Fitech repair

[Kain]

well i asked tech at fiTech will i have to load software and this was Joe reply

The failure was not necessarily caused by heat but was present when the temp of the ECU was increased. So it might have worked cold but when it got hot it failed. It does not hurt to run a spacer if hood clearance allows and anytime you can keep the fuel and electronics cooler the better. The new ecu will be loaded with the correct software but you will have to reset your default numbers, engine cubic inch etc and any tuning that you did.

Thanks,
Joe

 

[Kain]

so yes on the spacer

 

[anwat]

Are you still running a fuel command center?

No, No, NO...I've heard nothing but bad things about the Fuel Command Center. I went with the Delco 25163473 255 lph in the tank pump like the Vortec 350s use. Other than wiring it backwards the first time, haven't had an issue with it, and it fits in with the stock sender etc. That's the same setup I used when I upgraded the TBI on my 90. If you do decide to go with the GM TBI, take a look at the recommendations on Harristuning.com. I followed them and had the rebuilder set the engine up exactly as he recommends, with the exception of I used stock exhaust manifolds rather than headers so I could get through a smog inspection. I dynoed the truck wearing 35 inch tires and a 205 t-case and it was putting out 132 hp and 200 ft-lbs of tq at 3900 rpm consistently. And would still get around 20 mph on the flat open road. I was so happy with that engine that the truck is still sitting in my garage, I can't bring myself to sell it. The FiTech equipped truck is easier to tune and troubleshoot, but the GM TBI is very simple, and hasn't let me down yet. If you already have the wiring and sensors in place for the GM TBI, my vote would be to go with that and do some minor mods like Harris details to get a little more out of it but retain that reliability.

 

[anwat]

OK, one of you youngsters is going to have to explain to me how a computer failure can be present when heat is increased, but isn't caused by heat. What the hell does that actually mean? It means the ECU got hot and failed, but that the failure wasn't necesarily due to heatl And what type of electronics fail when hot but will then work again when they cool? Except something that has a thermal cutoff switch, it seems like if it got too hot and something melted, it's not going to work when it cools down ,either. Maybe I'm being cynical, but that kind of sounds like crap to me. They designed that unit to sit on top of a running engine....it should be ale to take that amount of heat without failure.

 

[Kain]

ya sounds like he was trying (but failed ) to say it was not a design flaw

 

[Bent77]

ya sounds like he was trying (but failed ) to say it was not a design flaw

I agree, but without any scientific proof. But I do work on a lot of electronics. Me and a few other techs in my area have done autopsy on some of the parts failures and found that heat caused failures due to cheap production items (think solder and silicone)

Cheaper to build can drive this type of issue. Not accusing anyone, but an observation from another field

 

[Kain]

There are a guy here that have bought several units and when one failed he took it apart and said same thing shitty solder
mind the diff is he is a master electrition and tore all of em apart and fixed it. still runs em has like 8 or 9 cars he puts em on

 

[Blue85]

OK, one of you youngsters is going to have to explain to me how a computer failure can be present when heat is increased, but isn't caused by heat. What the hell does that actually mean? It means the ECU got hot and failed, but that the failure wasn't necesarily due to heatl And what type of electronics fail when hot but will then work again when they cool? Except something that has a thermal cutoff switch, it seems like if it got too hot and something melted, it's not going to work when it cools down ,either. Maybe I'm being cynical, but that kind of sounds like crap to me. They designed that unit to sit on top of a running engine....it should be ale to take that amount of heat without failure.

Basically it has an intermittent connection somewhere that is affected by thermal expansion. A crummy solder joint can do this, as can a cracked resistor or capacitor. Sometimes these flaws are hard to see even with a microscope, but it's enough to affect operation. It's possible that ECU left the factory with some bad solder joints or another flaw and it's possible a cheap component failed after a lot of thermal cycles.

Short answer is that it's a quality control issue - either supply chain management, manufacturing or validation.

 

[Blue85]

Keep in mind that smaller companies don't make the investment to do Printed Circuit Board manufacturing or even circuit card assembly. They will typically work with a larger manufacturer to build their little boxes or even handle all of the design. Aerospace type extended thermal cycle testing of each ECU would cost more than the ECU, so this limits what testing is practical. With larger-run electronics you do full testing of smaller samples - say 1 box out of every 1000. Automated assembly, statistical analysis and process controls keep failure rates within targets. An OEM ECU has been validated in facilities costing 10's of millions of dollars and this helps you understand why aftermarket electronics typically have much higher failure rates.

 

[anwat]

Keep in mind that smaller companies don't make the investment to do Printed Circuit Board manufacturing or even circuit card assembly. They will typically work with a larger manufacturer to build their little boxes or even handle all of the design. Aerospace type extended thermal cycle testing of each ECU would cost more than the ECU, so this limits what testing is practical. With larger-run electronics you do full testing of smaller samples - say 1 box out of every 1000. Automated assembly, statistical analysis and process controls keep failure rates within targets. An OEM ECU has been validated in facilities costing 10's of millions of dollars and this helps you understand why aftermarket electronics typically have much higher failure rates.

So what is actually happening? the solder gets hot and soft and the connection fails, but then when it cools, it connects again? Is it contamination of the solder with something that doesn't conduct? I'm not a electronics expert or even close, but in my experience, usually electrical stuff works or doesn't work. Not at all saying you're wrong, just trying to understand what could be going on that would cause hot failure and then fix itself when it cools. I do get the expansion with heat, and how that might impact a resistor with a crack, even a circuit board expanding and breaking a connection, then cooling and it touches again. It seems odd to me that something would fail only when hot, but then work when it cools off, but I guess it's possible. Either way, though, it sounds like we all agree that this is a manufacturer-caused problem. In some way, their processes, supply line, quality control, or subcontractors caused this issue. Which is why, I'm guessing, FiTech offers a one year warranty, not a three year warranty.

