I’m trying to calibrate a new TPS on my 95 accord. I can’t seem to get both the closed and WOT voltage where they need to be at the same time. When closed is at .45v, WOT is at 4.28v. When I get WOT to 4.5v, closed is way off from where it needs to be.

Which one is more critical to be just right? WOT or closed? I’m trying to clear a code 7 and get the ECU out of limp mode.

For those who have gotten both readings right on the mark, how were you moving the TPS?

 

Just tried to calibrate it again, still off.

When WOT is 4.5v, closed is 0.61.

When closed is 0.5v, WOT is 4.44v (this is where I left it at tonight). Should this be close enough to clear the Code 7 CEL light?

Driving myself nuts here.

 

It's more of an approximation of voltages.
FSM calls for ~.5V at idle and ~4.5V at WOT.
I would keep the idle at .5V or less. This allows the ECU to know that throttle is closed and at idle, any higher voltage and the ECU may interpret it as the throttle opening.
It's not an exact measurement, the ECU is just looking for a change in voltage output from the TPS showing that the driver is opening the throttle.
WOT voltage is not really a concern as by the time the voltage gets that high the ECU already knows what the driver demand is and it is watching the other sensors to properly calculate air density for fuel and timing.

 

thanks! I did have it at 0.49 closed and 4.33 open at one point earlier today but it was still giving me a CEL code. Could it be an ECU issue?

 

So I noticed today that the CEL doesn’t come on immediately when I start the car. It takes about 20 seconds and then the light comes on and at the same time the rpms rev up to between 1500-2000, before that it idled at 1000.

Later today it got unseasonably cold outside (in the upper 30s). I went to start the car and the CEL came on as usual but the car was idling at 1000, even after the light came on. What does that mean?

Also, should I just replace the whole throttle body with one that already has the TPS on it?

 

It might be, but don't throw an ECU at it just yet, you will want to verify it is actually an ECU issue and not simply a bad connection/wire.
If the ECU is not seeing the TPS signal properly, or there is something else amiss then the idle may hunt. Idle should be 750rpm.
One other issue is that the speed density system on these cars uses most of its computing power on calculating air mass. And being OBD, its self diagnostics can be limited.
Just because it spits out a code doesn't necessarily mean that is the problem. It may just set the code to the closest or nearest possible issue it is encountering.
That's why it is best to always verify that the sensor/component that has a code set to it is indeed faulty before replacing it. As the code may just be a symptom of the real issue.
It could be the cold, it could be the programming is different for colder temps, it could be that the hunting was a freak occurrence or part of the re-learn if the battery/memory was disconnected.
No need to replace the TB if it is not damaged.
You had a TPS code, replaced the TPS and still have a TPS code.
It's probably safe to say that the TPS itself is not the problem.
If there was any work, or incident, done to the vehicle prior to the TPS issue I would look there first. No matter how inane the issue may have been.

Do you have an FSM for your Accord?

 

How would I check to see if proper communication is occuring between the ECU and TPS? I checked the red wire and there is 5v constant.

So, if the TPS is slightly off with calibration, it shouldn't be enough to kick a code out?

I was also thinking it may be something else. I'll tell you a few issues I have to resolve: The Air Intake hose is cracked in two places....plan to replace it this weekend. Also, the car sat for awhile before I acquired it...therefore the gas in the tank had gone bad. I siphoned most of it out and replaced the fuel pump, but I believe there is still a bit mixed in. The fuel filter and injectors also likely need to be replaced. Would any of those issues be close enough to cause that code?

And yes, I just realized I do have the FSM (just checked my PMs on my computer, thank you!)

 

I just installed a new TPS this afternoon (previously experienced off-idle bogging and slightly rough acceleration around 2500RPM). You might want to adjust your idle screw. My car did the CEL and high/bouncing idle thing before I did the idle adjustment outlined in the service manual:

1. Warm up car till the radiator fans come on
2. Turn off car and disconnect IACV plug
3. Start car and give light gas until idle is stabilized
4. Adjust idle to roughly 700RPM
5. Turn on headlights and see if idle raises to roughly 770RPM (half a tach needle-width higher)
6. Turn car off, reconnect IACV, and reset ECU (remove 7.5A backup fuse under hood for 10+ sec)
7. Start car and let idle for 1 min (should stabilize at 700RPM)
8. Turn on headlights and idle for 1 min (should stabilize at 770RPM)
9. Turn off headlights, turn on vent to max and AC on and idle for 1 min (should stabilize at 770RPM)
10. Turn off car, and idle should be set

I experienced a longer ECU learning time with the heater and AC on. Things got funky for a bit, with the idle ramping up and down for over 1 min, and I kept it going until things settled down. Now that that's done, the idle is perfect, and I actually have the correct idle up for headlights and AC, which never happened before (even on my first Accord). And yes, my new TPS didn't completely line up with factory voltage specs, so I erred on the WOT side, since that's where throttle position should matter most.

