My brake lamp light is on...

[Johan@SjoebergDP]

The BRAKE LAMP, if working correctly, is a so called "Ground disengages" type. Meaning, if the cluster is not seeing ground on the B3 pin when started, or if gound disappears during driving, it will light up.

Adding some references to this circuit from the service manual (1991) :

Best regards,

 

[Valhalla]

Thanks Johan, I'll dig some more. So I tested everything and still go no where. Grounds are solid. So I ordered the right brake light sensor and replaced it. Nothing. I've ordered the left (driver side) sensor and will replace that tomorrow. If that doesn't fix it, I'm left with the safety indicator or a new gauge cluster rebuild.

 

[Valhalla]

Pulled both tail lights out, rechecked all bulbs, connections and wiring. Replaced driver side light sensor too. Checked each bulb, connection and operation separately. Once I plugged in the new sensors, the light went off! Pulled out the key and tried it again at position 2 in the ignition and the light came back on with each new sensor separately. Going back through the system I could not get rid of the light, period. Damn....

Looked at the brake pedal sensor and the grommet was still there. So the only thing I can do at this point is to remove the dash and check for leaking capacitors in the gauge cluster. I believe there are 10-12 capacitors. So if I do find a leaking capacitor(s), I have three options;

1) Try to source another gauge cluster. Not easy to find and who knows when those capacitors will need to be replaced. I'll just be kicking the can down the road until it needs repair
2) Remove the gauge cluster and send it to BrianK so that he can r/r all of the capacitors. The cost is $125 and it will take a couple of weeks for the repair. He's done this service before and he too told me of the potential for dash issues, smoke and fire inclusive of burning up the entire car.
3) Remove the gauge cluster and send it to T3TEC in Japan. They can replace all of the capacitors and then bench test and re-calibrate all the gauges for accuracy. Their shop time is about two weeks also, but the cost is $450 US dollars. Looks like I'm pulling out the dash next week.

 

[hkz286]

Pulled both tail lights out, rechecked all bulbs, connections and wiring. Replaced driver side light sensor too. Checked each bulb, connection and operation separately. Once I plugged in the new sensors, the light went off! Pulled out the key and tried it again at position 2 in the ignition and the light came back on with each new sensor separately. Going back through the system I could not get rid of the light, period. Damn....

Looked at the brake pedal sensor and the grommet was still there. So the only thing I can do at this point is to remove the dash and check for leaking capacitors in the gauge cluster. I believe there are 10-12 capacitors. So if I do find a leaking capacitor(s), I have three options;

1) Try to source another gauge cluster. Not easy to find and who knows when those capacitors will need to be replaced. I'll just be kicking the can down the road until it needs repair
2) Remove the gauge cluster and send it to BrianK so that he can r/r all of the capacitors. The cost is $125 and it will take a couple of weeks for the repair. He's done this service before and he too told me of the potential for dash issues, smoke and fire inclusive of burning up the entire car.
3) Remove the gauge cluster and send it to T3TEC in Japan. They can replace all of the capacitors and then bench test and re-calibrate all the gauges for accuracy. Their shop time is about two weeks also, but the cost is $450 US dollars. Looks like I'm pulling out the dash next week.

it also could be the actual chip itself of which isn't available to purchase :s which means a new cluster

 

[Valhalla]

[MENTION=31273]hkz286[/MENTION] now you are just being mean.

 

[hkz286]

@hkz286 now you are just being mean.

yup that's pretty much how I reacted when I was told,

I was also told it was quite a common fault

mine illuminates all the time, ill take it out eventually to inspect

- - - Updated - - -

3) Remove the gauge cluster and send it to T3TEC in Japan. They can replace all of the capacitors and then bench test and re-calibrate all the gauges for accuracy. Their shop time is about two weeks also, but the cost is $450 US dollars. Looks like I'm pulling out the dash next week.

could you tell me how I go about sending my gauge cluster to T3TEC as I couldnt find a website or anything? and I don't speak Japanese

 

[Valhalla]

@Johan@SjoebergDP and @Kaz-kuzkNA1 & @NSXGB, Are you guys in the know about this "chip"? This is the very first I'm hearing of this part and a failure.
[MENTION=31273]hkz286[/MENTION] do you have any pictures of this "chip"?

