No speedo/odo after diff carrier swap

[rogersparknuisance]

Hi everybody, about a month ago I swapped the internal carrier of my 3.73 open diff to a 4.10 LSD that allegedly came from an '87 325is with an automatic. The housing is from my original unit. Everything bolted up nicely and the new diff works great mechanically and is tons of fun.

However, since the swap my speedometer and odometer stopped working. It would work very intermittently if I had been driving for a while at speeds over 65 when accelerating, but still bounce a lot and maybe give half the actual speed reading and drop when letting off the gas.

My understanding of the diff speed sensor is that it is a magnetic reed switch, so it goes between open/continuity as the fins on the axle pass through it. I tested it on a lift and while it fluctuated, it went back and forth between 1 and 1.2 ohms, so I figured the sensor must have coincidentally just gone bad. I ordered a new diff sensor and new wire connectors as the old ones were pretty brittle, installed it and the readings are the same. Looking into the hole with the sensor out I can see the fins are all intact and move with the axle as they should. Are there different size sensors or toothed wheels for these diffs? From my research they all have 9 teeth and are the same size and use the same sensor. Is this not the case? Am I wrong about how the sensor works? I also tested current from the leads with the ignition on and got about 7.5 volts, does that sound wrong, should it be more like 12? Could it be the gauge cluster coincidentally failed?

Any help would be greatly appreciated, I've put probably 1200 miles on it since I've lost the odometer and I'd really like to get it back.

[jeffnhiscars]

I'd try the sensor from the 4.10.

[rogersparknuisance]

I don't have a sensor from the original diff, I only bought the internal carrier and swapped it into my housing. Looking up the part for an 87 325i the only difference appears to be the connector, they appear to be the same size. Is there that much of a difference between these two?

For an '87: https://www.pelicanparts.com/catalog/SuperCat/0227/BMW_0227_ELSENS_pg3.htm#item13
The one I bought for an '89: https://www.pelicanparts.com/catalog...pg4.htm#item14

[jeffnhiscars]

The connector type shouldn't matter as long it's making contact. It has to either be the sensor of the internal wheel. When I swapped out a 2.73lsd for a 3.46 Torsen I had to change sensors

[rogersparknuisance]

I am hesitant to buy another $77+shipping sensor when I already spent $52 on this other non returnable sensor, and then $21+ shipping on new leads that I can't even use with the older style sensor, when they look identical in length.

'87



'89 speed sensor (the one I have currently):



Can anybody speak to the resistance test I did? Is there anything else I can test with what I have currently without throwing another $100 in non returnable parts at it?

[TobyB]

Two things pop to mind:

first, the most common voltage at the sensor should be 5v, followed much later by 12v. 7.5 doesn't seem right... oh, unless it's 12.5 - 5v, I guess.

second, meter across the sensor with an ohmmeter, and turn the wheels- you should see it switch from closed-
0 ohms (1-3 is fine)- to open 9 times per rev.

I am remembering vaguely from some E36 stuff long ago that there are different- length 'fins' on the tone wheels,
but a lot of pints have crossed the lips since then....

I've used early and late sensors interchangeably, without issue. I don't think you have a sensor problem.

t

[rogersparknuisance]

The resistance never indicates an open circuit with either sensor, it fluctuates between 1-1.2 ohms.

The fins on each diff look the same to me as well.
Original open diff that came out of the car:




The LSD that is currently in the car:


Is it possible I have 2 bad sensors, one that coincidentally failed right when I changed the diff and another bad brand new one? I should add that the brand new one doesn't work at all ever, where the original one would work somewhat at high speed after it's been driven for a while.

[TobyB]

Is it possibly an inductive sensor?
Then you'd read current spikes from it as it spun.
They're often very low resistance, as you're reading.
That seems odd, though, as an inductive sensor (like the CPS) usually doesn't work
very well at very low speeds.

How about attacking it a different way- wire up an easy-to-push switch across the
input, get to where you can see the speedo, and start tripping the switch like crazy?

If you're right about the reed switch (and I thought you were) that should simulate
it well enough to give you a speedo reading....

hth

t

[Gregs///M]

The speed sensor is a simple reed switch. You should read near 0 ohms (1-1.5 is OK) when switch is closed and open when the switch is open. The speedometer wants to see a pulse current signal so it does not matter if its 5V, 12V or somewhere in between. If the cluster output is 12v and you are only getting 7V, you have a serious voltage drop. In the end, you want a pulsed signal on the brown/red wire. It sounds like the switch is not opening and closing as the axles move. Maybe pull the sensor and test it outside the car first.

[ForcedFirebird]

Maybe pull the sensor and test it outside the car first.

This. Just hook up your multimeter and wave something ferrous in between the sensor prongs.

Honestly, I have never seen a diff sensor failure. Might want to do a continuity test all the way to the cluster - specially if it was intermittent right after the diff swap. Doesn't take much to break the copper inside the wire insulation on wires that have been exposed to the elements for ~30yrs, and go completely unnoticed on a visual.