Sorry that I can't help with the other issues, but I do suggest you check your airbag light. Swap over two light bulbs, Take out any of the smaller ones, like CEL or handbrake, and stick it into the airbag place. Make sure that works!

 

Yep, that's a good idea. Ugh, just as I thought I was done opening up the dashboard for a while!

 

Pulling the negative cable may not reset the airbag light. If you started the car without the gauge cluster harness plugged in, it will throw a code. If you search a member on here posted a guide to reset the light; it's a tap dance involving grounding two terminals on your diagnostic port in a specific sequence. Try that after checking the light bulb though, as I agree it should be lighting up if you have an airbag code.

Also, do you have a 3 or 4 speed? 3 speed is entirely mechanically controlled and the VSS is just there for your speedometer.

 

Oh for clarification I was referring to pulling the negative cable to reset the VSS/starter circuit codes. As mentioned I don't have any air bag lights or codes. I have never removed the gauge harness.

I have a 4 speed.

Sent from AutoGuide.com Free App

 

Well I just found another reason to unplug the gauge cluster ("combination meter" in Toyota speak). It turns out that the output from the VSS is FIRST sent to the gauge cluster, then from there on (besides being used for the speedometer) it's "refined" into a more precise signal that's then passed on to the ECM. So that would explain why the speedometer can still work even if the ECM thinks the VSS is bad... likely a bad connection in the gauge cluster.

Am I right in remembering that someone might have recommended re-soldering (basically carefully just "touching up" with a soldering gun) the connections on the back of the gauge cluster if you have gauge cluster issues?

Hey Haloruler - what is the bulb spec for the SRS light? I like your idea of swapping with a light that's working to test, but on the other hand it would be nice to have a replacement bulb handy right away with the cluster being out.

Still don't have any theories for the starter circuit code though. I can't really believe there's anything wrong with the starter relay as long as the car starts just fine... so I'd think the issue is somewhere else.

 

I honestly don't know. I haven't replaced a single one of those small bulbs. You might want to go for a dealer replacement. The bulbs themselves can't be removed from the twist base they're in, so you'll need to replace the entire assemblies.

 

Double post for I got it! I bought Toyota bulb sockets for 194 bulbs that allow you to insert and remove standard 194 bulbs, instead of using my original soldered sockets. You can probably buy these:

http://www.ebay.com/itm/3-8-Instrum.../271128632660?pt=Motors_Car_Truck_Parts_Accessories&hash=item3f20869d54&vxp=mtr

They're bulb sockets for the gauge clusters. They fit standard 74 bulbs, which are really cheap on eBay. No need to buy 360 degree ones, just front firing. Hopefully, they'll fit our Corollas, but I can't imagine why not. Fitment says no, but look at this:

http://www.ebay.com/itm/Toyota-3-8-.../370723135225?pt=Motors_Car_Truck_Parts_Accessories&hash=item5650d21ef9&vxp=mtr

Same thing, same size, but fitment says it'll fit my 96 Corolla. They're the same, so give it a shot!

 

Cool, thanks! I'll likely give it a shot.


Sent from AutoGuide.com Free App

 

Well boy do I feel stupid - my SRS light is indeed working... I am not sure what happened but when I started the car it came on and went off as normal. Then after I stopped the engine again, flipping the ignition off and then back to acc/on made the light come on again. This did not happen prior. I might have temporarily disabled the light somehow when I went to pull the codes for the VSS/Starter Circuit. And a pack of bulbs in the mail that I won't need since my SRS light isn't out after all

So that leaves me with only the two codes (42 VSS and 43 starter circuit).

For VSS: Keeping in mind that the speedometer works, and that the VSS signal goes THROUGH the gauge cluster before being "refined" and passed on to the ECU I guess I'll start by taking the gauge cluster out and cleaning the connections with a contact cleaner and cotton swab, maybe even pull the old Nintendo trick and get a pencil eraser out to clean

For Starter Circuit: Maybe I'll play around some more with the relay.