By the same token, though, the other manufacturers would be subject to the same issues, correct? I'm speaking of the other aftermarket ECU injection systems like Holley and Edelbrock etc. Which takes us back to the option of an OEM system that Boetiek 10 was asking about. Yet another reason to go that route. All the time and money invested in the OEM parts and design are going to give you a much more reliable unit than anything aftermarket.

What you're saying makes sense in terms of quality control, though. I know we've all purchased a brand-new chinese auto part and had it not work. I guess it depends on where your box was on the line...before or after the sample box they tested, right?

 

[Bent77]

So what is actually happening? the solder gets hot and soft and the connection fails, but then when it cools, it connects again? Is it contamination of the solder with something that doesn't conduct? I'm not a electronics expert or even close, but in my experience, usually electrical stuff works or doesn't work. Not at all saying you're wrong, just trying to understand what could be going on that would cause hot failure and then fix itself when it cools. I do get the expansion with heat, and how that might impact a resistor with a crack, even a circuit board expanding and breaking a connection, then cooling and it touches again. It seems odd to me that something would fail only when hot, but then work when it cools off, but I guess it's possible. Either way, though, it sounds like we all agree that this is a manufacturer-caused problem. In some way, their processes, supply line, quality control, or subcontractors caused this issue. Which is why, I'm guessing, FiTech offers a one year warranty, not a three year warranty.

By the same token, though, the other manufacturers would be subject to the same issues, correct? I'm speaking of the other aftermarket ECU injection systems like Holley and Edelbrock etc. Which takes us back to the option of an OEM system that Boetiek 10 was asking about. Yet another reason to go that route. All the time and money invested in the OEM parts and design are going to give you a much more reliable unit than anything aftermarket.

What you're saying makes sense in terms of quality control, though. I know we've all purchased a brand-new chinese auto part and had it not work. I guess it depends on where your box was on the line...before or after the sample box they tested, right?

In some of the non-automotive stuff we tested, we suspected the solder “may” have done something like that, but also found the terminals no longer soldered at all
I do know from working on instruments that they can be soldered, connected, but alter the OHM reading because it’s not soldered correctly

FWIW, the Edelbrock ECM has enough wiring with the harness that it ran be remote mounted like the OEM components. I have mine under the dashboard, similar to a TBI

 

[Bent77]

Any of the manufacturers may have this issue, but certainly one could if they used a low heat solder in production

Probably more noticeable in a region like @Kain is from, Texas

 

[Blue85]

There's actually a wide science behind soldering. Early days was one guy with an iron and he could watch each joint be formed. Today there are machines doing 1000-pin Ball Grid Arrays with a specifically created heat profile that a human operator could not apply. An easy example is a part with 30 pins on each side. One cold solder joint on pin 15 will be held in place by the other 29 pins on that side of the component and 90 other pins on the other 3 sides. But when you go from the 72F where the board was built to the 250F on top of your engine, parts can expand enough where that one doesn't make good contact anymore.

If the CCA design isn't careful enough, a component is stressed a little during EVERY thermal cycle until eventually it cracks. Really easy to say a circuit card works after initial assembly. Much more complicated to say it will work over time and a huge range of environmental conditions.

 

[mrk5]

In some of the non-automotive stuff we tested, we suspected the solder “may” have done something like that, but also found the terminals no longer soldered at all
I do know from working on instruments that they can be soldered, connected, but alter the OHM reading because it’s not soldered correctly

FWIW, the Edelbrock ECM has enough wiring with the harness that it ran be remote mounted like the OEM components. I have mine under the dashboard, similar to a TBI

I've been wondering if remote ECU's are the more reliable way to go.

My ECU failure on the Sniper wasn't long after the first time I pulled the camper up the high mountain passes and the motor got pretty damn hot a few times. Luckily in my case Holley covered the cost of the replacement even though it was several months out of warranty.

 

[6872xtc]

I've been wondering if remote ECU's are the more reliable way to go.
.

This was one of the reasons why I chose the Edelbrock. More like a OEM system that way.
I know that Cummins 11 and 15 liter, as well as Duramax diesel engines have the ECM mounted on the engine, but they are cooled with fuel. Entertaining since they are common rail which puts heat into the fuel. Makes me wonder what the temperature can get up to with those.

 

[mrk5]

The higher end Holley EFI also goes to remote ECUs. I think it's an interesting correlation and probably answers its own question.

 

[warrpath4x4]

This was one of the reasons why I chose the Edelbrock. More like a OEM system that way.
I know that Cummins 11 and 15 liter, as well as Duramax diesel engines have the ECM mounted on the engine, but they are cooled with fuel. Entertaining since they are common rail which puts heat into the fuel. Makes me wonder what the temperature can get up to with those.

currently the ECM's for the N14 are having solder issues also. It is so bad we have to take your old ECM, ship it to Mexico to be checked and see if its repairable, if so they fix it. If its not repairable then they will authorize the purchase of a Recon ECM.

Issues with a certain housing styles of the ISX, also with solder issues.

 

[Kain]

well i got it back today, sad part is i think ill wait till after my MRI before tearing into it so ill know i have a ride......

 

[Kain]

They at least cleaned it

 

[6872xtc]

currently the ECM's for the N14 are having solder issues also. It is so bad we have to take your old ECM, ship it to Mexico to be checked and see if its repairable, if so they fix it. If its not repairable then they will authorize the purchase of a Recon ECM.

Issues with a certain housing styles of the ISX, also with solder issues.

That is good information, but I not something that I really want to hear.


We have almost close to 30 Cummins powered trucks at work. I won't say anything else to tempt karma..