96/98 Alternate Service light reset

[cn90]

Spend 5 sec cutting a piece of black electrical tape and you will never care about this reminder light again.

[erikv11]

Spend 5 sec cutting a piece of black electrical tape and you will never care about this reminder light again.

Since you keep posting this option I will opine too ... I really don't want to have a piece of black tape on the face of the speedo cluster, I think the service light for a few minutes every day would be far less annoying. But I too, pull the bulb on my P80s.

[misha]

Does it appear on bulb check,when key is in positon II,along with all other little bulbs?

Is it orange light with "service" message on it?

It does test.

You're the 1st man on the planet who managed to reset service light just with disconnecting battery.

The ONLY explanation on this could be that any of previous owners "bridged" actual check engine light with service light...in this case service light will turn on instead of check engine light when something is wrong.

If BOTH of them turns on on bulb check,then actual check engine light is bridged with some else little bulb from the cluster.

[cn90]

...Since you keep posting this option I will opine too ... I really don't want to have a piece of black tape on the face of the speedo cluster, I think the service light for a few minutes every day would be far less annoying. But I too, pull the bulb on my P80s...

Erik,

I don't even see the tape because it is black and it blends in very nicely with the dash.
I find the light, even present for 2-3 min before going out, annoying.
Pulling the bulb is more work; black tape is much much faster.

Every day when I start the car, I don't even notice the black tape.
For the sake of trying, cut a piece of black tape and place it there and go on with your daily life and see if you notice it (pretending like you didn't install tape).
Psychologically, when:
- I start the car, I look for dash bulbs that should turn off once the engine is running.
- When I am driving, I only look at speedometer, tach, fuel gauge etc. I don't look down at the black tape, this is why it is not noticeable to me, especially when it is dark at night.

[abscate]

IF the tape doesn't match the car color, your flux core can warp....

[precopster]

Mike's DIY METHOD

There are so many shades of black I honestly wouldn't know which colour tape to use. I'm going down to the hardware store to bring home a Dulux colour chart then I'll take a photo of the dash panel with a digital SLR taken at medium light levels . After I calibrate my computer screen's RGB output, I'll download my printer's colour profiles then I'll print out the photo of the dash and bring it to the paint department for spectrum analysis. Then I'll have a small pot of low sheen enamel oil based paint made up.

Cut some masking tape in the shape of a frame with a hole the size of the service light and attach it to the dash exactly where the service light is located.
Fine sanding of the dash plastic with a miniature block will prepare the surface and protect the paint from premature peeling, after all this IS your pride and joy.

Use of a fine horse hair artists brush will avoid streaking. Avoid painting the dash panel in high heat days. Allow at least 24 hours to dry then carefully remove the tape (perhaps start the peel with a blade)

There you have it. If the service light is professionally paint matched you can all sleep easier at night.

[ThommyKent]

Does it appear on bulb check,when key is in positon II,along with all other little bulbs?

Is it orange light with "service" message on it?

It does test.

You're the 1st man on the planet who managed to reset service light just with disconnecting battery.

The ONLY explanation on this could be that any of previous owners "bridged" actual check engine light with service light...in this case service light will turn on instead of check engine light when something is wrong.

If BOTH of them turns on on bulb check,then actual check engine light is bridged with some else little bulb from the cluster.

I think it was just a lucky fluke. I never intended to try and reset the service light. It just happened with a change of the battery. When it comes on again in a few months, I will try disconnecting the battery and will report what the result was.

[mecheng]

ThommyKent,

1. There have been a multitude of cases where the '96-'98 850/S70/V70/C70/XC70 SERVICE light does not stay off (ie, remain off during the first 2 minutes after engine start) immediately after the battery is disconnected. This is the first time I've heard of where a battery disconnect did reset the SERVICE light.

Of course, it's well known the battery disconnect resets the CHECK ENGINE light. But, to my knowledge, the battery disconnect has never been reported previously to also reset the SERVICE light in the '96-98 850/S70/V70/etc (in a verifiable and predictably repeatable manner).

It could be that your SERVICE bulb just happened to coincidentally burn out. I think you just got lucky about that.

