Failed 3FE ECUs (1 Viewer)

This site may earn a commission from merchant affiliate
links, including eBay, Amazon, Skimlinks, and others.

@Spook50 Random thoughts...

Wild guess is that the first or most common failure one would see is cracked solder joints, particularly where the multipin connector is soldered to the board. Moisture would hurt too if it's a parts truck with broken windows dragged out of a field. Those electrolytic caps, like you suggest, also have a lifespan. But typically you see power filtering caps on big B+ rails having issues long before decoupling caps would - judging by the size, I bet a lot of those are for decoupling or else filter caps on a tiny B+ rail. The ceramics will last essentially forever. The carbon film resistors won't drift as much as older carbon comps will, so unless one of them is exceeding it's wattage rating those should last more than 40 years. Those look like 1/4w, and at worst case (14.7VDC) they can only handle 17mA. I would think most of what's happening in there is logic/switching network stuff that doesn't see significant current draw, but if they exceeded 17mA they could drift or even fail open. You'd probably see a toasty spot on the resistor if the latter was the case. The resistor point might be moot anyway, it looks like there's voltage regulators in there, possibly dropping the voltage to 5VDC for logic stuff, which means the resistors would be good to 50mA. Heck, I'd say a voltage regulator would go bad before the resistors. Out of all the gear I've repaired I've seen a few bad VRs, all about 30-40 years old. Not frequent though.

I bet a lot of those components are NLA or labeled with an in-house designation, obscuring what a replacement would be. It's too bad because you could just have new PCBs made and populate them with new components. That would maybe be more labor effective than refitting that board. Not that it couldn't be done redesigning the entire thing with modern production components. At the end of the day I agree with Godwin and other's assessment of how limited the market is. It would be a LOT of R&D (time + money) to sell a handful of them. You'd have to charge $$$$$ to recoup.

Not trying to piss in your cheerios, just sptiballing.
 
@Spook50 Random thoughts...

Wild guess is that the first or most common failure one would see is cracked solder joints, particularly where the multipin connector is soldered to the board. Moisture would hurt too if it's a parts truck with broken windows dragged out of a field. Those electrolytic caps, like you suggest, also have a lifespan. But typically you see power filtering caps on big B+ rails having issues long before decoupling caps would - judging by the size, I bet a lot of those are for decoupling or else filter caps on a tiny B+ rail. The ceramics will last essentially forever. The carbon film resistors won't drift as much as older carbon comps will, so unless one of them is exceeding it's wattage rating those should last more than 40 years. Those look like 1/4w, and at worst case (14.7VDC) they can only handle 17mA. I would think most of what's happening in there is logic/switching network stuff that doesn't see significant current draw, but if they exceeded 17mA they could drift or even fail open. You'd probably see a toasty spot on the resistor if the latter was the case. The resistor point might be moot anyway, it looks like there's voltage regulators in there, possibly dropping the voltage to 5VDC for logic stuff, which means the resistors would be good to 50mA. Heck, I'd say a voltage regulator would go bad before the resistors. Out of all the gear I've repaired I've seen a few bad VRs, all about 30-40 years old. Not frequent though.

I bet a lot of those components are NLA or labeled with an in-house designation, obscuring what a replacement would be. It's too bad because you could just have new PCBs made and populate them with new components. That would maybe be more labor effective than refitting that board. Not that it couldn't be done redesigning the entire thing with modern production components. At the end of the day I agree with Godwin and other's assessment of how limited the market is. It would be a LOT of R&D (time + money) to sell a handful of them. You'd have to charge $$$$$ to recoup.

Not trying to piss in your cheerios, just sptiballing.
You pretty much hit the nail on the head. Ultimately I'm expecting this will fall to "My ECU started causing problems, and here's what I was able to do to correct them", and that's even IF my own ECU is failing, which I have yet to verify one way or another.

The cracked solder joints could very well be a contributor to ECU "failures" that's at least worth looking at. I need to invest in that good digital microscope I've been wanting.

Completely agree on the resistors, and likelihood of electrolytic caps not causing critical failures.