If by replacing all the capacitors and that light is still on, I my just remove the light bulb when I reinstall. The tail lights all work, the wiring seems to be intact, the brake light sensors have been replaced and the instrument cluster will be refurbished.....

 

[Kaz-kzukNA1]

Hi, Valhalla.
It's the one marked in blue.
It's on the RPM gauge board and labeled as HC1.

[All photos are courtesy of Mr Toyoizumi at T3TEC, Japan.]


There are several surface mounted chips molded in the black coating so you can't see what's being used and also can't see the actual circuit pattern.
It's in-house hybrid IC so not sold on the market.....
From time to time, the cap leakage caused internal failure of this molded component that the only fix was to replace it with a spare one.

Kaz

 

[Valhalla]

Hi, Valhalla.
It's the one marked in blue.
It's on the RPM gauge board and labeled as HC1.

[All photos are courtesy of Mr Toyoizumi at T3TEC, Japan.]


There are several surface mounted chips molded in the black coating so you can't see what's being used and also can't see the actual circuit pattern.
It's in-house hybrid IC so not sold on the market.....
From time to time, the cap leakage caused internal failure of this molded component that the only fix was to replace it with a spare one.

Kaz

Thanks Kaz for this detailed explanation. So much like the EPS having its sensor box mounted inside the steering rack, Honda decided to make a proprietary gauge cluster?! I'm sure there is no way on God's green earth that you can find one of those chips floating around. The only options would be to by another gauge cluster.

 

[NSXGB]

Thanks Kaz for this detailed explanation. So much like the EPS having its sensor box mounted inside the steering rack, Honda decided to make a proprietary gauge cluster?! I'm sure there is no way on God's green earth that you can find one of those chips floating around. The only options would be to by another gauge cluster.

I would imagine that in the smaller chance that chip is faulty, you would only have to buy the RPM gauge, not the full cluster.

 

[Valhalla]

Ah, didn't realize that that chip was on the individual gauge and not the cluster board. That might be a little easier to source, should I need to. Thanks to all for your awesome help.

 

[NSXGB]

Ah, didn't realize that that chip was on the individual gauge and not the cluster board. That might be a little easier to source, should I need to. Thanks to all for your awesome help.

That is my assumption from Kazs' post anyway...

 

[Johan@SjoebergDP]

I haven't looked at this specific circuit so I cannot tell. However since I'm S2K instrument cluster converted I have a spare original NSX instrument cluster which I can sell for cheap if you need one for spare parts. I can let you know when I'm ready to sell it. I still need it for R&D...

Best regards,

 

[Kaz-kzukNA1]

Hi, Valhalla.
The special chip marked in blue in my previous post is on the RPM gauge side of the circuit board.
Depending on the year model and country spec, some of the early model ones are no longer available as a new parts.

T3TEC has small number of spare chip available so if you can’t find other solutions, you may want to contact them.

Thanks Kaz for this detailed explanation. So much like the EPS having its sensor box mounted inside the steering rack, Honda decided to make a proprietary gauge cluster?! I'm sure there is no way on God's green earth that you can find one of those chips floating around. The only options would be to by another gauge cluster.

For your reference, T3TEC already developed the replacement EPS rack torque sensor I/F board and they can overhaul, test and setup the EPS rack using the real NSX EPS controller with their own jig so we always have backup plan.

OEM I/F board on the left, T3TEC one on the right.

[All photos are courtesy of Mr Toyoizumi at T3TEC, Japan.]

 

[Valhalla]

Oh now that is just showing off! Thanks for the update and options Kaz, love having you checking in on us Yanks.