I'll post updates when I have them... in the mean time I'm happy to take any new suggestions.

 

So I tinkered a bit more today. First of all I pulled the battery to reset the codes and verified with my scanner that I had lost both codes. Then I went on to do the following:

VSS Code 42:
Pulled out the gauge cluster and cleaned the contacts both on the cluster and on the harness. Used contact cleaner and cotton swabs. They weren't corroded really. My thought behind doing this was that the speedometer receives the speed signal from the VSS sensor and then passes it on to the ECM, so since the speedo is working fine there might be a loose contact between the speedo and the ECM. Started the car (didn't drive it) and the code came right back. Which means I didn't fix it with my cleaning. But I was puzzled as to how I could have triggered the code since according to the A245E repair manual there are only two conditions that would trigger the code, and I did not incur either:

Condition A:

• 30 seconds or more elapses after the PNP switch goes OFF
• The throttle opening ratio and engine RPM are within the shaded area (referring to a graphic that shows RPM being above 2000 and higher)

Condition B:

• 1 second or more elapses after the PNP switch goes OFF
• A VSS signal of 14 km/h (8.7 mph) or more was input.
• The stop light switch is OFF (brake pedal is released)

I assume the "PNP switch" is the Park Neutral Position" switch that is ON if the car is in P or N, and OFF otherwise. It was in P the entire time the engine was running for this test. The other weird thing is that the manual also states that the OD OFF indicator should light up when the code is set. I didn't notice that either.



There are a couple more voltmeter-based tests I could do on the wiring harness that connects to the gauge cluster, so I might have to pull the cluster out again to do those tests. Sigh. I'm still very puzzled though as to how I could have triggered the code outside the possible conditions listed.


Starter Circuit code 43:
I was going through a bunch of wiring diagrams to find possible wiring issues causing this. It appears in addition to trigger the starter there is also a signal that travels from the starter relay (behind driver side kick panel) out to the large fuse box under the hood, then back to the engine ECM. I made sure all the contacts underneath the fuse box were snug and I couldn't see any corrosion. Still got the code right back when I started the engine. There is also a junction point somewhere inside the center console, before it travels to the ECM. Might have to check that.


I'm starting to fear there's an issue with the ECM itself, especially given the strange circumstances causing the VSS code to be triggered. Ugh. Don't want to replace the ECM.

 

VSS Code 42

I had that problem where the speedo needle went to zero and it felt like the transmission shifted to neutral when I tried to apply the throttle.

Replacing speed sensors did not help and so I bought a gauge cluster taken out from a 97 YM that had over 150k miles on the odometer. Tweaked the odometer to match the odometer on the outgoing cluster and so far I have no shifting issues or code 42. Some owners have pointed out the weak soldering on the older 7th gen gauge clusters. I noticed that on the my old gauge cluster and so their conclusions are not unfounded.

So, you can use a soldering iron to fix the speedo head unit or buy a used gauge cluster.

 

Hm, so the darned voltage test seems really hard to do. I also don't understand WHY it's asking me to jack up one of the front wheels while doing the test.. that makes no sense to me. The instructions are (with my comments in italic brackets):

1. Shift the shift lever to N range
2. Jack up the front wheel on one side
3. Turn IG switch ON [which I assume means to the "ON" position, not starting the engine]

And then it says: Measure voltage between terminal 9 [no idea which one this is - the illustration is vague and the wire color is indicated as "V - W" whatever that means?] of combination meter connector "B" [this is the blue connector] and GND.

Apparently if it's constantly 0V you have eliminated the cluster/speedo as an issue. If it's constantly 5V the cluster/speedo is suspicious and requires further troubleshooting (for which I don't have the manual to do). If it fluctuates between 0V and 5V then the ECM is toast.

There is also a voltage test on the ECM terminals... but I cringe at what I'd have to take apart to access that.

