My Samurai is a 1992 TBI 1.3 model
I picked my son up from school this afternoon...ran perfect on the way there, as always.  When I started the samurai back up, check engine light came on after a few seconds, and engine ran really rough.  Long story short...after a quick check under the hood with no obvious problems, I limped home at about 35 MPH.  A check of stored codes reveals a 22 and a 32 (TPS and MAP sensor circuits).  Both sensors look fine, and the wiring harness isn't chafed anywhere.  All grounds are intact and in good shape.
The reasons I suspect my ECM:

BOTH sensors failed at the same time?  Nah. Only thing common to them is the wiring harness and the ECM.
 
The code blinks a 22 three times in a row, followed by the 32 three times in a row.  Never seen it blink the same code multiple times in the fault stack...

The problem began on a startup, not while driving or any specific event

The ECM is original

SO...I plan to pull it out and pull the cover off, but I can't get to it for a day or two.  In the meanwhile, maybe somebody here has a piece of advice?  Make me think of something I haven't already thought of?

I can do component replacement and PC board repair, but I have no way of troubleshooting the ECM; I'm hoping for something obvious when I pull the cover off (I've never pulled one apart, but I'm aware of the leaky capacitor issue).  If I can't repair it, any recommendations on rebuilders?  
I see these 2 places:
http://www.avproecm.com/
http://www.foreignecurepair.com/ECMs_ECUs_samurai.htm
Both are about $200.  

Edit:  Need to get it fixed before the fall classic..or I'll be driving the deuce through all of those narrow trails in the woods.  Won't be pretty  

Replace the capacitors.  Plug it back in and try it again.  If it is still throwing codes and running crappy, pull the ECM again and test some of the copper traces around the capacitors (sometimes the caps leak and it eats through the traces. The "break" in the trace usually can't be seen, you will have to do a continuity check with a meter from "point a" to "point b".
If you can't get it figured out or fixed, there are a couple of guys on zuwharrie.com that will do just as good of a job (if not a better), than either one of those other places.....and for less $$.

Sometimes you can't see any damage until you remove the caps. If its original then they're 20 years old and most likely are well past the point that they should have been replaced.

Well, I pulled my ECM and there's no obvious damage or evidence of leaky caps.  I'm going to pull them anyway and see what's underneath...I'm really hoping to find an open or high resistance trace, because if there isn't then I will have pulled it apart for no reason, and I still have some other issue 
Anybody heading down hwy 26 to the coast from Portland this weekend with their EFI Samurai
I'd like to plug my ECM into a known good running rig and see if it exhibits the same symptoms...

why not adjust tps and clean maf before throwing $$$ at it? (those are the codes mentioned)

reset ecu, see what it then tells you...?

22 and 32 indicate that the 5V supply to the sensors is bad, either that or the wiring has a break. Its more likely to be an ECU failure than for both sensors to fail simultaneously.

Ohh, if there is a 5v (reference voltage) failure, then there is a failure
in the chip that controls that voltage, I had to test that circuit on the
MegaSquirt when I built it to make sure it worked before moving on to
the next stage of assembly.

That or a broken wire like Rhino mentioned, check your harness for
rubbing where it goes thru the firewall and in the engine compartment

Wild

I pulled the caps; no evidence of any leakage.  I tested all of the traces near the caps for continuity and all tested good.  I ordered new caps (I had the original ones that are known for leaking), and see if my problem goes away.  I did inspect the harness pretty thoroughly before I pulled the ECM and it looked good.  If the problem continues once I have the ECM repaired and installed, I will continue the troubleshooting...

why not adjust tps and clean maf before throwing $$$ at it? (those are the codes mentioned)

reset ecu, see what it then tells you...?

The caps cost me about $3.00 total...and I figured that I should replace them before they damage the PC board anyway. 
I may be wrong, but the Samurai's MAF isn't like a standard pass-through MAF; it just looks like a pressure sensor. 