 

[Valhalla]

For all those who helped, thank you. I sent the unit off to BrianK and he found issues immediatly. He replaced all of the capacitors and sent it back, very well wrapped up. I opened the box and this is what I saw;

I can only imagine that UPS was playing football with my box! Upon confering with Brain, had you ever seen this before? He stated that he didn't. Neither of us really knew what to do so I pulled off the sheild and very gently nudged the dial counter clockwise until it returned to it's normal stop peg at zero. I'll take if for a drive soon to see if all is right with the world.

So I installed it and turned the key;

 

[NSXGB]

Excellent news.

 

[icefire]

Hi Kaz-kzukNA1,

I just read through this thread and followed every bit of your recommendations to debug this issue as my '91 has also fallen victim to this infamous problem. However my debugging so far has come up empty. Here is a description of the symptom and finding;

Symptom: SRS and Brake Lamp failure warning light stays ON regardless (they first surfaced together)

Debugging Steps Taken:

1. Observed all brake lights - all GOOD
2. Grounded Orange/White lead of the 22-pin connector in the trunk - SRS & Brake Lamp remained ON
3. Grounded Orange/White lead of the 22-pin GREEN connector under the dash - SRS & Brake Lamp remained ON
4. Removed and disassembled instrument cluster, check PCB electrolytic capacitors for traces of acid leak or circuit trace contamination - NONE
5. Visual inspection of Hybrid IC appeared in good condition, no contaminants around base of solder joints, epoxy pack looked in good condition

Furthermore I'm puzzled by the fact that since SRS warning light can only be triggered by SRS controller and is independent of the safety warning light system (including brake lamp), why would SRS and Brake Lamp Failure lights go faulty at the same time as if they were interdependent?

I believe what I had done covers everything you recommended. At this point I am hitting a brick wall. Any suggestions on further steps to take?

 

[Kaz-kzukNA1]

@icefire, exactly at what timing did 'they first surface together'?
Was it while you were cruising at speed, pressing on the brake pedal or at the time when you turned the IG key into P2 ON position?

If both first showed up at the time of IG in P2 ON, then not surprised.
They both go through the bulb check mode at the same time but at different check process (BRAKE LMAP; about 2sec, SRS; about 6sec).


For SRS, please check #1 fuse (10A) at the fuse panel inside the cabin.
From the factory, it is covered by a single yellow cover but it may be already gone.


For the BRAKE LAMP warning light, when testing at the instrument cluster, you need to check the green connector at the LEFT side (looking into the cluster while sitting in the driver seat).
By the way, both green connectors are 30pin, not 22pin.

With both green connectors connected to the cluster and if you GND the ORG/WHT wire, the BRAKE LAMP warning light should switch itself off in 2sec after IG P2.

If it stays On, please double check that you are using good GND and also making sure that the ORG/WHT wire is not pulled up by other sources.
You may want monitoring the voltage of ORG/WHT wire while testing.

If it still stays On, I'm afraid the issue is inside the cluster.



For the SRS warning circuit, the idea is the same as the ABS, etc ones.
You need to keep discharging the RC circuit to prevent the Zener diode within the warning circuit from reaching the Zener voltage.
Or, vice-versa depending on how the circuit is designed to act.

For SRS light, you need to apply 10+V at the BLUE wire at the RIGHT (looking into the cluster while sitting at the driver seat) green 30pin connector at the cluster.

If the SRS light doesn't switch itself off, your issue is inside the cluster but very rare.

More likely to be blown #1 fuse at the fuse panel inside the cabin (i.e., short circuit inside the SRS unit).
If you have aftermarket steering wheel/boss/hub, please make sure that whatever the method you used to simulate the presence of SRS is still fine.
With 91, dual SRS or seat belt pre-tensioner were not available so should be easy to check.


Kaz

 

[icefire]

@Kaz-kzukNA1

, exactly at what timing did 'they first surface together'?
Was it while you were cruising at speed, pressing on the brake pedal or at the time when you turned the IG key into P2 ON position?

