a broken Corolla...

[eastman51]

I am going to preface this by saying I haven't checked the timing yet, nor have I tried the code reader again.

That being said, I believe that the sensitive knock sensor issue may be on the right track.
I drove the van on Sunday down to the next town to meet up with family. Nice sunny day, around 60F ambient temp, even used AC for the first time all year.
I noticed the issue a little on the drive there, but I can't say with certainty due to running the AC and there being some substantial gusts of wind. On the drive back I took the interstate rather than backroad highway, and don't believe I noticed it at all. Interesting.

This morning I decided to drive the van again since I'll be driving my daily the rest of the week and through the weekend.
It's 34F ambient, not much wind, and had to defrost the van before leaving. Didn't have any issues for the first half of the drive, and then the power loss and pulse in the seat was the worst I think I've ever experienced it (other than maybe the first time I noticed it).
Coming into the town that's halfway on my commute, I noticed this problem even at around 85-90kmh. Interesting, because before I had thought it was entirely speed dependant.
Going through town wasnt a problem, of course; mostly coasting anyway.
Back out on the two lane highway, I noticed it almost immediately once I was at speed. Luckily there wasn't any traffic so I was able to experiment a little, and found that sometimes if I held it, it would kind of clear up a little and accelerate close to normally but not quite fully there. Other times it would just bog really badly. Still no check engine light of any variety.
And finally, on the second to last turn, it did something completely different to the typical symptoms thus far.
For this turn, you slow down from posted 50mph limit to make a 90 degree right turn onto a 55mph posted limit road (though I rarely get up to that speed before turning into work); I made the turn (around 25-30mph usually) and gave it a little gas (probably around 55kmh, 2nd or 3rd gear not certain) and noticed it had no power, then downshifted and accelerated normally before instantly hitting me with it after shifting into 4th at 78kmh.

I doubt the mechanical timing is off with how inconsistent this seems to be. It's almost got to be the knock sensor. When I'm giving it ample throttle and it's in lower gear, it has great power all through the higher revs. It's really just at lower throttle inputs when cruising or accelerating gently. And now that I really start to think about it, I've noticed it even accelerating from a stop; I've always felt that the Estima felt pretty slow off the line if I didn't give it a lot of throttle. But I always assumed it was just the 4wd making it seem slower than my 95 2wd.

I'd rather not modify the wiring to add a resistor that shouldn't be there to begin with, does anyone know if I could simply replace the knock sensor with a new one to perhaps resolve this?

[eastman51]

I definitely suspect the knock sensor for certain now. But I don't really know how to prove it.

I found out by doing a bit of light research that basically all JDM cars, of this vintage, run off of some form of OBD1 protocol. Most people seemed to think that typical US OBD1 code readers would work, but I could not find any hard confirmation or proof which was slightly annoying. I gave it a shot anyways, and it does seem to work to a certain extent but I'm not sure how accurate any of the live data is. Previa did not show up as an option outside of OBD2 on my reader, so I had to botch it with a different car (told the reader it was a 4Runner 4WD, and also tried it as a 91 Celica with nearly identical data readings).

I went for a drive this afternoon with the code reader, but I feel like I'm left with more questions than answers; though admittedly I have no clue what I should be looking at in terms of ignition advance.

At warm idle, the "Spark Adv Ref (*)" PID fluctuates between 42 and -18. Idle is around 800-ish according to the tach, and is butter smooth with no missing or vibration.
While driving around, even maintaining a constant speed on relatively flat road, this PID fluctuates constantly. I didn't take any pictures or videos due to safety reasons, but it would change from anywhere between 88 and -23.
For a while it didn't want to do its little power loss thing, even after a good 10-15 minute cruise at 55-60mph. I took it off of country roads and onto the highway where it finally decided to act up. It seemed like the reader was displaying either 51 or 57 when it was acting up. When I went to turn around, I noticed it had basically no power even at 3/4 throttle getting back up to highway speed. The reader continued to read around that 51 and 57 mark but occasionally fluctuating to seemingly random numbers. It continued to be low on power all the way until I got back into town and sat at a stop light; from there on it felt fine with the city driving back to the house. Temp gauge was fine the whole time, and using my IR thermometer under the hood had the radiator and coolant reservoir at 180F.