I also had the SERVICE bulb (apparently) burn out last year in my '97 850 T5. At that time, I had already learned that there is a bit in the response to the ECU 51 (COMBI) A50601 request that accurately reflects when the SERVICE light should be turned on and off. It is bit 3 in the Status byte returned in the E506 response to the A50601 request. This bit gets set/cleared regardless whether the bulb is burned out or not. The same Status byte is also available in the response to the ECU 51 A50701 request (along with a bunch more info). From that bit in ECU 51 A50601 and A50701 and from an examination of the ECU 51 B903, B904, B905, B906, B907, B908, B909, B90A data I could see that the SERVICE light should actually have been on. Hence, I concluded that it had coincidentally burned out.

2. If the response to A50601 is "84 13 51 E5 06 7B 4E", then "Engine is On, SERVICE Light should be On during 2 minutes after engine is turned on (assuming the SERVICE light bulb has not been removed and has not burned out)".

If the response to A50601 is "84 13 51 E5 06 73 46" and if the A50601 request was issued within 2 minutes after the engine is turned on, then "Engine is On, SERVICE Light should be Off during 2 minutes after engine is turned on".

3. You can check the bulb without removing the dash by issuing the A50601 request soon after turning the engine on -- and definitely before the ~2 minute SERVICE bulb illumination timer expires. If the response to A50601 includes the 7B and if your SERVICE light is not illuminating at the time that A50601 was sent, then the SERVICE bulb is very likely burnt out.

If the response to A50601 includes the 73, then repeat a few times to make sure you're actually issuing the A50601 request within 2 minutes of turning the engine on.

4. One way to send the A50601 request is with the latest version of volvo850diag downloadable from volvo850diag.zip You might do something like:

4a. Download volvo850diag.zip, extract, and setup.
4b. Get ELM327 device drivers installed.
4c. Display volvo850diag.html (extracted from volvo850diag.zip).
4d. In volvo850diag, click "Start Realterm".
4e. Turn engine on (and wait about 5-10 seconds).
4f. In volvo850diag, click "My Extra Scans". [That's a quick way to check the SERVICE flag using A50701.]
4g. Turn engine off
4h. In volvo850diag, click "ATZ Setup for ECUs 51/58/01..." and wait for it to finish.
4i. Click ECU 51 "Setup".
4j. Turn engine on (and wait about 5-10 seconds).
4k. Click either ECU 51 "ATSI" or ECU 51 "Read DTCs" (and wait for the Slow Init to finish).
4l. Click ECU 51 "Full Scan".
4m. On Realterm's Capture tab, figure out where the .log file has been created.
4n. Click "Close Realterm".
4o. Copy contents of that .log file into the "Paste..." box of the kwpd3b0_interpreter.html page.

5. Another way to send the ECU 51 (COMBI) A50601 request to monitor whether the SERVICE light should be on or not, even when the bulb is burned out, involves interactively issuing the A50601 request using an ELM327 compatible device and a terminal emulator. The list of ELM327 commands and KWPD3B0 requests, along with additional instructions of when to turn the engine on, involves:

5a. Issue these ELM327 commands:

ATZ
ATL1
ATE1 or ATE0 (depending on your terminal emulator and the ATPP 09 value for your ELM327)
ATI
ATSP 3
ATDP
ATH1
ATAL
ATKW0
ATTA 13
ATRA 13
ATIIA 51
ATWM 82 51 13 A1
ATSH 84 51 13

5b. Turn engine on (and wait about 5-10 seconds).

5c. Issue the following KWPD3B0 request:

A50601

5d. See if the response to A50601 is "84 13 51 E5 06 7B 4E" or "84 13 51 E5 06 73 46" or some other response. The first two responses are explained in item 2 above. Any other response can be interpreted by kwpd3b0_interpreter.html.

6. Also, you could display the B903, B904, B905, B906, B907, B908, B909, and B90A values, all of which are used in determining if the SERVICE light should illuminate. Their values will almost always help in determining if the SERVICE light should be on during the first 2 minutes after engine on.

See elm327_reads_volvo_850_svc70_mileage.html for some details on how the B9xx requests can be used to determine the SERVICE Reminder Intervals' counters and limits and whether one of the limits is exceeded, thereby latching the SERVICE on flag so the SERVICE light will come on during the first 2 minutes.

