O2 sensor/alternator/regulator mismatch

[TrevorB]

I wanted to start this post because the issue has been raised elsewhere but a specific topic does not appear to have been raised, well not to my knowledge.

If this is in the wrong area I'm sure one of the admins will move it to the correct space.

As you may have read on Facebook I have finally converted 'Merlot' to closed loop by purchasing an O2 sensor from Rock Auto. Conversion went surprisingly easy as the original 32 year old plug came out without too much hassle albeit needing a strong man with the right equipment to do it.

Now, I have the issue which has been raised in other topics not specific to this problem. Hence the reason for new topic to cover this issue.

When I connected the O2 sensor and started the engine the battery light remained on. I now know that this is related to the alternator and the regulator that is within. My alternator was replaced in 2020 with an alternator from SJS part no. SJ100E0010. The original alternator was not able to be repaired with the repair kit which was supplied in original group buy. I have disconnected the plug to the O2 sensor.

What I would like this post to explain is the reason why the regulator needs changing, I know that will mean technical explanations which will go over my head but I hope it will help any future members if they also experience this fault. e.g. What is the difference between the regulator in the alternator now and what is I believe the recommended replacement (HC-cargo 134447 regulator).

Information on what I need to do to change it with details of parts and will it be compatible to the alternator I have in place. Obviously the alternator will have to come out and a replacement regulator put in. Joao has already directed me towards his video on YouTube.

I also wonder if SJS is aware of this issue and if not do they need to be made aware?

Hope it makes sense and the excellent knowledge within this Forum will come up with the answers

Many thanks

Trevor

[Tuga2112]

Trevor.

im currently at work, so your getting the very short version of the technical explanation.

the wire that the lighbulb in your instrument cluster uses a yellow wire with brown trace (or vice versa, memory escapes me) in order to complete the circuit and tell you when the alternator isnt working
when this yellow/brown wire is at +12v the light is off.
when the yellow brown wire is at a low voltage, the light comes on.

the yellow/brown wire gets its +12v current directly from the regulator in the alternator.

the O2 sensor heating element is powered by the exact same yellow/brown wire.

when you add the drain from the heating element into that wire, what happens is that if the regulator doesnt put enough energy into the wire, the wire itself gets a low voltage, which brings the light on (not to mention the heating element only partially operating)

you need a regulator that is compatible with the alternator gubbins as well as capable of putting enough power into that wire so that both the heating element works. as well as enough surplus in order to keep the wire voltage high in order to not switch on the lighbulb.

[chrism]

...I agree with Tuga, and I presume that now you've disconnected the sensor the light goes out, as normal, when the engine runs?

I've found that a lot of very relevant stuff in LEC is often hidden away or scattered in other posts. Yes, this should be common knowledge; I sympathise cos I bought a Bosch O2 sensor from RA and the lead wouldn't reach the other Bosch one they offered did. Another 'known' problem is with one of the rear calipers offered by RA, I gather.



I've just looked at the SJS site and they seem to be advising that their alternators/regulators don't support O2 sensors.

[RonR]

The reason for the voltage drop which causes the light to come on is all to do with power supply and power consumption:

The light bulb is rated for around 2 Watts at 12 Volts, which consumes 1/6 of an Amp current (P=IV), so it has a resistance of 72 Ohms (V=IR).
The Bosch O2 sensor has been measured at a resistance of around 4 Ohms, so consumes 3 Amps current, or 36 Watts.
The Denso I think is nearer 10 Ohms, so 1.2 Amps or 14.4 Watts.
When adding an O2 sensor the two are connected in parallel, so demanding somewhere between 16.4 and 38 Watts from the regulator.

Most regulator warning light circuits are designed to just run the light bulb, so there wouldn't be any reason for it to supply much more than 2 Watts plus a safety margin. However our friends at Lotus (or Isuzu?) specified a regulator which was man enough to supply enough power to run the O2 sensor heater, meaning extra electronics wouldn't be needed and "Adding Lightness!".

When you try to "pull" current from a regulator which can't supply it, the regulator voltage collapses and the light comes on as it's connected between the regulator warning light output and the 12V positive IGN supply. Also the O2 sensor can't draw enough power to heat up as quickly as it should (not sure if it ever actually reaches the correct temperature).

20+ years later someone produces a replacement Alternator, tries it in an SE and it works because there's no O2 sensor. BUT if you have an S2 or have gone closed loop it's a different story.