If both first showed up at the time of IG in P2 ON, then not surprised.
They both go through the bulb check mode at the same time but at different check process (BRAKE LMAP; about 2sec, SRS; about 6sec).

I hadn't driven my car in over a month. So when I recently first started it both SRS and Brake Light Failure Lamp came ON and stayed ON.

For the BRAKE LAMP warning light, when testing at the instrument cluster, you need to check the green connector at the LEFT side (looking into the cluster while sitting in the driver seat).
By the way, both green connectors are 30pin, not 22pin.

With both green connectors connected to the cluster and if you GND the ORG/WHT wire, the BRAKE LAMP warning light should switch itself off in 2sec after IG P2.

If it stays On, please double check that you are using good GND and also making sure that the ORG/WHT wire is not pulled up by other sources.
You may want monitoring the voltage of ORG/WHT wire while testing.

If it still stays On, I'm afraid the issue is inside the cluster.

I disconnected the 22-pin white connector in the trunk, and then put a needle pin into pin 18 (ORG/WHT) of the left GREEN connector while plugged into the cluster. With IG in PII it measured +11.1V at pin 18. I then grounded pin 18 with an alligator clip to GND (on steering column). Brake Failure Lamp unfortunately remained ON constantly.

For the SRS warning circuit, the idea is the same as the ABS, etc ones.
You need to keep discharging the RC circuit to prevent the Zener diode within the warning circuit from reaching the Zener voltage.
Or, vice-versa depending on how the circuit is designed to act.

For SRS light, you need to apply 10+V at the BLUE wire at the RIGHT (looking into the cluster while sitting at the driver seat) green 30pin connector at the cluster.

If the SRS light doesn't switch itself off, your issue is inside the cluster but very rare.

More likely to be blown [URL=http://www.nsxprime.com/forum/usertag.php?do=list&action=hash&hash=1]#1 [/URL] fuse at the fuse panel inside the cabin (i.e., short circuit inside the SRS unit).
If you have aftermarket steering wheel/boss/hub, please make sure that whatever the method you used to simulate the presence of SRS is still fine.
With 91, dual SRS or seat belt pre-tensioner were not available so should be easy to check.

First of all I did find Fuse #1 (10A) blown. I replaced it and shortly after it was blown again. Not instantly but rather slowly over time !!! Sounds like an internal short!

As for applying 10+V to the blue wire, I did so and SRS light immediately went OFF. Just curious, what is the BLUE wire?


Based on my findings would it be correct to conjecture;

1. SRS - likely an internal short in the SRS module? If not, what else could repeatedly blow the 10A fuse?
2. Brake Failure Lamp - instrument cluster problem? I did take apart the cluster and examined the circuit board closely and saw no signs of defective electrolytic capacitors or other burnt components. Further suggestions?

Jerry

 

[Kaz-kzukNA1]

Hi, Jerry, [MENTION=10822]icefire[/MENTION] .

For the SRS, if you blew the #1 fuse immediately after replacing it, very likely to be the same cases as in my blog; http://www.nsxcb.co.uk/blog.php?1811&blogcategoryid=48
As it's a safety device, you should replace it with a new one (although the LHD SRS unit is NLA) and not repairing.
As your NSX is 91 model, according to another owner on here, the internal circuit board is the same as the old days Honda/Acura Legend.
I don't recommend using used parts with unknown history though....

The BLUE wire is the output signal from the diagnosis circuit within the SRS unit.
If the diagnosis test passes all the protocols, it will keep the output at high level and switches off the SRS light.



For the BRAKE LAMP warning light, you mentioned that

I disconnected the 22-pin white connector in the trunk, and then put a needle pin into pin 18 (ORG/WHT) of the left GREEN connector while plugged into the cluster. With IG in PII it measured +11.1V at pin 18. [Kaz-kzukNA1; Did you power cycle the system here?] I then grounded pin 18 with an alligator clip to GND (on steering column). Brake Failure Lamp unfortunately remained ON constantly.

