Alfa Romeo Forum banner

MoT fail: ABS lamp not illuminating on startup

11K views 55 replies 12 participants last post by  279666 
#1 ·
Apologies for doubling up on a couple of the posts below: but I've hit a snag with the '96 Phase 1 Spider I've been getting back on the road.

I put it through its MoT on Friday and was pleasantly surprised that it failed on two points only: handbrake adjustment and ABS warning lamp not illuminating when switching on the ignition.

On the ABS warning light, I've pulled the speedo and checked that all the LEDs are present (they are) and have checked the resistance is the same on the ABS warning light as on the handbrake & pad wear LEDs (it is). So I'm assuming the lamp itself is OK.

Reading the below posts, it seems that the lamp failing to light can be caused by an ABS ECU fault.

However, I can't test the ABS with AlfaOBD or MES as I have the old system (Bosch 2Si - pic below) and these tools only support the later Bosch 5.3 ABS.

Can anyone advise on next steps to diagnose this problem? And is it possible to replace the ECU without replacing the entire ABS unit?

It would be a major pain for this to be the only thing standing in the way of an MoT pass for this car!!!

Other posts on this subject:

http://www.alfaowner.com/Forum/alfa...738361-spider-abs-light-does-not-come-on.html

http://www.alfaowner.com/Forum/alfa-gtv-andamp-916-spider/966450-abs-light.html
 

Attachments

See less See more
1
#2 · (Edited)
I bought my project car with exactly the same fault (knew when I got it) as this, I'm afraid I can't give the answer to the fault yet as I've not got to this point in the resto yet. Unfortunately it's bit of a nightmare with this particular ABS unit as it was only fitted for a short period and only to the GTV/Spider and Fiat Barchetta. To compound the misery all of the research I've done shows the ECU can't be fixed by the usual experts and it really is throw it away time.

There is a way round (bodge) it to pass an MOT and making it appear the ABS unit is working correctly to a tester but I won't elaborate as you still won't have a working ABS system, I'm fully intending fixing mine!!

The ECU does come away from the ABS unit fairly easily, not a tricky job but.... The previous owner of my car did exactly this and blew up the new ECU straight away, given their rarity now I'd strongly advise changing the whole ABS unit as it looks likely the fault will not be in the ECU board itself.

I've done a very small bit of fault finding on mine (checking insulation resistance on ABS sensors and car wiring) and haven't found any short circuits yet so it is looking increasingly likely it is the internals of the ABS unit at fault. That said I'm intending to really pull all of the ABS cabling apart before powering the ABS up, got a couple of spare 2SI but I can honestly see this being a real problem for early Ph 1 cars in the future.

Did look at converting to 5.3 but it isn't as straightforward as you'd think, good luck anyway
 
#6 ·
Thanks all for your advice - really helpful, as always.

I've been looking through the workshop manual over the last couple of days and have found wiring diagrams, connector pinouts and test procedures for the ABS unit. This will at least give me the chance to check the wiring through from the dash warning lamp to the ABS unit - and if it works OK I can try fault-finding via the "flashing codes"...

If anyone's already tried this, I'd love to hear how they got on.

The ECU does come away from the ABS unit fairly easily, not a tricky job but.... The previous owner of my car did exactly this and blew up the new ECU straight away, given their rarity now I'd strongly advise changing the whole ABS unit as it looks likely the fault will not be in the ECU board itself.
If the whole unit is unresponsive and needs replacing, has anyone done this? I'm worried by the fact the manual says that if air enters the ABS unit, "any kind of air bleeding is impossible"...

Finally: does anyone have a 2SI unit in case I need one? wal5 - sounds like you might have!
 
#4 ·
Have you checked that some previous owner hasn't put insulating tape or similar over the w/l to hide a fault condition?

If not, test the LED

( as you've measured resistances, I'll suggest this test ).

Remove the plastic cover at the back of the instrument pack and identify the two wires coming through the board from the ABS LED. There'll be three small rectangular devices ( resistors ) next to the leads.

Get a 3v button cell. Must be 3 volt - nothing more, nothing less. Nick one temporarily from the kitchen scales or somewhere. Carefully bridge the two LED wires with the button cell battery. If the LED doesn't light, try the cell the other way around. The + side of the cell probably needs to be towards the multiway connector.

If the LED lights, you then need to trace out the wire between the instrument pack and the ABS connector in the engine bay. If you have continuity there, you're probably looking at the ABS unit.
 
#9 ·
Removing the Abs from the car is legal and will pass the mot. I know one chap who has done this very successfully. The block to join all the pipe work in place of the abs unit looks like it should be there. I'll try to find a pic.