Unfortunately the two types of regulator look physically identical and have the same terminal names, the only way of telling them apart is by the part numbers, but if looking for a replacement regulator you can usually tell the correct high-output one as it has something like "L terminal supplies Choke" in the description.

FYI: The original Denso regulator was part number 126000-0440

[dapinky]

Not sure if it helps at all, as I have no idea where Steve gets his 'alternative' alternators from, or which model he buys......

...... but the one I bought from Rock Auto last year (A Remy 14971 @ £55 + VAT/Shipping) was from their Isuzu Impulse listing, and it works fine with a closed-loop system using a Bosch 15701 O2 sensor.

[chrism]

I like your Maths, RonR, so much so that I've got a spare used Denso 'on the shelf' and a Walker O2 sensor in my m100 and I'm going to measure their resistances tomorrow to find out which is highest - with a view to installing it in the car to give the alternator less work to do.

I remember measuring the Bosch 15701 (RA) at ~4ohms and the Denso (SJS) nearer 12ohms a while back thinking the Denso was faulty since it was so high. I didn't measure the Walker assuming it would be okay cos it was brand new.

I now have a suspicion that my Walker (blue) is low resistance/high power because it seems to heat up immediately compared to the Denso (black):-



(ps I know I'm being a tedious anorak but I think P=IxV should be W=IxV cos I reckon P is the symbol for momentum...)

I'll post those resistance values tomorrow as an edit and say which one I'm going to install/leave in place.

edit:- Walker 2.5ohms, Denso 13ohms ie 58W/11W I reckon as soon as I can get a sensor removal tool I'll swap them over...

[chrism]

Apologies for skewing the original post, but there's a very strong connection here - the likes of which I wish I knew when I was '1499-ing' my '90 SE a couple of years ago (and folks are still queuing up to take advantage of RonR's excellent dual chip upgrades)

I re-installed my old (on-the-shelf) 13ohm Denso sensor and took a stationary-from-cold EScan to find that it maintained the a:f ratio at 14.7:1 from when it 'sparked up' properly around 90s after start-up , showing a regular scope pattern. So I'm leaving it in...

Incidently, when I first took it out I thought it was trashed and washed it in paint thinners, brushed it, sprayed it with carb cleaner and did all those things that doom laden, sharp-intake-of-breath people say that you shouldn't do...and it still works okay

I hope this helps aspiring SE->1499 upgraders to vastly improve their fuel cons and emissions figures. (Imo it'd be useful if the resistance of these items could be marked on the packaging!)

[Tuga2112]

chris.

your last 2 posts have only confused me.

can you please give a summary of what you are trying to tell us ?

[chrism]

...trying not to cause confusion but hoping that 1499-upgraders from SE (aren't you doing a video of this?) don't show the same naivety as me in assuming that all lower-resistance O2 sensors are 'better' and that a 13ohm sensor is not 'broken' wrt a 2.5ohm sensor. 'Highers' might heat up a little slower, but work just as well in spite of a 90s warm-up (as I attempted to show)...

When upgrading to 1499 we have to 'grab what's going' sensor-wise (their resistance is not labelled) and bearing in mind RonRs spread of wattage, plus having had a blown 3A fuse in a fuel-related VSV circuit (due to prev owner 'improvements' before I 1499'd it), then perhaps we ought to be aware of not overtaxing the alternator since its output is needed (certainly for SE->1499 upgrade) to control our fuelling in the way you mentioned earlier in the thread.

ps I don't know if S2 O2 sensors work in the same way but it might be relevant, too. I hope someone will inform us 'don't knows'. If they have a different current supply then that'd be of interest to SE upgraders?

Sorry, not intending to confuse; just hoping to gain and spread some insight

[RonR]

Thanks Chris for confirming the resistances, so it looks like the spread of power requirements is actually wider than I thought!

All turbocharged M100 ECUs are identical apart from the contents of the Memcal's eprom chip no matter which version you have. The closed loop conversion uses exactly the same code as the S2 and 49-States USA cars (California is slightly different).

As far as I know all M100s started out with an alternator regulator capable of running an O2 sensor, but replacement alternators may have been fitted with the 'wrong' regulator, which would work fine in an SE running open loop. BTW Simon's rebuild kits have the correct regulator, so if you've rebuilt yours with one of his kits you should be OK to go closed loop.

Anyway, as far as engine management goes there's no difference between a UK SE converted to run closed loop and an S2 or USA car. Of course the converted car won't have a catalytic converter or any of the S2 'goodies', but ECU-wise they're electrically identical.

