Auto trans won't shift

[96 XJ Sport]

Well, here's what the testing yielded:

-Battery voltage at C8, C10, D14, and D16 was 12V give or take 0.2 (with key and shifter as indicated for each pin).

-Battery voltage at C9 was 0.00 (with key on and selector in D as indicated, but should be 12V).

-Voltage at D2 (TPS input) was 0.7V with closed throttle and 3.8V with gas pedal to the floor. Test indicates that it should be the opposite?

-TPS ground at D3 was 0.00V.

-Lockup solenoid at C14 read 12.2 ohms and 14.5 ohms on subsequent readings before and after driving.

-Solenoid #2 at C15 read 32.8 ohms and 35.5 ohms before and after driving.

-Solenoid #1 at C16 read 513 ohms and 340 ohms before and after driving.

So the correct diagnosis would appear to be to replace solenoids #1 and #2. Is that correct? What about C9 reading 0 volts in the D position? I don't know what to do with that one.

 

[lawsoncl]

So the correct diagnosis would appear to be to replace solenoids #1 and #2. Is that correct? What about C9 reading 0 volts in the D position? I don't know what to do with that one.

I would agree on the bad solenoids. There is a possibility it's wiring damage between where you are measuring at the trans computer connector and the solenoids themselves. The wiring runs through the firewall in front of the trans computer, down the firewall, over the tranny, and into the side near the bottom on the drivers side. Best way to eliminate the wiring is to drop the pan and measure directly on the solenoids. If this isn't a daily driver, I'd probably pull the pan and order solenoids if they confirmed as bad (I'd say they probably are).

There are some writeups around here on replacing the solenoids. Pretty easy, just a bit messy with pulling the pan and filter. You could use junkyard replacements if they measure good, or buy new http://www.naxja.org/forum/showthread.php?t=996259.

The C9 voltage means you also need to clean the Neutral Safety Sensor. That's a pretty common problem which might explain some of the erratic lockup behavior. You might also notice the reverse lights don't always work. http://www.jeepin.com/features/nss/index.asp or http://www.go.jeep-xj.info/HowtoNSSrebuild.htm.

Congrats on learning firsthand just how helpful a meter can be!

 

[96 XJ Sport]

I would agree on the bad solenoids. If this isn't a daily driver, I'd probably pull the pan and order solenoids if they confirmed as bad (I'd say they probably are).

The C9 voltage means you also need to clean the Neutral Safety Sensor.

Congrats on learning firsthand just how helpful a meter can be!

It's not a DD, so I'll pull the pan and confirm 100% before ordering solenoids. I think I'll just get the whole set of 3 while I'm into it. You know I just cleaned that NSS about a year ago, but I guess it could be dirty again. Something that I didn't think of last night is that even though I'm getting reverse lights, it could still be slightly out of adjustment. I'm going to test C9 again with the lever in 3 and maybe halfway between D and 3 and see if I get any voltage. Hopefully that's all that is.

Thank you again, sir! I'm going to be running around the house now testing voltage on everything :idea:

 

[kastein]

Yikes! Yeah, those solenoids are WAY out of spec (spec is under 20 ohms, iirc.)

Make sure you have the probes in the correct sockets on the meter - measuring current you may need to use different sockets, and measuring voltage while the meter is set up for current accidentally (or volts while set for ohms) will result in some fireworks and pretty blue smoke. Ask me how I know, and why I have been very careful about setting the meter correctly for ten years now...

 

[96 XJ Sport]

Yikes! Yeah, those solenoids are WAY out of spec (spec is under 20 ohms, iirc.)

Make sure you have the probes in the correct sockets on the meter - measuring current you may need to use different sockets, and measuring voltage while the meter is set up for current accidentally (or volts while set for ohms) will result in some fireworks and pretty blue smoke. Ask me how I know, and why I have been very careful about setting the meter correctly for ten years now...

Yeah, the one that read over 500 ohms shocked me, too. I still don't know what it means exactly, except that it's WAY off. I triple checked every reading and that the meter was set as appropriate to either DC volts or Ohms, but that is certainly a fair question since it was my first time using a meter. I followed that chart in the link provided by lawsoncl to a "T".

 

[kastein]

The solenoids need around an amp or so to make a strong enough magnetic field to pull in the valve actuators. If they have that much resistance it probably means there is a broken wire and you're reading a leakage path or intermittent, dirty connection. At that point it might as well not even be connected.

I'd unplug and clean the connector for the transmission wiring harness by the dipstick tube. Just unplug it and spray it out with contact cleaner, then pack it with dielectric grease and plug it back in. If that fixes it, you found the problem. If it doesn't fix it, you need to measure the resistance of the wire all the way from the TCU connector under the dash to the same connector at the dipstick, or preferably to the solenoid connectors in the pan with the solenoids unplugged. Electrical stuff is kinda annoying on cars because there really is no debugging method other than "divide and conquer" one step at a time.

EDIT: not sure if you've ever dropped the pan on an AW-4, but it's easiest getting it off of the dipstick tube if you spray a little PB-blaster (or your favorite penetrating lube) into the junction between the tube and the tube connected to the pan, then heat it for a few seconds with hot air or VERY carefully with a torch (keep an extinguisher in the other hand... can't really recommend this one, I only did it because the tranny was sitting on blocks in my living room and I had already swabbed almost all of the gunk and grease off of it) and then twist and pull.

 

[96 XJ Sport]

Ok, I'll check/clean the connector by the dipstick tube before dropping the pan. Dumb question, what is the expectation for resistance when both ends are unplugged?

