Fit ECU failure

[miberns]

My 2 month old FIT ECU failed on pregrid at the Seattle majors event. As I happen to be a design engineer, I checked it out and found that the failure was caused by incomplete soldering of the processor.

Honda won't even look at it and wants me to buy another ECU for $1200 and suggests that I carry a spare.

In modern electronics, this kind of failure is simply not acceptable. If Honda doesn't care enough about reliability to do failure analysis, I think SCCA needs to seriously reconsider our relationship with the company.

[Jon Jeffords]

Ooof. With an early failure like that, one would hope they at least take a look at it to determine if it has a defect due to materials or workmanship.

[Dee]

Jon,
sorry to hear about your uncle, he was a true friend>
dave
pd ca.

[Rick Kirchner]

is the processor dead, or can you re-flow the solder and it will work again?

If it's dead, can you buy the processor (assuming it has no firmware that cannot be externally loaded) and solder it in?

I had a couple of those AIM tach amplifiers die (they're really a $90 fuse) and the IC was only 89 cents apiece - so I had one of the techs at work replace two of 'em for the cost of a olive garden gift card.

I don't recall the rules - stock ECU with a firmware update, or is it custom?

If custom, agree that they should stand behind the product pretty much regardless of age/use - they knew what the environment was when they spec'd it.

funny thing - when EFI came into FC I don't think people really thought about the sparing issue. Before EFI all you needed was a spare float, an accelerator pump diaphragm, points, a battery, and maybe a fuel pump.

After EFI you need a couple of connectors, a pin tool, pins, crimpers, just about every sensor in the system (or at least the ones with a high failure rate) an ECU - plus the battery, fuel pump, at least one relay, probably a couple of other things I don't know about - and unless you're just box plugging, you are probably going to need a laptop and some specialized test software.

Yeah, in general EFI is more reliable, but when it has a problem, you're usually done - no limping about.

[Raceworks]

My 2 month old FIT ECU failed on pregrid at the Seattle majors event. As I happen to be a design engineer, I checked it out and found that the failure was caused by incomplete soldering of the processor.

Honda won't even look at it and wants me to buy another ECU for $1200 and suggests that I carry a spare.

In modern electronics, this kind of failure is simply not acceptable. If Honda doesn't care enough about reliability to do failure analysis, I think SCCA needs to seriously reconsider our relationship with the company.

I had exactly the opposite experience with HPD when I had a similar failure on an ECU on the Piper back in 2011. The only difference is I didn't break the seals on the ECU.

[Raceworks]

is the processor dead, or can you re-flow the solder and it will work again?

If it's dead, can you buy the processor (assuming it has no firmware that cannot be externally loaded) and solder it in?

I had a couple of those AIM tach amplifiers die (they're really a $90 fuse) and the IC was only 89 cents apiece - so I had one of the techs at work replace two of 'em for the cost of a olive garden gift card.

I don't recall the rules - stock ECU with a firmware update, or is it custom?

If custom, agree that they should stand behind the product pretty much regardless of age/use - they knew what the environment was when they spec'd it.

funny thing - when EFI came into FC I don't think people really thought about the sparing issue. Before EFI all you needed was a spare float, an accelerator pump diaphragm, points, a battery, and maybe a fuel pump.

After EFI you need a couple of connectors, a pin tool, pins, crimpers, just about every sensor in the system (or at least the ones with a high failure rate) an ECU - plus the battery, fuel pump, at least one relay, probably a couple of other things I don't know about - and unless you're just box plugging, you are probably going to need a laptop and some specialized test software.

Yeah, in general EFI is more reliable, but when it has a problem, you're usually done - no limping about.

The FF ECU's are a custom unit that is manufactured by HPD specifically for the class. The GCR lists them as "sealed units" and doesn't permit tampering. You're not even allowed to mess with the engine wiring harness either. The ECU has two tamper-evident labels: one at each end of the box. This is why while I have had a problem that's very similar to what the original post describes (albeit after a MAJOR impact) it never occurred to me to pop the box open. I exchanged it for a "fresh" one from HPD and never heard about it again.

