'09 Suzuki Bandit 1250 S
Suzuki Bandit 2.jpg (234080 bytes)

Purchased in 2011with 5100 miles on it. 

The previous owner made the cases and put them on really nice SW-Motech mounts.

It's got a Holeshot Performance Touring Slip-on exhaust and a Dobeck Performance EFI controller adjusted to the settings recommended by Dale at Holeshot.  

I picked up a Healtech Gear indicator and mounted it on the dash, which is nice for minimizing attempts to shift into 7th gear, and a Healtech SpeedoHealer to fix the ridiculous 10% high the speedo (and therefore the odo) reads compared with a GPS.

Also have the Healtech Suzuki OBD connector which has proved very useful in clearing DTC codes.

The first ride I took on it, on the way home from buying it, I got the dreaded C44 FI light that is a problem with aftermarket EFI controllers on Suzukis.  Naturally, the previous owner had somehow never experienced it before.   Imagine that.  Anyway, here's what I'm doing about it:

June 2011

Contacted Dale at Holeshot Performance, since the Dobeck EFI controller was supplied by Holeshot as a EFI and exhaust upgrade package.  He gave me a name at Dobeck to contact about the EFI controller.

Contacted Matt at Dobeck and was told that the C44 code was stored despite the light staying off after the bike was turned off for a few minutes, and that the code would have to be cleared using the Suzuki Diagnostic Software.   It's apparently like $750 to buy, or you can have the dealership clear the codes for an hour's worth of labor.

Bought the Healtech OBD software and installed it on my garage laptop (also has the TwEECer software for my Mustang).  It does diagnostics and clears codes very easily for $220.  Determined that the DTC I was getting was C44/P0135.  C44 is the Suzuki error code for the O2 sensor and will flash on the LCD panel when the bike is placed in "Dealer Mode".  P0135 is a OBD diagnostic code indicating that the O2 Sensor Heater is not functioning properly.  After some practice, I could make the FI light come on in 1-2 minutes by maintaining ~4400 RPM in 6th gear.

The Dobeck EFI controller supplies a false O2 Sensor signal output to the ECM but it doesn't simulate the O2 Sensor Heater.  That's done with an "O2 Sensor bypass", which is just a 270 Ohm 1/2 Watt resistor in the Heater circuit.  The actual O2 Sensor is left unplugged. My research found that most people addressed this by clearing the code and changing the resistor to a higher Wattage rated resistor of the same or slightly higher resistance. It also indicated that this rarely actually solved the problem and most people just live with it or stay out of the conditions that give the code.  "Learning to live with it" isn't good enough, so I cleared the code and replaced the resistor with a 0-1kOhm 2 Watt potentiometer, thinking I could adjust the resistance until the FI light didn't come on.   I tried all sorts of resistance settings from 200 to 900 Ohms, and always the FI light came on at about the same time.  It always gave me the same DTC code, C44/P0135 - O2 Sensor Heater Circuit failure.

I followed the procedures in the Service Manual for diagnosing these DTCs, and everything seemed to be reading properly.

July 2011

Several emails back and forth to Matt at Dobeck and I came upon an idea.  If the problem is that the ECM is looking for a response from the O2 Sensor Heater, and according to the wiring diagram it is, why try to fake that signal if you don't have to?  So I went through Eastern Beaver and ordered several sets of Sumitomo MT series connectors so I could make a harness adapter to use the bikes O2 sensor Heater circuit (O/W power and B/Br output wires), but not the Signal circuit (W/G signal and W/B ground wires).  This is what I made:

HarnessSmall.jpg (22235 bytes)

The R and B/R wires are the O2 Sensor Heater circuit wires.  The smaller B and W wires on the right are the Signal and Ground wires from the O2 Sensor.   I left them unconnected because a) the Dobeck EFI controller supplies a false O2 Sensor signal to the ECM so these wires are not needed, and b) I might want to datalog the voltages at these wires under operating conditions in the future.  The smaller B and W wires on the left are the O2 Sensor Signal and Ground wires to the ECM (which should show the voltages supplied by the Dobeck EFI controller to the ECM instead).  They will also be used for possible datalogging later.

With this harness installed and the codes cleared, I tested the bike in the garage.   Startup and warm up: no FI light.  Bump throttle to 2000, 4000, 8000 RPM, no FI light.  Since it's raining, I'll test the bike on the road later.

I took the bike out for a ride.  In the 10 mile ride to the highway, the FI light did not come on.  That's normal.  Got on the highway and ran it up to ~4400 RPM in 6th gear (~76 mph).  I made it south about 15 miles in about 12 minutes and still no light - much longer that it normally goes.  Got off the highway, turned around, and got back on.  Went about 16 miles north (just past my exit, of course) and the FI light came on.  Damn.

Back in the garage, I hooked up the Healtech OBD software and read the code.   It was a new one; C44/P0130.  This code has to do with the O2 Sensor Signal not adjusting fast enough.  Since this code took longer to throw than the original C44/P0135 code, I suspect the P0135 code prevents the ECM from looking for the P0130 code since there's already an error in place that would skew the O2 Sensor Signal out of whack.

So, the Good News is that using the O2 Sensor Heater circuit seems to address the C44/P0135 DTC, and the Bad News is that the C44/P0130 DTC is generated because the Dobeck EFI controller isn't sending the proper signal to the ECM and there's nothing my little harness can do about that.

I've emailed Matt at Dobeck on these findings, and hopefully after the weekend we'll be able to figure out a plan.  I will be datalogging the O2 Sensor Signal voltage so the EEs at Dobeck can use that to see what the ECM is looking for, and perhaps they can adjust the model accordingly.

