ECU repair P1117 O2 heater circuit

Dear all,
Warning - this thread is not for the faint-hearted, it contains details about open heart ECU surgery.

I have been chasing a P1117 ("left side, rear O2 sensor heater high reading") code for a while and have narrowed it down to the ECU (well I think so). I am wondering if any one has delved into the ECU to effect a repair? Or worked out a simple bypass or O2 simulator for the heater circuit.

What I have discovered:
1. The O2 heater is not a critical component, in the old days O2 sensors did not even have heaters. Obviously affects cold start running and emissions though. Otherwise engine runs fine.
2. On the bench, the O2 heater is working fine , I have replaced both front and rear sensors and load tested them with 2A current. Cold static resistance is about 3 ohm. Start up current is 2-3 amp which drops steadily to 1 amp over 1-2 minutes.
3. Harness continuity has been load tested and is fine
4. Relays have been swopped, left to right. The O2 sensors use the fuel pump relay.
5. Fuses are fine
6. ECU connectors were tested, cleaned and inspected. Some slop in the spade-slot connection was remedied. Pin 1 is used for the rear left O2 heater.
7. Measuring the current draw at pin1 of the ECU I see it randomly oscillates from 0.5 to 2.5 amps so the ECU is actively varying the current. The driver side ECU does not do this , the current drops steadily from 2A to 1A as expected.
8. Removing and opening the lid of the ECU, one can trace pin 1 to a transistor BTS132 which is a special sort of MOSFET (TEMPFET) with a built in temperature sensor shutoff. It is a discrete component , and I could easily replace it if failed, but I will try and test it in situ first. Anyone no a good test procedure, I have the datasheet.
9. Everything inside the ECU looks clean , no signs of burning or corrosion

Anyone electronics specialist care to suggest another course of action, maybe a O2 heater simulator? Maybe one for Cribbj?

How the O2 heater circuit works:
12-13V is supplied from the battery via the fuel pump relay to the O2 sensor which is then current controlled and earthed via the BTS 132 in the ECU. I believe there is active feedback control of the current via this transistor (TEMPFET) that has built in temperature control and shutoff (current goes to zero at 160 degrees C).

Seems a shame to throw the whole ECU away for such a small $5 item. Or maybe I have overlooked something.

Regards,
Ron

PS as a by-product of all this ECU connector removal and cleaning, I found that "SLOWDOWN" lights are easy to reproduce if the pin 6 slot is loose.

 

Dear John,
Yes, I have load tested the harness, by bypassing the ECU completely. I cut the brown wire that goes to pin 1 about one inch from the ECU. From this breakpoint, on the sensor side I load tested the harness and sensor with a 1 ohm resistor connected to ground and see a steady 2 amp dropping to 1 amp when warm, sounds good to me. I then load tested the ECU side with a 12v supply and 8 ohms resistor and see the oscillating current.
That reduces it to the connector or the ECU. I then load tested the connector and sensor (no ECU in circuit) and it was fine.
There remains the possibility that the spade on the ECU is still loose in the slot connector but I dont think so, the current is swinging so fast that only an active device like a MOSFET (with current/temperature sensor) could do this rather than a loose connection. Interestingly pin 1 is the last pin to connect as the 88 pin connector is swung into position from the bottom up, so this connection is the most vulnerable.
Relay D powers the heater and yes I have swopped C and D.
Thanks for jumping in,
Ron
PS in your diagram, 2nd fuse from left is the right bank O2 heater (handles both front and rear), you can remove the fuse and insert a current meter to monitor current changes.

 

Dear John and Mark,
My soldering skills are pretty good, I have done a lot of DIY electronic projects. Even high voltage ones - direct drive amplifiers for electrostatic speakers keeping one hand in my pocket!
I did a current load test on the MOSFET (TEMPFET) last night - I used a 6V battery to switch on the gate of the FET and loaded the drain-source with 1.5A and the current was steady as a rock, so the power side of the device looks good.
Leaves me 2 possibilities - the gate logic driver IC is switching the device on and off rapidly (too difficult to solve? capacitance issues?) or the pin 1 slot in the connector is bad, broken or burnt inside. I will replace this connector or just do a hard wire bypass and retry the ECU in the car.
Pin 1 being the last to connect makes me suspicious, the slot opening was a bit on the large side , see picture - top row right corner is pin 1 , - I have now closed it up a bit.
Thanks,
Ron
Image Unavailable, Please Login

 

Dear Jim,
Thanks for your interest. The circuit flow is as follows, 12v from the battery goes to the fuel pump relay, if the ignition is on and engine running then the relay is closed and current flows to the right hand fuel pump and, via the fuse 2, to the two O2 heaters (front and rear). One branch flows to front O2 heater and then to pin 30 of the ECU. The other branch flows to the rear O2 heater and pin 1. Pin 1 and pin 30 each have a separate mosfet that connects the current flow to earth, the on resistance of the mosfet is very low (0.065 ohms in datasheet) so only a small voltage drop to earth across the mosfet.