Technically removing the seals and soldering in a new chip would count as "tampering" but I'm not a technical steward so I'd get a "professional" opinion on how that would gel with the GCR.

On the Hondas the only things I've seen problems with are usually the fuel pump (particularly if you do something stupid like drain the fuel & leave the pump on for 15 minutes). The only thing I've been told that will keep the engine from starting completely is the Crank Position sensor (relative cheap and very easy to replace). Anything else and it will still fire but run poorly (plus if you wire the trouble light in correctly you'll get a code).

[Rick Kirchner]

well that puts a whle different perspective on it, although even if the tamper seals are broken the failure mecanism should be evident, as would any attempt at real tampering/replacement attempt, etc. In other words, just popping the cover off shouldn't cause a manufacturer of such a specialized item to balk at replacment. consumer goods, yes, but this thing, no.

Interesting that they don't want you messing with the loom. They make a custom one for each installation? Or do the british thing and put in something that's way long everywhere and just tie up the excess?

Crank position sensors are a known failure item throughout the industry. two of my last three vehicles have been towed because of them.

[problemchild]

I fully sympathize with mibern's tough situation.

FWIW, I thought that sales of electronic items like ECUs were "you bought it, you own it, don't bring it back" in the automotive industry. They tend to get fried when other problems exist in the vehicle. The "no return" policy prevents the discussion that the seller will lose, regardless of the true cause. For that reason, I won't loan out our spare ECU. I have sold it twice however.

[Rick Kirchner]

I've always felt that the "no return" policy on automotive electricals was simply because for decades mechanics simply couldn't comprehend them.

Nowadays, with input protection, polarized connectors, etc, there's no excuse.

[Dave SanF 50]

In my almost 25 years of racing Formula Ford, my Kent's have never, ever had a ECU failure. And before you even ask, never broken a crank, rod, piston, cam, rocker shaft, clutch or clutch disk. I did break a valve spring last year and a few years before an accelerator pump diaphragm.
Of course, they don't make 120 hp either.
YRMV

[problemchild]

In my almost 25 years of racing Formula Ford, my Kent's have never, ever had a ECU failure. And before you even ask, never broken a crank, rod, piston, cam, rocker shaft, clutch or clutch disk. I did break a valve spring last year and a few years before an accelerator pump diaphragm.
Of course, they don't make 120 hp either.
YRMV

I broke a crank in my first FF race. I have a buddy who blew up in his first FF event. We both quit the class within a year. That, like your comments, is irrelevant to the topic being discussed.

There is a kent FF in SEDIV that reportedly has a 122 hp engine. If you believe Honda's have 120 hp, I'm sure you will believe that too

[Dave Harmison]

From my personal experience with electrons I have found that they have a negative attitude, know they are invisible and take advantage of the situation.

If they would just have the courtesy of leaving a puddle when they leak I might get along with electrons.

Although that would lead to other issues, like someone would have to invent "electron-dry".

[Dave SanF 50]

I broke a crank in my first FF race. I have a buddy who blew up in his first FF event. We both quit the class within a year. That, like your comments, is irrelevant to the topic being discussed.

There is a kent FF in SEDIV that reportedly has a 122 hp engine. If you believe Honda's have 120 hp, I'm sure you will believe that too

Greg, sorry to have touched a tender spot but some of my post was meant to be funny.

You missed the first clue that part of my post was in jest which was the part about a Kent having a ECU.
Second, I did not say the Honda had 120 hp, I never mentioned Honda in my post. My statements were regarding MY Kents. MY Kents have never, ever made 120 hp.

I am sorry for the confusion I caused.

[problemchild]

Greg, sorry to have touched a tender spot but some of my post was meant to be funny.

You missed the first clue that part of my post was in jest which was the part about a Kent having a ECU.
Second, I did not say the Honda had 120 hp, I never mentioned Honda in my post. My statements were regarding MY Kents. MY Kents have never, ever made 120 hp.