The more I ponder and look online, it looks like should any passive component fail, it would be the result of one of the transistors going out. The few transistors I've looked up so far can still be gotten (some gray market, but supposedly Toshiba brand), and the ICs I guarantee are custom in-house jobs like you mentioned, so in the unlikely situation that an IC tanks, that whole ECU is good for nothing but the other components in it.

Honestly at this point I just want to play with one and see what I can learn :lol:
 
dark socket is an 80 series ECU #87661-60090
yellow socket is a 62 series ECU #89661-60021

the 80 series uses different mounts and doesn't have the same screw mounting holes

last pic is a 62 series ECU that has a pin diagram on it, not sure what # it is

others I have are 2 x 89661-60021 and 2 x 89661-60020

the "bad" one is 89611-60021

DSCF6115.JPG


DSCF6116.JPG


DSCF6117.JPG
 
Out of curiosity what are the transistors? 2SK___?
All I've been able to look up are what I see in the pic I posted, so as far as the others, I'm not sure.

I might pull my ECU this weekend and take a peek inside if the weather isn't crappy. Also take a look at the solder joints if I break into it.
 
Can’t see any of the faces of the transistors where they’d be labeled. Let me know what you find in yours. I also need to hunt down some Toshibas for a project I’m working on, I think 2SK170s. A lot of those have been NLA’d, so it’s eBay or bust, and a lot of those are fakes.

I’ll go back to my bet that the alphanumeric code on the transistors is an in-house code that gives you zero Google results. An off the shelf transistor relabeled for Denso, but you’ll never be able to find out which one it was because it doesn’t have the standard nomenclature.
 
Can’t see any of the faces of the transistors where they’d be labeled. Let me know what you find in yours. I also need to hunt down some Toshibas for a project I’m working on, I think 2SK170s. A lot of those have been NLA’d, so it’s eBay or bust, and a lot of those are fakes.

I’ll go back to my bet that the alphanumeric code on the transistors is an in-house code that gives you zero Google results. An off the shelf transistor relabeled for Denso, but you’ll never be able to find out which one it was because it doesn’t have the standard nomenclature.
The weird thing is that as far as transistors go, nearly everything has a modern one that's a drop-in replacement. Problem with the "relabeled" ones is finding an equal part.

I wish I still had access to a curve tracer because that would make it so much easier to analyze these transistors and find any replacements, if any are needed.
 
The weird thing is that as far as transistors go, nearly everything has a modern one that's a drop-in replacement. Problem with the "relabeled" ones is finding an equal part.

I wish I still had access to a curve tracer because that would make it so much easier to analyze these transistors and find any replacements, if any are needed.
Right, you CAN find replacements, but it takes a full analysis of the existing one, comparing it to specs of what’s available, buying some new ones, then testing & sorting those to find the individual with the closest specs.

Have done. 10/10 would not recommend.
 
And you still might have to swap a resistor or two to get the new one to bias the right way and operate correctly for a given circuit!
 
Right, you CAN find replacements, but it takes a full analysis of the existing one, comparing it to specs of what’s available, buying some new ones, then testing & sorting those to find the individual with the closest specs.

Have done. 10/10 would not recommend.
Oh I agree it's a HUGE pain in the arse. Possible, but certainly not fun, moreso if you're analyzing a failed one. Best method I had when trying to trace parts years ago was analyze as many as I could on the curve tracer and find what available parts had the closest specs to the average I got. More parts = better average.

Damn I miss using that thing. Older than dirt but still worked a treat.
 
Oh I agree it's a HUGE pain in the arse. Possible, but certainly not fun, moreso if you're analyzing a failed one. Best method I had when trying to trace parts years ago was analyze as many as I could on the curve tracer and find what available parts had the closest specs to the average I got. More parts = better average.

Damn I miss using that thing. Older than dirt but still worked a treat.
Yep, more data is better! Although, if the existing transistors have a critically narrow operating range, they would have been carefully selected for that - meaning one data point might be very valid for choosing new ones. But you wouldn't know that until you tested them all, and possibly tested all of the transistors in multiple units.

