If you buy an older luxury car there are two things near certain: the first
is that it will have electric seats, and the second is that at least one of the
seat functions won't work! So how hard is it to fix a defective leccy seat?
Obviously it depends a lot on what the actual problem is and the car in
question, but as a guide let's have a look at fixing the seats in an E23 1985
BMW 735i. The seat architecture in other cars will be different, but if you
don't have any idea where you'd even start to fix such a problem, this story is
sure to be useful to you.
The front seats in the BMW are amongst the most complex that
you'll find in any older car. They have electric adjustment for front/back
travel, front of the seat up/down, rear of the seat up/down, head restraint
up/down and backrest rake forwards/backwards. However, they don't have electric
lumbar adjust and they don't have airbags. (If the seats that you are working on
have airbags, you must read the factory workshop manual to ascertain the safe
procedure for working on the seats.)
The seat functions are all controlled by this complex
switchgear, which is duplicated on the passenger side of the car. As can be seen
here, the driver's seat also has three position memories. Incidentally, the rear
seat is also electric, with an individual reclining function for each side! But
in this car, the back seat was working just fine.
The driver's seat had three problems.
- The button which moved the seat rearwards didn't work. However, the
seat could be moved backwards with one of the memory keys.
- The front
of the seat couldn't be raised.
- The head restraint wouldn't move up
or down, although in this case the motor could be heard whirring uselessly
whenever the right buttons were pressed.
Getting the Seat Out
The first step was to remove the seat from the car so that
access to all the bits could be gained. The seat was electrically moved forward
and then the two rear floor-mounting bolts undone.
But how was access going to be gained to the front mounting
bolts? Pressing the adjustment button didn't cause the seat to move backwards,
and by this stage the memory button had stopped allowing that action as
well! The answer was to manually apply power to the seat to activate the
motor. All the connecting plugs were undone and those plugs containing the
heaviest cables inspected. (There will be wiring for seat position transducers
and things like that in the loom, but the motors will be powered by noticeably
Using a heavy duty, over-current protected, 12V power supply
(this one was made very cheaply - see DIY Budget 12-volt Bench Supply
power was applied to pairs of terminals connecting to the thick wires until the
right connections were found. The seat was then powered backwards until the
front mounting bolts could be accessed. These were removed and then the seat
moved forward until it sat in the middle of its tracks, making it easier to get
out of the car.
Fixing the Head Restraint
This is what the BMW seat looks like underneath. Four
electric motors can be seen, plus there's a fifth inside the backrest. Each
electric motor connects to a sheathed, flexible drive cable that in turn
connects to a reduction gearbox. As I later discovered, inside each gearbox is a
worm that drives a plastic gearwheel, which in turn drives a pinion operating on
a rack. At this stage, though, a simple test could be made of each motor by
connecting power to its wiring plug and making sure that the function worked as
it should. Every function but the head restraint up/down worked, so the problems
other than the head restraint showed that they must be in the switches, not the
motors or associated drive systems. But how to fix the head restraint
The rear trim panel of the seat came off by the simple
undoing of four screws. As with the other seat motors, the mechanism consisted
of a brush-type DC motor driving a flexible cable that went to the adjust
mechanism. The motor worked fine with power connected, but the head restraint
didn't move. Feeling the outside of the drive cable sheath indicated that the
drive cable inside was turning, so the problem must lie in the mechanism closest
to the head restraint itself.
The adjustment mechanism was held in place with one screw,
which was accessible with the leather upholstery disengaged from small metal
spikes that held it in place. The legs of the head restraint clipped into
plastic cups on the mechanism (one is arrowed here) and these were able to be
popped out with the careful use of a screwdriver.
The whole upper part of the adjustment mechanism was then
able to be lifted out of the seat back and placed next to the seat. Note that
the electric motor stayed in place - it didn't need to be removed as well.