 

You have two printed boards. The first is on the back of the cluster. This is really just a ribbon cable of sorts. There is nothing to it, just copper tracks and one or two components; the solder joints you see are probably jumpers. It actually connects to the gauges via those screws on the back. If you take this off though, the gauge itself has a printed board on it with several pieces to it. My old speedometer went out. Replaced it with a junkyard unit, got Code 42. I tried everything that didn't involve replacing parts; cleaning contacts on the VSS and speedometer, new PCB on the back of the cluster... still got it. I realized if my VSS was bad I'd have no speed at all or it would be fluttering. I just lived with it until this fall when I bought a tach cluster on here and replaced it. And what do you know... no Code 42. I would say that if you are getting the code but still have a speed readout, it all but rules out the sensor and wiring, narrowing it to the ECM (still unlikely) or the gauge itself.

 

Thanks! Yeah might just have to bite the bullet and get a tach cluster. No way I'm getting a new cluster without also getting the tach

If I get inspired I might try to re-solder one of these days though.

Any suggestions on the starter circuit code?


Sent from AutoGuide.com Free App

 

Nope, sorry. Never had to deal with that. If you're going to resolder, do tear it down and check the board on the back of the speedometer itself for bad components. I'd start with capacitors, which are probably the most common failed part on older electronics made in the last 20 or so years.

 

Does anyone know how the VSS output is relayed to the Cruise Control system from a wiring perspective? Because Cruise Control still works even with the 42 VSS error code.

If it feeds the VSS signal from the speedometer to the cruise control system then I think I can rule out speedometer issues. If it taps the VSS output directly though the speedometer will still be suspect.

I've looked through wiring diagrams but can't find any wiring to Cruise Control.

(As a reminder, my speedometer displays the speed correctly, but also has the additional responsibility of relaying the VSS speed to the ECM and I'm wondering if that part has issues, or if it's the ECM itself that has issues.)

 

So I haven't found anything for the 7th gen cruise control but assuming it's not radically different on a 2004 Corolla I found this:

According to page 22 of this PDF from a 2004 Corolla the VSS signal goes through the "Combination Meter" (aka gauge cluster) and is then being passed on to the Cruise Control ECU, just like it's passed on to the vehicle ECM. I have to double check that cruise control works flawlessly but I believe it does. Which means the instrument cluster is fine... and unfortunately the issue is in the vehicle ECM... ugh.

"The vehicle speed sensor circuit is sent to cruise control ECU assy as a vehicle speed signal. For each rotation of the shaft, the vehicle speed sensor sends a signal through the combination meter assembly to the cruise control ECU assy"

Link: http://www.toyotadostlari.com/uploads/diagnostics-Cruise-Control.pdf

 

I've determined both codes very likely are related to either the ECM or the wiring connecting to the ECM. I saw there are a couple of TSBs for the 1994 Corolla related to wiring harness corrosion but I haven't been able to obtain them to see if they could at all relate. My rationale for blaming ECM or wiring immediately close to ECM:

Code 42 VSS:

• The Speedometer displays speed correctly, so the VSS is fine
• Just verified while driving yesterday, the cruise control works flawlessly. The Cruise Control depends on the same output from the gauge cluster (same wire) as the ECM to read speed, so the gauge cluster must be working correctly. The small caveat is that this is based on what I was reading for Cruise Control for a 2004 Corolla, but even down to wiring color the description was the same, so I can't believe they would have changed this design.
• So the only items left is the short stub of wiring between the junction connector that takes the gauge cluster output and routes it to the ECM, and then obviously the ECM itself.

Code 43 Starter Circuit:

• The starter relay tests without power, i.e. the resistance tests, checked out fine
• The car starts fine, never any issues
• Which really leaves me with the same two suspects, wiring between starter relay and ECM (to send signal that you're cranking the ignition), and then the ECM itself.

Since the wires come in from the gauge cluster and the starter relay through different junctions to the ECM, it doesn't seem very plausible that the two different wirings both are bad, while mysteriously everything else in those wire harnesses work fine.... So the common nominator here seems to be the ECM... and as someone pointed out earlier, might be a failed capacitor.