Quick tips: the idle adjustment screw needs a very thin tipped flathead screwdriver. It's also not in the most convenient spot, so a bit of finessing to get the screwdriver lined up is in order.

 

Thanks, I’ll try that tomorrow. Was this on the same generation Accord? And it cleared the CEL? Was it running in limp mode prior to that?

 

Yes, 96 LX F22B2. The idle was at 2K+ at cold start and 1.5K when fully warmed up, with random idle bouncing while at stop lights/signs, after the CEL (code 7) lit up immediately on the first start. Power was there and all, and was actually smoother than before, so I don't think it was limp mode. The idle adjustment process requires resetting the ECU to learn fast, so no CEL showed up after that. I took a 10 mile drive after the process, and the CEL didn't come back on. Everything is smooth (finally!), and idle is probably as perfect as it can get.

 

That's the reference voltage. Ideally you would watch the voltage changes at the ECU, back probe the signal wire and watch that the voltage changes is linear. Start the car let it warm up and keep watching the voltage. If you start getting fluctuations then you know the signal is being interrupted or something else is amiss.
Section 11 'Fuel and Emissions' is going to become your favorite go-to in the FSM.
Best to follow the TPS diagnostics from the FSM.

If the car has been sitting and you have the car as is with unknown history, it is best to get back to basics and verify the engine itself is mechanically fine.
Pull the spark plugs and verify they are not worn out and burning clean.
If your idle/closed throttle setting is .45V-.5V it should not set a code. Even if it was slightly higher the only thing would be the off would be the fueling/timing. But a code I don't think it would.
On the B1 there is the FIA system. One vacuum hose does connect to the intake tube. If it is not getting the correct signal it may affect the idle control.
TPS code not likely, fueling codes(via an HO2S code) might.
Put in some 91/92 gas, gas dryer, and your preferred fuel system cleaner/stabilizer.
Get a fuel pressure gauge.
With vacuum hose off the FPR, you should see 38-46PSI
With vacuum hose on the FPR you should see 30-37PSI.
If you are getting lower than anticipated fuel pressure, artificially raise it by pinching the return line. If pressure increases then you know the pump is working, need to find out what is causing the pressure loss.
If you are getting higher than anticipated fuel pressure, there may be a pinched return line or junk in the system.

Verify battery is healthy, a battery that is not properly charged can cause strange electrical issues due to under voltage.
Verify the grounds, bonding jumpers and battery cables are in good condition with clean tight connections.

Cars that have been sitting may have other issues that the previous owner is unaware of, or just plain forgot.
Try not to focus too much on the TPS at the moment, make sure everything else is in good order before chasing your tail on a possible phantom issue.

As for idle setting, before doing any kind of adjustment.
Pull out the air bypass idle screw.
Count the number of turns it takes til you feel it 'click' at the end of the thread.
Pull out the screw and clean any coking/soot off the needle.
Then clean out the air bypass port.
Reinsert the screw the correct number of turns.
Unless the TB is fubar'd, the engine is weak on compression, or something else is going on, the idle screw air bypass rarely needs to be adjusted. Mostly soot/coking just need to be cleaned off. Just like with the IACV.
Saucy.
That will be your go to, IIRC the TPS diagnostics is a few pages long, well worth investigating. Also if your car is an automatic IF the TCU is bad it can also cause certain backfeed issues that would screw with the ECU. But first verify the wiring and connections to teh TPS are fine and follow the FSMs diagnostic chart(s). Page 11-46.
If the car was sitting, there may have been a pest infestation that gnawed on the wires, that would account for the CEL.

 

Thanks, very helpful info! I did download the FSM and separated the TPS pages and have been reading through it. Here’s all the things I have done to date since acquiring the car:

Replaced spark plugs and wires
Replaced distributor
Replaced original ECU (but it was with a used unit off eBay)
New Battery
New Alternator
New Fuel Pump
New Main Relay

There were signs of rodent activity when I first inspected the engine bay (droppings but no visual gnawing activity, could just be hidden). I do know that my fuel injector power wire had a break somewhere that I couldn’t find....ended up running a new wire to the main relay and it worked.