BTW - I just looked it up in my Electronics textbook, which says "P=VI" - However it WAS published in 1980, and so much has changed in the Physics world I wouldn't be surprised if P has morphed into W in the last 44 years.

[Tuga2112]

Chris.

i really dont follow.

1 - as you mentioned. we have no control over the 02 sensors resistance until we have them at hand and can test.
2 - we know from experience from now 2 confirmed cases (mine and trevor) that new alternators from SJS cannot operate in closed loop converted SE's
3 - simonP did the research about the repair kits, and its been confirmed now in both my alternators, that the regulator from car-go mentioned in trevor's post (which RonR told me in a PM and i forwarded to Trevor in a facebook message) will work with the cheapest o2 sensor from Rockauto i coudl get ( ULTRA-POWER 2343014 )

i understand your focus on the internal resistance of the o2 sensors. which may potentially. lead us to find a specific unit that can operate without the voltage regulator requirement (i personally find that very unlikely, but its certianly not impossible), but that would require an individual to have a un-modified SJS alternator. and a collection of o2 sensors at hand to test.

i have no idea why your 3a (A10?) blown fuse is relevant to this topic. that fuse is not connected the wire from the regulator to the O2 sensor. my alternator didnt blow any fuse, and trevor has not reported anything.

we also know from Pinky's experience, a cheaper alternator from rockauto has a compatible voltage regulator with whatever o2 sensor from rockauto he has installed in Angie's car.

the SE upgrade is the same as the S2 instalation. when you read the electrical suplement you can see how the conversion is effectively mimiking the suplement details ( despite the fact their vague)

I think your adding a level of complexity to a topic that doesnt require it.

[chrism]

RonR - Truly sorry; you're absolutely right and I stand corrected: Capital P for power in watts, lower case p for momentum...

And Tuga, I think I'm getting that we need to care less about the wattage consumed by the O2 sensor and more about making sure that if/when we replace our alternator we must ensure that its rectifier is higher output by asking if "L terminal supplies Choke" etc before expecting to run closed loop.

I'm wondering why there wasn't a separate, less complicated 12V feed for the O2 sensor but I suppose our current system design of using the charge light circuit is one way of ensuring that we (should) get enough heater current, as soon as it's needed, once the engine is fired up. (That turns off when the engine does.)

Anyway I've got a spare Denso O2 sensor to lend now if anyone gets 'stuck' - and it seems it doesn't really matter if it's run under a tap to clean it

[Simon_P]

I now have a suspicion that my Walker (blue) is low resistance/high power because it seems to heat up immediately compared to the Denso (black):-

Interesting! I'd always assumed that they'd all be the same.

(ps I know I'm being a tedious anorak but I think P=IxV should be W=IxV cos I reckon P is the symbol for momentum...)

Nothing wrong with being an anorak ..... P= Power W=watts the symbol is P the unit is W (J/s),... but the symbol for work done is W (J) so P=W/t also - I was helping my Niece with her physics revision and I'm sure it much more confusing than it used to be.... p = momentum in kg m/s (and yes I did have to look it up!) (I see you have also commented)

your last 2 posts have only confused me.

Easily done!

I think your adding a level of complexity to a topic that doesn't require it.

I don't think that you have understood its relevance. Chris and Ron have pointed out that O2 sensor heaters aren't all the same - put simply the current draw is different, and it is the current draw that is causing the light to come on. We can see form Chris' post that the original Denso sensor has 1/5th of the current draw of his other sensor and hence load on the regulator there is a chance that the regulator is correct but the load from the O2 sensor is too high. That's also pertinent to the life of the regulator particularly as they are all now made in China.

I'm wondering why there wasn't a separate, less complicated 12V feed for the O2 sensor

Money - it would be a lot more expensive to put in a switched feed when there is already one a foot away that can do the job.

Trevor - KenS' post here has the technical spec viewtopic.php?f=29&t=27159&hilit=Charge&start=20

[Tuga2112]

I think your adding a level of complexity to a topic that doesn't require it.

