XJ won't downshift.

Two problems.

I have a 90 XJ.

1) This July we were running through a river and some mud pits, shortly thereafter I realized that my XJ wasn't downshifting as normal. It seems to shift shortly and when I floor it I get no response except wasted gas. After a couple of weeks it seemed to clear up for a while. I took it to the car wash to get some of the grime out of the engine bay and the same problem occurred. I left for Texas and left it sitting there. When I came back it worked for about 15 minutes then started acting up again and now it won't down shift at all. I kind of narrowed it down to two factors - water or cold. It seems like everytime that is works, the engine temperature is around 160 degrees. I doubt that this is the reason, because there are alot of times when it isn't working and the engine temp is 160. The second factor is water, it started when I immersed my XJ in water, and when I washed out the engine bay.

2) I can no longer get my shifter into 1-2. I have no idea why, it just happened one day.

Thanks

- Seth

Maybe its low on fluid.

Fluid levels are fine.

It's likely there is some moisture in the TPS and/or the connector and/or the adjustment and/or an iffy ground after the TCU.
My upshift points went crazy shortly after cleaning my TB and accidently sparying some Carb cleaner on the TPS. I've heard many people warn about getting the TPS wet during engine washes and such.
The previous owner of my 88 XJ sprayed the TPS connectors inside and out with WD40, which really messed with the resistance and adjustment. The TPS has a very low working voltage, the outputs are resistance/voltage changes to the TCU and the harness goes through numerous connectors berfore finding it's way to the TCU (and ground at the dipstick holder). A lot of opprotunities for faulty outputs.
But my TPS problems seemed to affect upshifts more than downshifts.

IMO there is some interaction between the hydrolic shifting at the valve body of the AW4 and the electronic shifts controlled by the speed sensor, the TCU and the TPS. Though the tendancy is for the electornic shifting to be dominant. I don't think I've ever done a speed sensor test, but it seems logical this would have some influence on downshifts, along with the TPS and possibly the TV cable (full throttle downshifts and upshifts). Speed sensor would be low on my list of possible causes though.
A TV cable adjustment seemed to affect dowshifts (on my 88), mostly how much pedal was required (how near full pedal) to get a downshift into passing gear. I've heard this debated, but I've changed the downshift point, by trial and error on mine before, out of curiosity (one notch up and down on the TV cable adjustment). It does seem to have some influence. The TV cable also has some input to the pump pressures, so the recommended setting is likely the best and safest. It's hard to find a good understandable explanation for adjusting the TV cable ( a search may turn up a good expanation), a simple task that seems to cause large problems in explanation. The Chev truck manual (post 90) TV adjsutment for the 700R4 has about the best explantion I've seen, identical process for the Jeep.
I'd do a TV cable adjsutment as a why not type of thing, then think about replacing the TPS and clean the TPS connectors well. Maybe do a resistance test on the harness between the TPS and the TCU. Then clean the ground. I found a backed out (loose) ground pin at a TCU connector while doing a resistance test on my harness and connectors.
I've taken the TPS off and layed it on a heater overnight, then let it cool down and heated it again (a few times over the period of a day). trying to get the moisture out (moister air out and dryer air in), with positve results. But eventually bought a new TPS nad relegated that old TPS to the spares box.
There is a a plasic wedge in the "T" handle that pushes down on a rod that runs down the middle of the shifter shaft. The "T" handle is held onto the shaft with a plastic sleeve with teeth or a half doze or so plastic hooks. If the "T" handle is loose on the shaft (or shifter) the rod doesn't go down far enough to release the pawl. If *a groove is worn into the wedge* or the end of the rod is worn off, the same thing happens. Occasionally greasing the pawl (and mechanism) under the shift indicator cover, often helps with finding reverse, park or 1-2 without having to force things. There is also often burrs on the lockouts, that may need a little dressing with a small file. The shifter mount has been known to pop a weld and float a little, causing shifting problems.

It does effect my upshifts as well, it shifts way too soon. Where can I find the TPS?

First thing to do is a little research. Search TPS and find out how to adjust it.
When you say find it, do you mean it's location or a parts dealer?
Front of the TB, actually a pancake sensor half for shifts, half for ECU throttle position input. Two sets of wires going to two connectors from the same sensor.
The screws holding it on, have been known to snap off when trying to remove them. I believe they are torx screws (my 87 was phillips head, my 88 is torx head), so you'll likely need a set of torx sockets. A trick is to put a torx socket into the screw end and then tap it with a hammer, which helps loosen the oxidation that jams up the screw. Try to loosen it with a little feeling and don't try to force them much. The time spent in trying to loosen them, whithout snapping them off isn't wasted, drilling them out and finding replacements takes a whole lot longer. I kind of hold my breath every time I try to loosen a set, that haven't been removed before. if you get them loose without snapping them off, coat the threads with grease before reinstalling the screws.
Personally I wouldn't attemp it without some kind of reference material handy. One of those simple processes, that takes a lot of explanation to do right. An FSM is best or maybe Alldata. Alldata is handy anyway for wiring problems and not that expensive.