I have dropped the pan before to do a filter change and I swore I would never do it again because of that stupid dipstick, lol. Hopefully it's loosened up this time.

 

[kastein]

When both ends are unplugged, I would expect no more than ~5 ohms resistance. Hopefully no more than one to two ohms. Make sure you zero out the meter using the adjust knob (if analog) or subtract off the probes-touching measurement (if DMM) to get an accurate reading.

EDIT: hahaha, yeah, I had issues with the dipstick tube too. I had the darn thing sitting on blocks in my living room and twisted it around forever without getting it to come out, so I tried the heat and penetrating lube out of frustration. Came out 2 minutes later. Unbolting the hold-down strap (goes to one of the bellhousing dust shield bolts) will help you a LOT, in fact if you do that and remove the zip-ties holding the wiring harness to it, you may simply be able to snake the whole assembly out the bottom.

 

[96 XJ Sport]

So I need to do the probes-touching measurement every time I measure resistance on my DMM? I didn't do that last night...

 

[kastein]

Nope, only when measuring low-resistance stuff. It will probably be ~1-2 ohms, on the 2k scales and up that digit isn't even shown, I ignore it if the value read is over about 50 ohms. Since solenoids and injectors and stuff like that are all under 20 ohms, it becomes a little more important.

 

[96 XJ Sport]

Got it. Thanks. Out to the garage after the kids are in bed...

 

[96 XJ Sport]

Maybe I misunderstood...I'm not getting any reading with both ends unplugged. On the 200 scale, it just stayed at 1 like before the reading, so out of range. So up to the 2k scale, same thing, and so on all the way up. Plug it back into the trans under the hood and I get similar readings as before. Under hood plug was clean as a whistle inside, but cleaned it with some electrical cleaner anyway. Gonna drop the pan and get a reading directly on the solenoids...

 

[kastein]

Oh. You read from one end of the wire to the other end of the same wire - you should measure infinite resistance (like you did) from the frame/ground to each wire. Sorry I caused confusion...

 

[96 XJ Sport]

Well, had to torch that tube this time and made a huge freakin mess, haha. Here's what I got on the solenoids working from the back to the front of the trans. 1st one 12 ohms, 2nd one 225 ohms, 3rd one 12 ohms.

Seems like a victory in that one is confirmed to be crazy bad, but why did one of them that read in the 30s from the TCU plug now read 12 ohms when tested directly?

Edit: Nevermind, I changed the grounding location I was using at the TCU plug and now I'm getting readings very close to what I got directly on the solenoids. Even though only one tested bad, I think I'm still going to replace all three because my problems were so intermittent and I am NOT going to take this apart again. Plan to order them in the morning unless anyone thinks that's a mistake.

 

[joe_peters]

Well, had to torch that tube this time and made a huge freakin mess, haha. Here's what I got on the solenoids working from the back to the front of the trans. 1st one 12 ohms, 2nd one 225 ohms, 3rd one 12 ohms.

Seems like a victory in that one is confirmed to be crazy bad, but why did one of them that read in the 30s from the TCU plug now read 12 ohms when tested directly?

1. A bad reading--it happens.

2. Problem in the harness between the TCU and that particular solenoid. You can go back and test the harness resistance from the TCU connector down to the connector for the solenoids. If the harness resistance checks Ok, chalk it up to a bad reading.

 

[96 XJ Sport]

1. A bad reading--it happens.

2. Problem in the harness between the TCU and that particular solenoid. You can go back and test the harness resistance from the TCU connector down to the connector for the solenoids. If the harness resistance checks Ok, chalk it up to a bad reading.

Yep, a bad reading. I think you were typing when I edited my post. Picked a better grounding point and now TCU plug and direct solenoid readings are consistent. Going to order solenoids in a couple of hours unless somebody stops me :clap:

 

[kastein]

Sounds like a plan to me. Where are you ordering from? According to my info, the part number for a solenoid is 4797 402 or 4797 403 (depending on which one it is, there are two 4797 402s and 1 4797 403) and they cost $150 each even from mopardiscountparts :shocked: RockAuto wants around 200 each for them.

At that price, I'd go to the junkyard and pull two entire valve bodies for 30-40 each, measure the solenoids, and take my chances. If you use a gasket instead of RTV (NAPA/rockauto have the right kit, ATP B112 is for the AW-4, comes with filter and gasket - NAPA's computer will give you the kit for the 3spd auto tranny so ask for that part number specifically) it is a hell of a lot easier to drop the pan again.

EDIT: those solenoid part numbers are for 95-97 and maybe later. 94 and earlier use a different set of part numbers though the later ones most likely swap in fine.

 

[96 XJ Sport]

I just ordered them from here for $152 for the whole set of 3. http://www.transmissionpartsusa.com/AW4_A340_transmission_solenoids_kit_Toyota_p/940-00094393a.htm lawsoncl had posted this place in a sticky somewhere that I found.

Yeah, I used a rubber gasket the last time I changed the filter, so this time the pan just fell into my hands...will def use a rubber gasket again. And I don't have to clean all that RTV off the pan and valve body again, YAY!

I should have the solenoids by Friday and installed Friday night.

 

[96 XJ Sport]

Can anyone identify which solenoid is which? And are any or all of the solenoids interchangable? Maybe this will be obvious when I get the new ones in my hands later today, but I want to be ready to knock this out tonight...need the heep for some towing tomorrow

 

[kastein]

The one at the front of the valve body (on the right in this picture) and the next nearest one are the same part number according to the catalog, and have the same style metal bracket in that picture, so I believe they are the ones that are the same. I'm not sure which is which though - you can probably trace the wiring or ohm it out and figure out which is which.