Sent from my SM-G930F using Tapatalk
 
#17 ·
I collected a spare unit from Corkyd10 on Friday and I switched over the ECUs - unfortunately, this didn't fix the problem.

Fortunately, I've located another spare from eBay (a Lancia version; same Bosch part number) for £35 which is on its way.

So a question: as per wal5's post earlier - could a fault in the ABS hydraulic unit damage an otherwise healthy ECU?

And therefore when I get the new unit, should I again switch the ECU only - or the entire unit?
 
#20 ·
GrahamL and spankdanny: I've done the following testing so far, based on details from the workshop manual:

- Checked the 60A and 10A fuses on the battery and aux. fusebox - both OK

- Disconnected the main connector to the ABS unit - checked there is a voltage on pin 1 (feed from the dash warning light) - there is; when I earth pin 1 then the warning light illuminates, as you'd expect
- Checked resistance of the four wheel sensors - all are showing approx. 1.6k ohm
- Checked continuity to the diagnostic socket - both good
- Reconnected and attempted the "flash test" by earthing pin 1 of the diagnostic socket and turning on the ignition - no activity from the ABS light

- Disconnected four-pin connector from the ABS ECU and tested all four circuits OK (12V, switched 12V, earth and feed from warning lamp) - all OK
- Earthing the lead from the ABS warning lamp illuminates the lamp in the dash

So, my conclusion is that the vehicle wiring is OK but the ABS is dead.

With the replacement ABS unit: I first switched the ECUs; then in case the hydraulic unit is implicated, I put the ECU back into the replacement unit, and plugged the whole ABS unit in to my car (there is just room to do this with the old unit in situ).

In both cases, I get the same symptoms: no ABS light when turning on the ignition.

Now awaiting the arrival of my second replacement unit...
 
#21 · (Edited)
Glad you managed to get a couple of ECU's but I'd still strongly advise (see my previous post No 2 above) against just swapping the ECU on its own.

If the fault is a short circuit on the electrical element of the hydraulic part of the whole ABS unit or on the ABS system wiring then you'll in all likelihood blow one of the devices on the ECU board (hope you haven't but fear you've already just done this). As soon as you apply power they will pop.

There will be wiring within the ABS unit itself and I'm starting to suspect that this is the relatively common failure on these 2SI units, wish I had a bit more time to spend on mine as I'd do some remote fault finding for you (and try and fix mine as well!!).

Thank you by the way for the précis of your fault finding, filed away for late Summer use on mine!! Good luck
 
#24 ·
I think a lot of phase 1 models have airbag light wired up to ABS light, these ABS units have been going wrong fairly regularly for years. Even when we were breaking a couple hundred GTVs/Spiders a year we never managed to build up a stock of the early phase 1 ABS unit (Bosch number ending 001 seemed the most problematic).
 
#25 · (Edited)
Just out of interest did you ever find someone who could fix the ECU boards for this unit or find a root cause for the 2Si failures? Had a very quick look at disassembling the actual valve block just to see if I could see anything obvious but it all looks understandably "tamper proof".

Although I've got a spare I can see as much as it pains me that I may have to live with a bodge eventually when all my spares are exhausted. Either that or look at doing a Bosch 5.3 transplant which I know isn't easy.
 
#33 · (Edited)
Still waiting for my replacement 2Si ABS unit to arrive, so did a little further testing on my spare (non-functional) unit and took a few pics to assist anyone else... wiring diagram for the unit also posted.

Regarding the link between the ABS ECU and the pump/hydraulic unit... the six pin lead from the ECU to the hydraulic unit carries outbound signals to the hydraulic valves on four wires (green or yellow) from four pins; with the returns (black x4) to the remaining two pins.

Power is sent to the Pump Motor from the Motor Relay via a metal cylinder which is connected to the ECU by a screw. Return to earth is via G202 - the stud and nut at the bottom of the motor casing.

So I tested continuity on the valve circuits and the motor (all good); and applied 12V to the motor (it ran - noisily).

I attempted to remove the relays for testing - but they won't budge. I read somewhere they are soldered to the ECU so I'm reluctant to risk further damage. Yet...

It occurred to me that I could test my new replacement ECU in situ on my car without connecting it to the pump/hydraulic unit, thereby eliminating the risk of damage from that source. I quickly tested my current ECUs this way, but they still didn't work. I wasn't surprised, tbh...

So now, just waiting...
 

Attachments

  • Like
Reactions: wal5
#34 ·
Great bit of work Jeremy and you've saved me a bit of time!!