I don't think that you have understood its relevance. Chris and Ron have pointed out that O2 sensor heaters aren't all the same - put simply the current draw is different, and it is the current draw that is causing the light to come on. We can see form Chris' post that the original Denso sensor has 1/5th of the current draw of his other sensor and hence load on the regulator there is a chance that the regulator is correct but the load from the O2 sensor is too high. That's also pertinent to the life of the regulator particularly as they are all now made in China.

i understand your focus on the internal resistance of the o2 sensors. which may potentially. lead us to find a specific unit that can operate without the voltage regulator requirement (i personally find that very unlikely, but its certianly not impossible), but that would require an individual to have a un-modified SJS alternator. and a collection of o2 sensors at hand to test.

i think i understood the relevance. aparently this sentence didnt transmit the message clearly.
in my opinion, it seems unlikely that a "non-chocke" regulator would be able to power a O2 heating element.
its not impossible to find out SJS units have a "crap choke regulator"... (i can take mine out and provide details later) which may work with SOME very specific sensors

you gone to the effort of finding a suitable regulator, its a confirmed solution against all sensors tested so far, which makes it ,at least in my opinion, the right solution to this problem.

when i first discussed this with Peter, he suggested i could create a circuit to feed the sensor directly fromthe battery when the brown/yellow wire was at 12v using a relay. i think he's right that it woudl work. and i have the bits in my house to create such circuit . yet i gone in, and spent another 3 hours taking the alternator out again, to put the regulator you sent me years ago.

im trying to preach "do it right, do it once" (lets just pretend no one knows about the regulator plug coming loose on the way back from CC incident)

PS.
i personally hate it when you need to search a part with lazer focus because of some arbitrary condition, thinking in 5 years time someone says "i changed my o2 sensor and now my alternator light is on" and you tell them "you need sensor XYZ" and it turns out "XYZ" is discontinued when we already know the nitty gritty details of the problem and a long term solution

[KenS]

Might be useful for future reference - I have just measured the factory fitted O2 sensor on a brand new old stock 1991 engine, its marked " 48 0185 - AC ROCHESTER MADE IN USA " and the heater measures 3.9 Ohm, confirming that the factory did fit high current O2 sensors. (maniford is currently off while getting the wastegate ported). AC Rochester were a part of GM and taken over by Delphi in 1995, but I cannot find any tech data on the sensor.

o2.jpg

[chrism]

Now that ties it all up quite nicely, KenS...

...ie if you want to run, buy or replace an O2 sensor then don't worry about its wattage - just make sure you've got' the 'super' rectifier in your alternator...

[Tuga2112]

just make sure you've got' the 'super' rectifier in your alternator...

you mean regulator

the rectifier (or diode bridge) is a different component


im sorry to be pedantic. but my experience with the M100 alternator rebuild caused some trauma...

[chrism]

[TrevorB]

RonR visited yesterday and with the aid of a newly installed Elan Scan to my windows11 laptop was able to diagnose some other issues related to my conversion to 'closed loop'. Ron completed some tests with O2 sensor plugged in and came to the conclusion that the wiring was not in place. He did think this unusual as Merlot is a late 1992 SE version which he thought should have had the wiring in place.
Any thoughts on that would be appreciated.
Does anyone have any pictures or diagrams, information on the wiring required. There were some indication that some wiring is present, I'm sure RonR will expand on this. Obviously it needs sorting at some stage.

Back to the alternator. Dave Pink, did you require a longer drive belt when you installed the Rock Auto alternator? I'm thinking this will be a better route than taking out existing, dismantling, replacing regulator and then putting back.

RonR did a thorough check of all the data collected after taking Merlot for a run with Kilimanjaro chip configured for 'open loop' and she passed with flying colours.

Thanks everyone for your invaluable help.



Trevor

[dapinky]

Yes, the belt is a '713' unit, so a little bit longer - easily available in the UK (a variety of sizes will work, but that one gave me the best location of the alternator for reaching the tensioning bolts.)

As far as the wiring for the O2 sensor, all Elans have the same engine bay wiring loom, so the part where the sensor plugs in will always be present.....however, there is a small wiring loom (only 2 wires) which connect from that engine bay loom to the ECU, for the sensor - it is these 2 wires which are not always present - it all depends on what they had in stock on the day of assembly. There is no specific 'age' of cars which do or don't have them.

It is a matter of running 2 cables from the engine bay (down by where the gearcables run) to the ECU - they are only about a foot long, and connect to the ECU terminals.

I have kept it a bit vague, as I don't expect you to be doing the job yourself, and whoever gets the task will be able to see what is required themselves (but I can try to explain further if you need it!).... to fully explain would take longer than I have at the moment, and won't help you personally!!!!

Nick Walker had to do it on his car, and it was only 30 minutes work (I'm sure Ron could do it in the same sort of time)...... I've never needed to do it, as all 3 of mine have had the 'full' loom fitted.