To see what was going on inside the unit, it needed to be
pulled apart. It was obviously never designed to be repairable, and so the first
disassembly step involved drilling out the rivets which held the plastic sliders
in place on their track. With these out, the action of the pinion (a small gear)
on the rack (a toothed metal strip) could be assessed. Neither looked
particularly worn and applying power to the motor showed that in fact the pinion
wasn't turning. So that meant that the problem was inside the gearbox
The gearbox was held together with four screws, each with
an oddly-shaped internal socket head for which I don't have a tool. However, knowing that
I could always find replacement small bolts, I used a pair of Vicegrips to undo
them - that is, it didn't matter if they got a bit mutilated in the process of
Inside the gearbox the worm drive and its associated
plastic gear could be seen. Initially I thought that the plastic cog must have
stripped, but inspection showed that this wasn't the case. So why wasn't drive
getting out of the gearbox? Again I applied power to the motor and watched what
happened. What I found was although the cable could be heard rotating inside its
sheath, that drive wasn't getting to the worm. Pulling the worm gear out and
inspecting the square-section drive cable showed that the end of the cable was a
little worn and it was slipping back out of the drive hole of the worm. (The
slippage was occurring inside the area marked by the arrow.)
The fix was dead-easy - simply pull the drive cable out of
the sheath a little, crimp a spring steel washer on it (backed by a plain washer
that here is out of sight - it's fallen back into the mouth of the sheath) and
then push the drive cable back down in its sleeve. With the crimped washer
preventing the worn part of the cable from sliding back out of the square drive
recess in the worm, drive was restored to the gearbox.
The mechanism could then be reassembled. New screws were
used to replace the Vicegripped ones, while the drilled-out rivets were also
replaced with new screws and nuts (arrowed). The gearbox was re-greased before
assembly and a smear of grease was placed on the tracks that the nylon sleeves
run on. Back in the seat, the mechanism was checked by applying power - and
So in this case the fix cost nearly nothing, except some time.
Since all the motors had now been proved to be in working
order, fixing the electric rearwards travel and front up/down motion could only
be achieved with the seat back in the car - it looked as if it had to be a
wiring loom or switchgear problem. But while the seat was out, it made sense to
wipe over all the tracks and exposed cogs and re-grease them.
Fixing the Rest
Under the driver's seat is a control module containing both
relays and the seat memory facility. Close inspection of the plugs and sockets
on both the unit and the associated loom showed that some corrosion had
occurred. (Perhaps at some stage a drink had been spilled on it.) The corrosion
showed itself as a green deposit on the pins and some tedious but careful
scraping with a small flat-bladed screwdriver removed it. Once that was done,
the associated plug was inserted and pulled out 20-30 times to scrape off the
deposit inside the pins of the plug, which were otherwise impossible to access
And after that was done, all the seat functions again worked!
At commercial rates, fixing the seat would have cost hundreds of dollars -
both in labour time and in a complete replacement head restraint up/down
mechanism. No one would have bothered repairing the gearbox drive - they'd have
just replaced the whole thing. The corroded pins? That would have been cheaper,
but the total bill would have still been prohibitive.
So if you have a car with an electric seat problem, it's certainly worth
taking a long hard look yourself - it might be something impossible to easily
remedy (eg a stripped gear) but it might also be a problem as easy to fix as
this one proved to be....
Restoring the Bounce
While the seat was out of the car, I tried something which I thought might
just work - and it was so successful it's worth covering here.
Compared with the front passenger seat, the driver's seat
had noticeably sagged - the base was flatter and less supportive. Inspection
showed that the seat used a simple series of coil springs, rather than the foam
rubber more usually employed in modern cars. Replacing these springs would have
been an immense job, but adding some helper springiness was dead-easy.
The rubber that was used to add some support was this high
density pipe foam insulation - it's usually found slid over the pipes coming
from domestic hot water heaters. It's available from hardware stores.
The rubber cylinder was manoeuvred into the seat base and
then slid between the adjacent coils of a row of springs, being held in place by
the springiness of the rubber and the coils. A second cylinder was then placed
in the other row of springs that supports the seat base.
The added support is quite noticeable, without making the
seat uncomfortable. In fact, the driver's seat now feels very much like the
passenger seat - which was the aim.