12075
12074

All I can really make sense of is that mechanical timing is perfectly fine, I don't think a timing light is necessary.
The only other thing I can think of, is that perhaps the knock sensor is perfectly fine. I read during my research that Japan has higher octane fuel at the pump and many people were saying that use of octane booster is required when running a JDM ECM as they aren't tuned for use with US pump gas. I've heard this in passing from other people at car meets and such as well. I didn't really come across any information that seemed to really prove this point, but with how often I've heard it I'm wondering if there's some truth to it. A few forums claimed that pump gas in Japan is roughly equivalent to 94 or 95 octane on the US scale, where the highest octane fuel at the pump (that I have seen) is 93 octane. However, I will say that this van behaved virtually the same on both 89 and 93 octane fuel; so I'm left in pretty much the same place as before.

[tbkilb01]

Any codes?
Does the throttle position PID
change when You step on gas pedal
how about under load?
Do you hear any pinging/knocking Under load?
Can you check any other engine performance data PIDS
using the ACTRON
MAF looks good...
have you looked at it under the same load
as when it’s acting up
:)>:

[eastman51]

I don't remember if it listed a TPS PID. I will bring the Actron next time I go for a drive and check. It does list rpm, but doesn't seem to be very accurate; usually reads higher than actual.

I definitely don't audibly hear anything when it acts up. It sounds like normal, from what little I can hear (even with window down), she's pretty quiet with the factory exhaust. I just feel jolts in the bottom of the seat and feel like it has no power + watch the speedometer drop as it slows down (until I give it enough throttle to kick down, where it will accelerate fine).
I'd have to pull up the scanner again, but there really wasn't a whole lot of PIDs that it let me look at; some of them also didn't seem to actually report anything either, such as the start command and idle command.

While driving and it acted up, I only paid attention to the IGN Advance where it would stay around 51 or 57 deg the entire time I watched it until it either kicked down or I released throttle. I will try to make a video next time so you can get a better idea of what its doing.

I didn't specifically check for codes, but I know that crossing the pins results in an all clear at the MIL lamp in the dash and the MIL is not illuminated while the engine is running. It doesn't flash or anything when it acts up either. Whatever the cause of this is, it seems the ECM is detecting no issues.
idk if it'd be worth the effort, but I've got two supercharged vans. I could try swapping the ECMs and see if the 2wd Previa acts up and/or if the Estima behaves as expected. Worth a try? idk if the drivetrain difference would matter in swapping the ECMs?

[eastman51]

I am armed with more information.

No there is not a TPS PID or any other throttle PIDs. I recorded a video on a short highway stint, hovering around 90-110kph depending on hills and how it was acting up. I have the video uploaded as unlisted on YouTube. Here is the link: https://youtu.be/o8ryFWAI5T0

It does its thing a few times during the video (and I used a custom data set to cut out the useless PIDs that did nothing and/or weren't relevant), but I didn't think to indicate when it was occurring until near the end. I honked a couple of times at the end to indicate when it was happening, first occurrence of this is at 4:26. If you want to examine the video in full, be my guest.
The tldw is essentially that IGN Advance locks to 51.

However, I also have another bit of info. I used the code scan function of the Actron and it returned a Code 21 for the front O2 sensor heater circuit. I do not remember getting this code when I initially jumped pins in the diag port last year; admittedly I have not jumped the pins since this issue first arose as the MIL has not illuminated while the engine is running. Today, however, I stopped at the gas station on the way to the highway to get some octane booster (which they did not have, somehow) and when I came back out and started the van the MIL was on for a good few seconds before turning off. I don't think this code would relate to this issue since the heater circuit should only apply when the sensor is cold, not while the van is running at operating temperature. In either case, the O2 sensor is a potential suspect due to the code...

Still don't hear any indication of ping or knock. I am tempted to try octane booster, but I am also reluctant to spend the money on something that may end up being a waste.

[tbkilb01]

Looking at data
and it seems to be
A bit too glitchy fir me...could be all false data...
That PID ..TP... varies allot and settles to 0
It was at 119 when I heard the honk
then it went as low as 14
while RPM varied... Although a little or not at all...
but hard to tell if actron working on estima
plus all the rest of data is changing allot...As well
Missing Knock Sensor signal ?
pull codes?
Did you check your coil while you were by the distributor
Have you turned off overdrive
and gone thru same load process...
did it act up the same

[eastman51]

Coil is most likely original. Visibly it looks perfectly fine. I've tried with overdrive off and it seems to be much more random. It usually doesn't act up in lower gears unless its acting up immediately before turning around or turning off to another road. Sometimes it feels like it still lacks power when it downshifts when I add more throttle when it acts up, but hard to tell for certain as it is willing to accelerate more. Its a really weird problem, very frustrating.