I am sorry for the confusion I caused.

Your (and my) comments were still irrelevant to the topic. This poor guy is caught in the middle of a tough but understandable situation. I wish him all the best in finding a solution.

[racerdad2]

No warranties on electronic parts is hard to take. No warranties on racing parts is understandable, still of little comfort. With the wiring loom untouchable, us it possible to relocate the ECU in a more electronics friendly location with some amount of vibration dampening ? And, I imagine an onboard switch to backup ECU is verboten. ... Here's to good fortune & long-life on the next round of electronics !!!

[BLS]

If the problem with the unit is as stated by the OP, Honda should want to see it and stand behind it, sealed unit or not.

[miberns]

Hi Greg,

I've been running Kents for 20 years in one form or another. I build my own motors, a practice that bears further discussion but not now. I've cracked 2 cranks, but found both in a routine rebuild before any disaster. Both had been exposed to more than 7800 rpm in downshift overrevs I've never broken a motor without driver (me) abuse. I'd check carefully if the flywheel can hit the ground, oil starvation can occur etc before blaming the motor.

Regarding FIT power, I know that the FIT with a 29.5 restirictor makes the same peak power as a Kent prepared with 1 race loose rings, perfect porting and exactly tuned for weather conditions (timing and mixture). Oh and the Kent is running on $50 / Gal fuel, and the humidity and barometric pressure haven't changed since it left the paddock.

However, peak horsepower is not the whole story. The FIT makes about 50% more power than the Kent at 5000 RPM, and will happily run against the rev limiter, which will break a Kent. These are things that happen in real racing conditions, at least with a typical amateur racer (me).

While the Honda conversion cost as much as I would have spent rebuilding Kents for the next 15 years, I would never go back. My lap times are noticeably faster with the 30.5 restrictor, and my ability to pass is dramatically improved.

Regarding the ECU, there is no way a special part made for racing can compete with a production part for reliability. Less than 200 of our ECU have been made. Honda has lifetime tested more than that of anything that gets to production. The production parts are watertight and use far superior custom made connectors to what can be bought in small numbers. From what I gather the field failure rate of the ECU is in the range of 3% to 5%. This is about 100X what can be tolerated for a production car. However, if this part is to have reasonable cost, there is really no other alternative. Honda can do what they want, but I think that they should offer a spare ECU at cost to anybody who bought one at full price. The BOM cost of the PCB is about $100, assembly, test and packaging could cost as much as another $250, or even more if the lot size is small.

In my ECU, the processor stopped running. So there was no MIL indicator.

[Will Velkoff]

$50/ gallon fuel?

50% more horsepower at 5000 rpm?

Please explain. I also have seen a car first hand before and after a Honda conversion, and from what I've seen, the car is not noticeably faster in total lap times.

[Jon Jeffords]

The FIT makes about 50% more power than the Kent at 5000 RPM....

Source of data please? Are you refering to an unrestricted Honda?

[miberns]

Sorry everyone about the 50% remark. The Honda has more power at the bottom end but no close to that much. I'm just frustrated. My Honda crapped out again this weekend due to a throttle position sensor.

I'd like to bellive that putting in the Honda made me a better driver. Regardless of how it happened I'm going a lot faster.

I don't know the status of the hot fuels. The fuel rules were intended to get rid of them, but when I was using the ERC version, it just got more expensive to make it meet the new rule. My dyno tests showed a significant gain, 1.5 at the peak and 3 at 5500. Somebody please chime in and let me know if they are still around? At the time ERC was about $20 per gallon and ELF was $40.

The harness is too short to get the ECU behind the dash which is where I want it. I guess I'm illegal because I had to repair my harness to make it work in the first place.

Mike

[Daryl DeArman]

Somebody please chime in and let me know if they are still around? At the time ERC was about $20 per gallon

Yep. Right in your backyard.

http://www.epictrainingsystem.com/