Thus I would say the mosfet connects the heater current to earth rather than the mosfet drives the heater, just a different perspective. The heater is a resistive load to the mosfet drain. The effective heater resistance (dynamic in circuit) is about 6 ohms when cold (2A current) and greater than 12 ohms (1 A current) when warm. The static cold value is 2.5-3.5 ohms.

Yes, the temperature sensor is just a fail safe device that switches the mosfet off at 150 degrees C (very high current). I was expecting to see a current sense resistor as a feedback control device but there is nothing in the drain source path so I guess the mosfet already does that. The Ferrari high/low heater pcodes are set to trigger at a high current of 2A and low current of 0.2A, so I wonder how the ECU senses that. Also I wonder if the heater is always on at a low current of say 0.5A when warm or whether it switches off entirely.

A mosfet is often used as a variable resistor to control the current flow and has a constant current region of operation. Not sure how the gate circuit is arranged, it could be as simple as a simple switch on voltage of 4-5V.

I have ordered a new BTS132 and will first try a hardwire bypass of the suspicious, pin 1 female connector.
Thanks,
Ron

 

Jim,
I will try some more tests tonight. By the way, the Siemens datasheet I have says Rds=0.055 not 0.55 ohms. The pulse drive possibility is interesting so maybe I should remeasure AC voltage change across an inline test resistor using my Fluke scopemeter not my cheapo current meter since it will not handle pulses. Funny, the driver side sensors show a steady reading but the passenger side does not.
What was it Sherlock Holmes said "When you have eliminated all that is impossible, then whatever remains, however improbable, must be the truth."
Regards,
Ron

 

FBB,
Can you explain how these female pins stop working? Is it lack of spring tension, corrosion or what?
Thanks,
Ron

 

Jim,

Well I hard-wired pin 1 with a flying lead and put an oscilloscope on an inline temporary 0.1 ohms resistor on the pin 1 (=mosfet drain). Indeed I see regular, rectangular pulses of 50mV and 60 millisec duration, so the ECU is indeed driving the O2 heater with a pulsed current. That explains why I see fluctuations in current with a simple multimeter.
Putting the simple multi-meter inline to measure "DC" current I see oscillations as before but now the max. current is only 2.0A not 3.5A, so I am hoping the bypassing of the loose female connector will stop the P1117 code.

The max. current drops from 2A to 1.7A over a few minutes so it looks like all is working.

Followup questions , will the old P1117 codes remain stored in the ECU forever, or disappear if switch battery switch off for a while. Will a generic OBDII scanner remove them. If no new P1117 codes are triggered will CEL light stay out and OBDII readiness signals appear?

Need a number of drive cycles before I will find out.

Conclusions so far:
1. mosfet is OK, just need to be aware it uses pulse drive
2. simple multimeters will show an oscillating reading 0 to 2A due to the pulses
3. pin 1 and pin 30 female connectors need to be checked for tightness or replaced or bypassed.

Thanks,
Ron

 

Just noticed that "spreading pins" is also mentioned in this thread about washing the car:
http://www.ferrarichat.com/forum/456-550-575/414415-watch-550-washing-guys.html
and FBB tells the full story here about loss of spring tension in the female clip
http://ferrarichat.com/forum/technical-q/312756-sri-gold-connector-pin-kit-strikes-gold.html
Regards,
Ron

 

Dear FBB,
I have been catching up on your old threads on the SRI kit, and I am now definitely a believer in the need to do something about those loose female clips. I realize now that trying to coax more life out of the old clips by bending them in is unlikely to be successful.
I also wonder if all those attempts to cure "slowdown" lights and cat ecus issues may be more to do with "loose female slots" !
Also you and Dave Helms mention that more than 5 disconnect/reconnect of the ECU connector can cause failure, my time is up since I have reached 5.
The hardwire bypass is just a work around until I can do the complete SRI kit.
Regards,
Ron

 

Jim and Mark,
The Bluetooth scanner price looks good, can you read real-time data when the car is in motion as well? My scanner is meant to read real-time (at idle) but the numbers look wrong, Ferrari proprietary interface?
Taz has mentioned the key cycle sequence a number of times , and I have somewhere the proper drive cycle instructions to get the CAR AIR EGR etc monitors ready.
Mark, yes I am aware about 2 not ready is OK for Texas inspections, I have just reached that condition after 3 *20 min runs today. Do you know if you can pass with a pending code (amber light on scanner) as long as no MIL or CEL is lit? The P1117 is still pending even after clearing with the scanner maybe an old pcode in the ECU long-term memory.
Regards,
Ron
PS Is Taz sick he is usually one of the first to jump in? I saw him lurking on the F50 threads.