Just to see what I might see, I hooked up the DMM to the W and B wires on my little harness with the bike idling.  The output voltage from the O2 Sensor is between 0 and about +0.8VDC compared to Sensor ground.  That is in the range indicated as proper in the Service Manual, and what should be sent to the ECM.   However, the voltage was +2.54 VDC and steady as read on the W and B wires on the other end of the harness.  That +2.54VDC is way out of spec for the ECM.  I suspect herein lies at least part of the problem.


I have instrumented the O2 Sensor signal output voltages (against body ground) and the voltages being supplied to the ECM by the Dobeck EFI controller (also against body ground).  First step is to clear existing codes, then attach the Fourier DAQpro to the harness.  I am measuring both sides of the O2 Sensor signal against body ground, and both sides of the ECM signal supplied by the Dobeck EFI controller against body ground as well.  If one lead of each signal is always about 0V (as I'd expect from the Dobeck unit), I can use the other as the actual value.  If not (as I'd expect from the O2 Sensor), the difference between the two will be what the ECM would see if the Dobeck unit were not installed.

Here's the setup:

DobeckSettings.jpg (55129 bytes) OBDsetup.jpg (72854 bytes)
Dobeck EFI Controller settings

 

First, clear the existing C44/P0130 code.  Hook up the hardware...

 

OBDscreencap1a.jpg (27313 bytes) OBDscreencap2a.jpg (21451 bytes)
Connect Healtech ODB Tool...

 

...and clear the code.  just that easy.

 

DAQconnections.jpg (74538 bytes) DAQready.jpg (79764 bytes)
Cat5e cable. O2+ = blue, O2- = green, ECM+ = copper, ECM- = bronze

 

Here's the wiring setup.  Fourier DAQpro 8-channel.   I'm using 4 channels.

 

Next step is to go out and generate the code while the DAQ system is running.


Ok, that was weird.  You know how when your car is acting funny and you take it to the shop and they say "We couldn't find anything wrong with it"?   That's what happened.  I made three passes down the highway with the DAQ system in place doing the things that I KNOW WILL LIGHT THE FI LIGHT because it does it every time.  Didn't do it this time.  Not once.  Not even a little bit.   Got back and checked with the Healtech OBD software, and sure enough no error codes were generated after three passes.  I downloaded the data into an XL sheet (one Book for each run), and added notes, headers, and a couple "delta V" columns to indicate what the voltage across the O2 sensor + to - and the ECM from Dobeck Unit + to - would be.  Here it is: 

BanditRuns1_2_and_3.xls

As you can see the Dobeck unit outputs a nearly even 1V to the ECM once the bike is off idle.  The O2 Sensor signal is all over the place, as would be expected since the fuel map is not actually being adjusted by the O2 Sensor output so the O2 content of the exhaust varies widely.

Now the fact that the light didn't come on intrigues me.  Since I know the old TOIWTOTTPOO adage (The Observer Interacts With The Observed Through The Process Of Observation), I used the DMM to measure the input resistance of the DAQ device with it turned on and with it turned off.   Turns out that with the DAQ device turned off, the resistance between the ECM from Dobeck + and the ECM from Dobeck - is ~700kOhm.  Not Open Circuit resistance, but still pretty high.  With the DAQ device turned on, however, it's ~250kOhm.  That lower resistance under the Observation conditions may be allowing enough current to flow that it tricks the ECM into thinking there's a response from the O2 Sensor instead of a simple voltage being applied.  It's 250kOhm and only 1 volt, but still...

Since the bike didn't catch fire and I noticed no odd performance characteristics during the tests, the next step will be to build a harness that uses the O2 Sensor heater circuit AND has an adjustable 250kOhm - 1MOhm resistance across the ECM from Dobeck leads in the bike wiring, and test it to find the highest resistance that will still not throw the C44/P0130 code.

OK, I went to Radio Shack and picked up some 1MOhm 1/2Watt 5% tolerance resistors.   I put four of them in parallel to give me 250kOhm, and used the DMM to verify it was near that range.  I added it to a new harness that uses the actual O2 Sensor Heater Circuit (R and B/R wires) and places the 250kOhm resistor across the + and - of the ECM O2 Signal from the Dobeck Unit.  These resistors will simulate having the DAQ system attached.  If the FI stays off at the next trial, I'll cut one 1MOhm resistor out of the bundle, increasing the total resistance to 333kOhm and test again.  If no FI light again, cut out one more to give 500kOhm total.  test again, and go to a single 1MOhm resistor.  If that still keeps the FI light from coming on, I'm going to call it solved.  Here's some pix of the New Harness:

Resistors1.jpg (44889 bytes) Resistors2.jpg (43957 bytes)
4x 1MOhm Resistors.  1/2 Watt, 5% tolerance

 

Resistor bundle.  Total R = 250.3 kOhm

 

NewHarness1.jpg (42046 bytes) NewHarness2.jpg (64154 bytes)
Resistor bundle installed in New Harness

 

New Harness installed on bike between O2 Sensor and bike wiring.

 

Well crap.  That didn't work.  Got the FI light within 2 minutes of getting on the highway with the 250kOhm bundle in place.  Came back and checked: it was C44/P0130 O2 Sensor Signal timing, same as before.  Crap.

Some more experimentation is in order.


Now I've made runs with the harness I used to test the voltagaes but without the DAQ installed and also with a 100microF cap across the ECM from Dobeck + and - leads to keep the +1VDC signal but allow the AC noise to pass.  Took longer, but got the same result - C44/P0130 both times.  Crap.

I found this O2 Sensor Bypass from Magnum Tuning.  It's supposed to do what I want done, that being clear the FI light, but I'm leery about whether it'll work with the Dobeck EFI controller.  I've emailed them to see what they say.

Updates forthcoming...