 

Then it is most likely a damaged wire harness. Urine can be fairly corrosive, don't dismiss it. Locate all the soiled areas, use some simple green and a hose to wash off the nasty.
Even if there isn't a full cut, a nick in the insulation will allow the wire to be exposed. Which can then corroded and break.

But first things first.
Do the TPS diagnostics and backprobe the ECU (per FSM 11-46) to figure out where the fault in the wiring is.

 

I had a chance to try the first step of testing the voltage of a bridge between the Yellow/Blue wire and the Green wire....only got 0.11v, so there’s an issue. While connected to the TPS, I do get the proper voltage from the yellow and red wires individually. I tried the next step on bridge between the yellow wire and chassis ground and still nothing. Will test at the ECU next.

 

Ok, just tried the ECU and used a paper clip to make a jumper between D20 (yellow) and D22 (green) and there was barely any voltage (0.02v) between the two. I did test the D20 individually and there were 5v.

So based on the FSM flow chart, am I to conclude that it’s a bad ECU?

 

I have a working ECU I picked up at a junkyard awhile back, but it’s from a manual (I have an automatic). Could I use it to test without causing any damage?

 

Per the FSM you do not jumper D20 to D22, you measure the voltage between the two with a voltmeter.
You are measuring the difference of potential(voltage) between the two terminals.

If you jumper the two together you make the difference of potential nil. Jumpering the two together makes the two circuits as one so there would be no voltage difference.
What did you measure it against? Ground? Battery voltage? The wind?
Follow the FSM directions all the way through to conclusion.
Verify your conclusion by testing again while moving/wiggling wires to verify the wire/pin/connector is not damaged.
When you are satisfied that your results are true then follow the recommended repair.
Don't think so. You haven't followed the FSM completely. You cannot skip steps, have to trudge through them to completion to get the correct results.

It's science, fudging steps will give you fudge results.

What I can see from what you have already wrote.
Turn ignition on(II)
CEL is lit and you retrieved Code #7
Turn ignition off(I)
Disconnect 3P TPS connector.
Turn ignition on(II)
Measure voltage between Yel/Blu(+) and Grn/Blu(-) terminals of TPS connector.
Q: Did you get 5V?
A: No(post #14)
Per the FSM on 11-46 we move to diagnostic column B.
Now measure voltage between Yel/Blue(+) and body ground
Q: Is there approx 5V?
A: ???

Next you posted about jumping D20 and D22 with a paperclip, which is not mentioned in the FSM at all.

You will need to go back and finish column B diagnostics on page 11-47 to finalize your diagnostics and confirm there is an issue with either the wiring or the ECU/ECM.
I would not change out the ECM until you can verify the results are indicative of an ECM issue, and not a wiring or sensor issue.

As for the MT ECM in an AT car.
Engine will run but there is no communication between the ECM/TCM. TCM will most likely not command a shift or if it does it will cause a limp mode function and put the trans into 2nd gear. And your D4 lamp will most likely illuminate and set a code for the FAS(communication) wire.

But get back to your original diagnostic steps and complete them as per the FSM.

 

Dang, I feel like a dummy....I should know better. Ok, I measured the voltage the correct way at the TPS and ECU. Ended up at the D22-D11 test and did not get any voltage at all, even with the TCM harness disconnected. According to the FSM, it recommends replace TPS (which I already have) or repair open or short on the Red/black wire between the TCM and TPS.

 

Update: I ended up running a test wire from the TCM & ECU to the TPS and it worked! CEL went away soon as I started it up.

It still revs high at Idle, so that will be another task.

Any suggestions on best method to splice and run a new wire? I want to make sure not to damage the harness.

 

Update 2: I was looking at places to run a new wire and I believe I came across the source of my issues. I found a bunch of wires chewed completely by rodents I assume. Attached is a picture....this is in the back right of the engine bay (at the firewall behind the steering wheel end). I don’t even know where to start to fix these or even what they all lead to.

 

Simple fix is to merely find the matching wires and crimp them back together.
Clean up the area, simple green and some clorox wipes to aid in cleaning up the busted wires.
Identify pairs of gnawed threw wires and get to crimping and heat shrink covering them up. Should be as simple as color to color.
If you can remove that section of wire harness from the car and work on it out of the car that would be easier as it is cramped there behind the spring tower and under the resistor box.

 

Thanks, I actually got it all figured out earlier today. What happened was, some of the wires were chewed all the way to the under the wrapping tape and I couldn’t see them initially because the ends were hidden. Once I pulled the wrap back, I could see the ends and match them up. Some needed to be extended with new wire.

The VSS also went out on me and I had to replace that as well. That FSM helped me diagnose that one easily!

Now it runs like a champ (so far)!