It would appear I have three choices:

1. I could leave things as-is. Which is essentially what I've been doing for the last year or so since I first discovered these codes. The only known issue is that the ECM will use a fail-safe mode to shift gears when it doesn't know the vehicle speed. It instead relies on RPM. It actually works relatively fine.
2. I could remove all the lower dash components and dig out the ECM, crack it open, and attempt to solder in new capacitors. Something I've never tried before.
3. I could get a used ECM from eBay and attempt to swap. I see they usually come with "6 month warranty" for whatever that's worth.

Any suggestions? I'm very tempted to just stick with option 1

 

I'm bringing alive an old thread as it appears I've solved the issue. I got a soldering iron recently, and pulled the instrument cluster to inspect and potentially re-solder. I took it apart, and removed the four screws that hold the speedometer in place. I then inspected all the speedometer solders and the capcitors, they all looked great. But as I'm doing it I'm wondering if it could simply be one of the four screws I removed that was too loose. After all the screws are also working as conductors between the speedometer itself and the circuit board on the back. So I put the speedometer back, tighten the screws firmly, and install the cluster again. Had the battery disconnected long enough to reset the codes. Took the car for a couple of trips. Checked for codes again - gone! So all this grief can apparently be attributed to loose screws holding the speedometer in place. No more Starter Circuit code and no more VSS signal code. Go figure....

Writing this update in case anyone else encounters this.

 

lol

But seriously, glad you figured it out. Don't you just love when it's something silly like that?

 

I should add it seems to have better shift points now that it shifts based on VSS signal and not the fail safe mode using the TPS sensor and RPM.

 

Good follow-up!

 

Good job on the repair! Sometimes the little annoying things gets their turn.

Hey, what hardware & software are you using for OBD1 monitoring?
Does it allow display of realtime data-streaming such as sensor-outputs, TPS-position, RPM, injector duty-cycle, O2-sensor, etc.?

 

You might be interested in this: https://www.facebook.com/groups/1505898572959292/

 

I have an Innova 3140b OBD-1/OBD-2 scanner. It will pull live data from an OBD-2 source but only error codes from OBD-1 as far as I know. Isn't that a limitation with OBD-1 itself? No live data? I just use the software that comes with the scanner, which for OBD-1 is worthless since it doesn't give you more than the codes and code descriptions anyhow.

 

The diagnostics port puts out voltages, nothing more. It takes interpretation of the voltages, and what each pin being jumped does, to get live data from it. And with no standardization in place, it's specific to each manufacturer. Someone on Facebook made one for Toyotas, plugs into the diagnostic port and gives you live data through an Android phone. I think it's slow though, but way better than nothing.

 

Just an update after a few days of driving with these codes reset and the ECM out of fail safe mode for shifting gears: It really makes a difference. It's one of those things that you never notice until you fix it and then all of a sudden you notice the car shifts so much better. The fail safe mode was fine, it really was. But shifting based on VSS signal makes for a better driving experience. Just never knew until I fixed it I think in particular it shifts to a higher gear a bit faster, so it feels smoother and more quiet. Perhaps saving a bit of fuel too.

 

First tank of gas after the fix yielded 32.5 MPG on winter gas. Mostly local roads at 45-55 MPH speed limits. Pretty happy about that number.

 

Might be inside the speedometer head?

 

I guess it could. I think I have code 42/43 fatigue.... bleh. Not very motivated to do more about it.

 

Throw the cluster, sans lens, into a mini oven for 5-10 minutes. Reflow the solder. Don't melt anything lol.

 

That's EXACTLY what I was going to advise you to do! Crazy as it sounds, I've personally seen 10 minutes at 350 degrees resuscitate an expensive video card and a motherboard. You can also do it with a heat gun, especially if you have a small tip for it. That's probably better than the oven method, as you don't have to heat everything else up.

 

That seems crazy but might be worth a try!

 

It does indeed sound crazy, but it's no joke-I've personally seen it work. We stuck a video card right in his oven, preheated to 350. Longest 10 minutes I think I've ever seen.