7. I think the main ways to truly reset the '96-'98 850/S70/V70/etc SERVICE light are:

a. Issue the ECU 51 B030 / B230 request combo, like both volvo850diag and the AST 5408 Volvo SRI Reset Tool do, and (presumably) the DRAPER Reset Tool, Vol-FCR Full, VST (Volvo Scan Tool / Volvo System Tester), or VST equivalents do.

b. Replace the Instrument Cluster with another one that doesn't have any of the ECU 51 B9xx SERVICE Reminder Interval counters exhausting their corresponding limits.

Then there's the ways to "mask" the SERVICE light state:

c. SERVICE bulb burns out (which could happen most anytime during these old cars).

d. SERVICE bulb is removed from the Instrument Cluster (followed by an optional smashing with the feet or a hammer).

e. Covering with black tape an otherwise functioning SERVICE bulb.

esl_97_850_T5

I have to ask, what do you do for a living and where did you get such knowledge. Very indepth, if it is correct.

[esl_97_850_T5]

mecheng,

> ...what do you do for a living...

My primary applicable work experience was: Software engineer. I'm now retired, so I have lots of time.

> ...where did you get such knowledge...

Experimentation -- lots of it -- and researching on the net.

See elm327_why_i_use.html for why I got started doing it.

See volvo_850_diag_links.html for some links which have been crucial for me. I especially should mention the rkam post, since it helped me immensely to understand the keyword D3 B0 protocol, and hence provided a framework for the experimentation.

> ...if it is correct.

Not everything on the jonesrh Volvo 850 OBDII portal is correct. Sometimes I make wrong inferences, since I don't have direct access to Volvo literature and my personal experience base is a single '97 850 Volvo T5, along with the very small amount of data which has been submitted to kwpd3b0_interpreter.html, along with whatever I can infer from others experience/info posted on the net. But when I do find something is inaccurate, I eventually correct it.

I try to provide enough detailed instructions so others can verify if the information is correct or not.

To my knowledge, everything in the post you quoted is correct -- at least for my car and I strongly suspect for any '96-'97 850. So it should be valid for ThommyKent's '97 850 T5. However, 3 things should be noted:

- The 7B and 73 cases that I cited were the most commonly seen in an 850 that has the engine running. Other somewhat common cases are 7F -- which can be interpreted as "Engine on, SERVICE light should be on during first 2 minutes after engine on (if bulb not burnt out), LOW FUEL light should be on (if not burnt out)" -- and 77 -- which can be interpreted as "Engine on, SERVICE light off, LOW FUEL light should be on (if not burnt out)".

- All 4 of those cases apply to 850, not S/V/C70.

- The main thing as far as testing the "SERVICE light should be on/off" flag is to look at bit 3.

I seem to recall that you have a '98 S70. For the '98 S70/V70/etc, the ECU 51 A50601 and A50701 status byte's bit 3 still works the same. It mimics when the SERVICE light should be on and off -- 1 = on, 0 = off. But the status byte containing that bit has different value for the SVC70 cars, and I don't yet understand it as well as I'd like. The SVC70 cars definitely have different values than the 7B and 73 that I originally posted for ThommyKent's car and the 7F and 77 values just mentioned in this post. It's not clear to me what they would be for your '98 S70. But for your S70, you could still test the "SERVICE light should be on/off" flag by simply decoding the status byte into its individual bits and looking at bit 3 where bit 3 = 1 equals 8 decimal.

Some of the kwpd3b0_interpreter interpretation's of that status byte for the SVC70 are likely incorrect. It seems that some of those status bits have a reverse polarity in the SVC70 cars vs. what they have in the 850 cars. I need more people to submit volvo850diag "Full Scan All ECUs" scans or ECU 51 "Full Scan" scans for '98 S70/V70/C70/XC70 to kwpd3b0_interpreter (while also noting the state of the dash, etc, and providing me an e-mail or PM contact to get my questions answered) before the interpreter works like it should for the SVC70 cars.

To emphasize my point that lots of experimentation is involved, here's a couple of recent examples of experimentation because I couldn't find the info anywhere on the net...

Experimentation Example #1 -- How does B90A interact with SERVICE Reminder Intervals?