No MAF on a sammy it has a MAP. Any electrolytic capacitor of that age is likely to fail, all electrolytics have a limited life and they are quite highly stressed in a low impedance TBI application.

22 and 32 indicate that the 5V supply to the sensors is bad, either that or the wiring has a break. Its more likely to be an ECU failure than for both sensors to fail simultaneously.

My thoughts exactly...

I don't have a wiring diagram that shows if each sensor has its own power supply to the ECM, or if they are tied together somewhere in the wiring harness.  If the wires are separate, I will continue to focus on the ECM.  If they are tied together somewhere outside of the ECM, then I will examine that wire run VERY closely...

Does anybody have a pinout for the ECM?  If my problem remains, I would like to give it some power and do a little troubleshooting. 

That or a broken wire like Rhino mentioned, check your harness for
rubbing where it goes thru the firewall and in the engine compartment

Wild

Short of pulling apart the entire wire bundle, I did check quite closely.  My guess is that each sensor has its own set of wires that go to the ECM; if that is the case, then the chance of two wires chafing through simultaneously is pretty darn low, and I'll focus on the ECM itself.  If they are spliced together somewhere outside of the ECM, then I do need to check that wire run closely...

The 5v reference voltage is usually common to each sensor, it's
the return value that has an individual wire to communicate to
the computer what's going on

And yes, in the other thread on this subject, there was a link that
had the pinouts for the 8V ECM on list says pin 14 on the chip is 5v

"Pin A23is dead syndrome:  +5v refr pin output to sensors.
+5V Vref traces like to blow up, Vaporize !
 Place jumper at R205  to  Pin A23 if that happens.
SEE DRAWING HERE.

http://www.kick-fix.com/ECU/5vdc-Refr-supply/8valve-5v-refr1.JPG


Wild

The 5v reference voltage is usually common to each sensor,

Wild

Good info there; thanks!

So, I got my hands on a decent wiring diagram for the sami, and it looks like the only common wire to the MAP and TPS sensors is a Gr/R color wire; when I get a chance tomorrow evening, I'll head back out to the shop and check continuity from the plug on the ECM to the sensors.  There is also a Gr/Y wire that is common to those sensors, but it is also common to the CTS and air temp sensors...I'll test it for continuity as well, but I'm assuming that since I didn't get any codes associated with them that wire is intact. 
Thanks for the help everybody; I'm going to learn a little something about the ECM in my zuk for sure...
As soon as I get my new caps, I'll do some voltage testing on the ECM and figure out if ref voltage from the ECM is dead. 

Got it!

So, the GR/R colored wire that is common to the 2 sensors does pass a reference voltage to them.  It connects to pin 20 on the passenger side connector on the ECM.  After figuring this out, I tested the trace from pin 20 through a few components and sure enough, it was open.  There was no evidence at all of any kind of deterioration or damage, and it was a good distance away from both of the caps (maybe 1/2").  There were several traces closer to the caps that I tested when I first pulled the cover off (all were good).  The trace that was bad I did not test at first because it was farther away from the caps and showed no sign of damage.  It was broken right where a surface mount cap was soldered to it (C22).  
I went ahead and replaced the caps with new ones; P/N P10776-ND and P/N P10773-ND from www.digikey.com.  To fix the break in the trace, I just re-soldered the surface mount cap (I used a 15W needle tip soldering iron).  Continuity was good after that, so I cleaned up the board with some isopropyl alcohol and closed it up.  Sami runs perfect now...nice to spend just a few $$ on a repair instead of hundreds of $$ on a rebuild  
Attached a few pics...showing the old caps installed on the board (the offending Rubycon ones circled in red), the board with the caps removed (the broken trace circled in red), and the new caps next to the old ones.  

Wildgoody and Rhinoman, thanks for the help.  Your posts got my troubleshooting started...would have taken me a lot longer to figure it out without your info...