It pulls a Code 21 as mentioned previously, but I don't know if its entirely related to this issue or not.
Yeah, the data fluctuates a lot. The rpm and vss pids are obviously wrong 99% of the time. Unfortunately Previa isn't an option for OBD1 on my Actron, and I'm using it on a JDM ECM too which doesn't help. Its better than nothing, but not exactly helpful.

[eastman51]

Unfortunately, the issue is still not resolved.

I replaced the O2 sensor in early June with a brand new Denso unit. The ghost Code 21 was gone.
But it has not resolved the issue, as I feared.

Since replacing the O2 sensor, I have noticed that above 70mph it behaves mostly normal. Sometimes it's slightly sluggish, but nothing like before. In the 50-65 mph range is where it's bad. Tip into the throttle and you feel it begin to accelerate before it hits a wall, zero power even when it kicks down and I push the pedal all the way to the floor. If I keep it floored or turn off overdrive to keep it in 3rd, it'll eventually "clear up" as rpm rises and I feel it misfire/buck before becoming smooth; then if I back off and/or turn overdrive back on it will behave as expected for a short time.

Where the problem gets really bad though, is when you've been cruising 50-65mph, it's acting up, and now you need to slow down (make a turn, slow traffic, traffic is turning, etc). When you need to accelerate again, it has exceptionally low power and almost refuses to accelerate past 15-20mph until it "clears up" around 3000rpm. This can be very sketchy and dangerous if making a left turn without a traffic light giving right of way, or generally just inconveniences other traffic being stuck behind a slow vehicle.

I just recently returned from a 1200 mile road trip with the van. It did great other than in cities where speed limits on the highway are 55mph and on country roads where the speed limit is 55mph. On open highways where speed limits are 65-70 it experienced very little trouble.
No overheating, no other noticeable issues. I haven't checked codes again, but I suspect I would find none. Definitely no flashing CEL or anything. No abnormal noises from engine, transmission, differentials, and oil level did not change at all through the trip.
I saw decent fuel economy, averaging around 20.8mpg. One leg I saw 20.98 and on the return route of that exact same leg I only saw 19.7, which was odd. This is a stark change in fuel economy from before replacing the O2 sensor, but I suspect it would be higher without the power loss thing.


I'm at a complete loss. Unless some new ghost code has appeared, there's nothing to base any theories off of. I've tried fuel system cleaner, to little or no effect. I wouldn't know what a clogged car feels like, but even then I doubt the cat is clogged at only 55k kms; even if it was slightly clogged I doubt it would still be after ~800+ miles at 70-75mph. It has new spark plugs, wires, dizzy cap and rotor, and O2 sensor.
I would assume that if it was the knock sensor, I'd have the Code 52 for it, even if it's just being too sensitive.
I have also thought it could be a clogged fuel filter, but then why is it so inconsistent and is fine when only driving low speed or only driving high speed?

What else could cause this? Should I throw a coil at it? I just want a working van for once, and now my US Previa is nearly undrivable too...

[eastman51]

I have finally solved this problem.

Earlier this year I decided to do some service to the van and changed the transmission fluid and filter. I initially didn't notice any change, but during a road trip a few days later the issue suddenly worsened dramatically; turned out the transmission fluid was low after stopping. I am confident that the fluid was not low before doing anything, however after topping off the transmission again and letting it cylce each gear multiple times and topping off as needed, the problem seems to have disappeared. My only theory is that the fluid had become contaminated from age, or the system had otherwise become less effective. I haven't driven the van too much since the road trip, but has been acceptable since. If it does it again, I fear that there may be an issue with the transmission itself.

[Previologist]

If it walks like a duck and quacks like a duck, its probably a duck. Or in this case, it was probably the fluid. The tranny fluid was either low, or no good, or a little of both. I take your word that it wasn't low, so maybe it was no good, or even the wrong kind. These trannies are pretty durable so I wouldn't worry unless it starts acting up. I got 8-9 years out a slipping Previa tranny by adding Lucas tranny stuff and it was still going strong (and not slipping) when the engine blew a HG and I surrendered it to the rust.

[eastman51]

The last time I had checked it I am certain it was full, but I don't really remember when exactly that was (prior to any work, of course); I do specifically remember the fluid color was not great, however. Most of my cars don't have a good way to check trans fluid levels, so I get really bad with checking it in the cars that do. For all I know it was low, but if it misbehaves again while full on fluid, then its definitely something else. I didn't do an actual flush, so there could still be some junk fluid hanging around from the converter or cooler; time will tell. But at least for now I'm glad I know what the issue is so I can deal with it more effectively in the future.