If I get a chance at the weekend I'll do some insulation resistance checks on the wiring on my (old) spare block and send you the readings, you've got me thinking. I've also had issues with relays (not on cars) before going short circuit and causing an earth, its not blowing the ABS fuse but maybe enough to pop one of the devices on the ECU. I'll also try putting 12V on the coils of the 2 relays and see if you get continuity across the contacts.

A few ideas anyway, just hope I get some time to "have a play".
 
#35 ·
Presumably the PCB with the microcontroller is under the cover where you have the "Power to Pump Motor" text on the "2SiABS02.jpg" pic, so try to remove that cover and get a look under there if possible. Or is it maybe under the cover on the same side as the relays?

Either way, it needs more stripping down to look at the electronics.
 
#36 · (Edited)
The ECU/PCB has been removed in Jeremy's picture no 2 (top right) in the above post, the only electrical parts left on/in the block in the picture are the ABS pump motor and valves plus associated wiring. The PCB is shown fitted in picture 1 and removed from the block in picture 3, it's a pretty "old school" board.

I know you're a whizz with Coupe electrics (I used to be on FC Forum), do you have any obvious ideas we're missing Graham?
 
#37 ·
The electronics must be on the PCB below the relays or on the reverse of that board then. There must be a microcontroller on there somewhere to monitor the wheel sensors, engage the ABS and handle the diagnostic communication etc.

If you have a multimeter it may be worth checking the big diode at the top of the ECU board and possibly that other component (another diode or maybe a PCB mount fuse?) between the 4 big pins at top left.
 
#39 ·
I took a deep breath yesterday and dismantled my spare, non-functional ECU. Here's what I found:

The ECU is comprised of three layers. The bottom layer contains the ECU proper - the integrated circuits. The top layer contains the electrical parts: the connectors, relays, etc. The middle layer interconnects the two.

First of all, I separated the bottom layer from the middle. Not pretty: having got past four rivets and a lot of sealant, I pulled the bottom layer away, tearing off all of the solder wires that link the electronics to the tags on the middle layer above. The electronics are also covered in clear gel.

Suffice to say, if the issues with our ECUs are in the bottom layer then repairs look pretty difficult... see pic below.

I decided to move on to separating the top and middle sections...
 

Attachments

#40 · (Edited)
Separating the top and middle layers of the 2Si ECU: what I should have done is as per pic 1 below.

However, lacking the correct Torx bit for the security screw, I resorted to levering the two sections apart once I'd ground out the rivets and de-soldered the six interconnecting tags - recognising that I was going to damage a small part of the metallic circuitry below.

But, with the correct tool for that screw, the two sections should separate and re-assemble easily - and with no damage.

So, Pic 2 shows what the top and middle sections look like once separated

Pic 3 is a close up of the underside of the top section, indicating the repair I had to make to fix the damage I'd caused above

I then set about tracing the circuits through and matching it up with the circuit diagram (also attached below).

I'm part way through this process - but so far, here's what I've found:

1) The circuitry tallies with the diagram (aside from the connections through the motor relay, which appear to be switched) and continuity is good throughout

2) BUT the components shown on Pic 4 are not shown on the wiring diagram - I'm trying to construct a schematic of where they fit in the diagram. As you can see from Pic 3 I will have to carve away some of the plastic to be 100% clear on the circuitry.

3) The relays appear to be spot-welded to the board - but all connections are good. I'll get some power across the relays to check them today.

Now - I could use some help in identifying the components in Pic 4:

C is obviously a capacitor
D behaves like a capacitor

But for A, B and E I cannot get a reading across any of these in either direction with my multimeter - can anyone help ID these, or advise how to test these further?

Electronics gurus - please step forward!

Will report back on the health of the relays later - and hopefully complete the wiring diagram.

And in the back of my mind is the following thought:

- The ABS warning light is not illuminating.
- For it to light, either Pin 1 on the 15-pin connector or Pin 3 on the 4-pin connector need to be earthed.
- There is no route to earth via the 15-pin connector. The only route to earth for the ECU is via Pin 4 of the 4-pin connector.
- Component A appears to sit between Pins 3 (lamp) and 4 (earth) of the 4-pin connector: if A is faulty, would this prevent the lamp from illuminating?
 

Attachments

#41 ·
Even more great work, pretty sure A,B and E are diodes with the band signifying the cathode. Can you see any other markings on them?

Won't insult your intelligence with how to test them but would have expected a very low resistance reading from anode to cathode, surprised they all seem "wrong" though.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top