I love old test equipment. Seems to have slightly less accuracy than new stuff, but has far fewer operational issues than the new stuff. Plus most of it looks way cooler than the new gizmos. I used to have a WWII Navy issue VTVM in it's own waterproof case. That thing was pretty deadly accurate, and I never once serviced it. Sold pretty much all my testing stuff a while back. I recently had to borrow my consumer-grade B&K tube tester from the friend that I gave it away to several years ago. At least I know where most of it is if I need it.
 
Yep, more data is better! Although, if the existing transistors have a critically narrow operating range, they would have been carefully selected for that - meaning one data point might be very valid for choosing new ones. But you wouldn't know that until you tested them all, and possibly tested all of the transistors in multiple units.

I love old test equipment. Seems to have slightly less accuracy than new stuff, but has far fewer operational issues than the new stuff. Plus most of it looks way cooler than the new gizmos. I used to have a WWII Navy issue VTVM in it's own waterproof case. That thing was pretty deadly accurate, and I never once serviced it. Sold pretty much all my testing stuff a while back. I recently had to borrow my consumer-grade B&K tube tester from the friend that I gave it away to several years ago. At least I know where most of it is if I need it.
I still want to get my hands on a tube tester (and a variac) so I can eventually start restoring the (now four) antique radios I have. Problem is there's a wide variety of tubes between them so I need to find one that'll cover at least most of the different tubes they use. Doubt I'll find one that'll cover ALL of them, but most would at least be a start :lol:
 
Interesting cottage industry. You could partner to diversify. You do the radios and I'll do the carburetors.
Screenshot 2025-02-20 at 08-49-10 zenith transoceanic radio - Google Search.png
Screenshot 2025-02-20 at 08-50-57 carter avs - Google Search.png
 
Not gonna lie that Zenith is cool as hell
I have one just like in the photo I pinched. It runs...

I also have an AVS like the picture... That one's broken :doh:
 
I still want to get my hands on a tube tester (and a variac) so I can eventually start restoring the (now four) antique radios I have. Problem is there's a wide variety of tubes between them so I need to find one that'll cover at least most of the different tubes they use. Doubt I'll find one that'll cover ALL of them, but most would at least be a start :lol:
The B&K 707 that I gave to a friend (and am now borrowing) is a pretty solid basic tube tester that can handle a very wide array of tubes. You can find them (and similar “consumer” grade units) for pretty cheap. Maybe $100-300. I’d recap it immediately and see how it does.

Plenty of old AA5 circuit radios practically being given away these days. Just be careful since they aren’t transformer isolated.

@cruiserdan Now there’s a business idea … tubes and carbs.
 
tubes and carbs.
Tubers gonna tuberate?


:hillbilly: :doh:


Annnnyway, @Spook50 : I too may have a junk 3FE ECM that I might be willing to lend to the cause. When I was trying to get my current rig running, I checked every system and finally ended up swapping in my spare ECM - and that's when it finally fired. I've been meaning to try swapping the old one back in, just to see if it's really junk, or if there was something else at play. If I plug it in and the engine doesn't fire, I'll yank it out and mail it to you -- for science.
 
Tubers gonna tuberate?


:hillbilly: :doh:


Annnnyway, @Spook50 : I too may have a junk 3FE ECM that I might be willing to lend to the cause. When I was trying to get my current rig running, I checked every system and finally ended up swapping in my spare ECM - and that's when it finally fired. I've been meaning to try swapping the old one back in, just to see if it's really junk, or if there was something else at play. If I plug it in and the engine doesn't fire, I'll yank it out and mail it to you -- for science.
Outstanding! Much appreciated!
 
Tubers gonna tuberate?


:hillbilly: :doh:


Annnnyway, @Spook50 : I too may have a junk 3FE ECM that I might be willing to lend to the cause. When I was trying to get my current rig running, I checked every system and finally ended up swapping in my spare ECM - and that's when it finally fired. I've been meaning to try swapping the old one back in, just to see if it's really junk, or if there was something else at play. If I plug it in and the engine doesn't fire, I'll yank it out and mail it to you -- for science.
Took me a minute to get the reference … to the sticker design my wife hand drew 🤣

F8EA28BA-CA77-485B-9401-B8021A82BE45.jpeg
 

Users who are viewing this thread

Back
Top Bottom