I've been doing some detailed testing for the past few months to see how ECU 51 B90A (the "Manipulation" Engine Hours per 1000 miles or 1000 km that is used to estimate "lost" miles when the ABS failed to deliver a speed signal) interacts with the SERVICE Reminder Interval counters and limits to cause the SERVICE light to trigger before the intervals themselves would normally get triggered. I was able to see from some posts about Brick-Diag Full or Brick-Diag Free at ecuproject.com and from the Brick-Diag Free v0.0.6.6 program's screens itself that "Manipulation (h)" was listed in hours. However, I wasn't able to discern if B90A was the "Manipulation (h)" value or not. By a process of elimination, I had come to the conclusion that B90A was most likely the Manipulation value, but I originally thought B90A was months, not hours. During previous experimentation, I learned how to change the Mileage limit with the B805xx request. This had not been published anywhere on the net (to my knowledge), but I made an intelligent guess as to how to do it based on rkam's post, since he said the B8 request was used to write values (which I had been reading for over 1 year with B9 requests). Recently, I've done some experiments with the Mileage limit as B905=1 (ie, 1000 miles for my car), B905=2 (ie, 2000 miles), and B905=3 (ie, 3000 miles). And I'm presently repeating the 2000 mile test, since I discovered the formula used to trigger at 3000 miles was slightly different from the one used at 1000 miles.

Experimentation Example #2 -- KWPD3B0 ECU 51 B031 will make instrument gauges swing back and forth thru most of their full range.

After coming close to finishing Experimentation Example #1 and thinking that I understood the formula used, I began to want to edit the B90A value to a lesser value. And if that was not possible, then I wanted to clear the B90A value back to the normal zero value.

I first tried to change my ECU 51 B90A value using B80Axx, where xx is some hex value. I previously had been successful at changing the Tire Size in B90E using a B80Exx request, so I was hoping the value in B90A -- the one that contains the "Manipulation Hours per 1000 Mi (or Km)" -- could be changed similarly. But alas such is not the case. I couldn't edit it and I also couldn't clear it with B80A00. Both attempts generated the "7F B8 12" error that implies this is not a possible ECU 51 B8 subrequest. I tried B80Axx with the proper ATSH command, B80Axxyy with its proper ATSH command, B80A with its proper ATSH command. None of them worked, for varying values of xx. Bottom line: ECU 51 B90A doesn't seem to be able to be written with some ECU 51 B80A variant.

I strongly suspect that B90A can be cleared and that it probably requires some sort of ECU 51 (COMBI) B0xx / B2xx sequence or just a simple B0xx request.

After remembering that B030 / B230 performed the resetting of the SERVICE light, I decided to experiment with B031 /B231 to see if it would clear the B90A value (since its index was just 1 higher and I figured doing the SERVICE light reset was more important that B90A clearing).

Voila! Guess what happened. The instrument gauges (except the Turbo Boost gauge) swung back and forth thru (most of) their entire range. Surprise!

Further experimentation with B031 all by itself, without also using B231, seems to show that B031 all by itself will swing the gauges.

It didn't accomplish the intended clearing of B90A, but I did discover how to perform a new function that, to my knowledge, has not been documented anywhere on the net.

So that completes my 2nd case of experimentation.

For the time being, I have *not* been willing to test any of the other B0xx / B2xx sequences other than the one that I did already -- the B031, that I just mentioned causes the cluster dials to swing back and forth once or twice. I'm afraid I'll screw up my COMBI and make it unusable. That will have to wait for: a) a trip to the dealer, or b) someone monitoring a VST or VST-equivalent with an OBDII Y cable with one leg of the Y connected to my OBDLink SX which has been updated with firmware which can both recognize the ISO 9141-2 style Slow Init performed by the COMBI as well as the immediately following bytes when the OBDLink SX is in ATMA mode which has been preceded by all the ELM327 setup commands for KWPD3B0 communication, or c) an infusion of boldness in which I recklessly experiment with some (or maybe all) of the B0xx / B2xx possibilities until I find the one that clears the B90A value, or d) I could just do nothing, since my SERVICE light is actually already burnt out and that suits me just fine, other than I would like to complete the testing of the B90A interaction with the triggering of the SERVICE light flag by first lessening my B90A value, performing some tests, then finally setting my B90A value to zero, and performing some final tests.

esl_97_850_T5

[Blockpartie]

Gauge test procedure is well documented on the net. Though did only the needles of the gauges move? That would indicate a non complete test procedure or rather a failing command from your side.
Turbo gauge can not be tested this way as it is purely mechanical and relies on engine vacuum or boost for actuation.

Happy to see my nomencalture for "isn't a VST but works like one" being used