Did you power cycle the system before GNDing the ORG/WHT wire at the timing as in the blue comment above?

The BRAKE LAMP light would be triggered as you disconnected the 22pin connector inside the boot/trunk when you first went into P2 ON.

I haven't tested this so I could be wrong but since the warning circuit is 'latched' type, you should turn IG into P0 OFF and GND the ORG/WHT wire first before going back to P2 ON.
Otherwise, the BRAKE LAMP warning light is latched and stays ON even if you GND the ORG/WHT wire, unless you power cycle it.

Not sure exactly which part of the steering column you used as the GND.
Best to use the existing known good GND.



Kaz

 

[icefire]

Don't know how many of you are still suffering from this infuriating Brake Lamp Failure gremlin but thought to share with you my recent "lucky" success in this issue.

Ever since the onset of this problem I researched many forum posts here, read time after time the service manual, got expert advices from kind folks like Kaz, and spent hours tracing circuits ... unfortunately to of no vail I was not able to pinpoint the specific cause that triggered the brake light failure lamp. Since debugging the peripheral hardware provided no clues whatsoever, I decided to shift my focus to the source of annoyance, instrument cluster. I removed and disassembled the cluster down to the circuit board level. Closed up examination (under magnifying glass) of the circuit boards showed no signs of component deterioration and leakage of the infamous electrolytic capacitors. All solder joints also appeared clean. So once again I hit a dead end. I know, some of the guys ended up sending the cluster to BrianK to replace all capacitors in hope it would fix the problem, but there was no guarantee it would fix the problem, not to mention the $200 cost and the 2~3 weeks down time. I just wasn't convinced.

So in a hail Mary attempt I got a bottle of electrical contact cleaner spray and a fine circuit board cleaning brush and thoroughly sprayed and cleaned every nook and cranny, every solder joint and every connector contact of the two cluster circuit boards. Granted there was no basis for doing this but if all's failed at least i would have a clean cluster circuit board to admire. But ... to my surprise, after reassembling the cluster and installing it back in the dash, as I turned the ignition key to P2, Brake Lamp Failure light came on through start-up diagnostic and then .... it turned OFF! Yes, it turned off! It is functioning once more!

One down and one more issue left to tackle. It's been a good day :smile:

 

[Kaz-kzukNA1]

That's the typical of the 'BRAKE LAMP' warning circuit failure mode.
If you managed to fix it by cleaning the board and never comes back again in the future, good on you.

According to Mr Toyoizumi at T3TEC Japan who already repaired 193 of gauge units by August this year, this warning circuit failure tends to show up under high humidity.
He has seen many intermittent cases and cleaning the circuit board did improve the situation but not a fix.

If you managed to improve the situation by cleaning the board, that means at least you found the potential cause so next time when you have a chance to take the gauge unit out of its place, I recommend replacing all electrolytic capacitors, clean the circuit board/lens cover, apply coating and re-calibrate the gauges.


Kaz

 

[goldNSX]

I'm glad the error went away but it might come back. The caps don't need to be leaking, they might look ok. It's well known that they loose their capacity over time. Caps that go bad can destroy other components (ICs) if they depend on working caps.

Guys, the caps are prone to go south after 20-25 years, it's a part with planned obsolence if you like. Why don't you replace them all BEFORE they fail to prevent them creating additonal gremlins. This forum is full of infos about which components need to be addressed. That's an easy winter project every NSX should undergo every 20 years.

As for the SRS: what's wrong with replacing caps with the same value and quality in the SRS unit? You just change the caps, that's all. Ok, if you live in the US a shop might refuse to work on it or require a strict liability waiver.

 

[icefire]

I recommend replacing all electrolytic capacitors, clean the circuit board/lens cover, apply coating and re-calibrate the gauges.

Kaz

Kaz,

What coating are you referring to?

Jerry