Blown injector driver inside Renix ECU

[lawsoncl]

Probably an internal Motorola part number that can't be easily cross referenced then.

I take that back. I checked the NTE cross-reference at http://nte01.nteinc.com/nte/NTExRefSemiProd.nsf. This looks like a likely match is an NTE270 http://www.nteinc.com/specs/200to299/pdf/nte270.pdf. It's a darlington.

If you don't have an electronics shop nearby, a TIP-120 from RadioShack might work in a pinch. It's rated for much less amperage and has lower thresholds than the NTE 270 suggests as the replacement. I'm not sure what the original was rated at, but I can't image the injectors drawing more than 5-amps continous.

 

[cygnus58]

good call Lawson.
what part number was that?

I just took my spare ECU apart and found it has
1 lm2931a
1 s886-2a
7 114-08
7 unknown Motorola number

Not so simple as I expected.

 

[mr_W]

There is something strange that I fail to understand.

If this is NPN transistor (which it is, I've checked with multimeter), how comes that collector is connected to injector output?? That would be true if jeep didn't have positive side switching (injectors are all permanently grounded).
In fact, all 6 emiters are connected to same point.

 

[mr_W]

Cygnus, I have only 6 motorola parts, and their collectors are wired to A1-A4 and B1, B2 which are inj. outputs.

My wrong, there are 7 of them. I'm so puzzled with this that I can't count anymore

 

[cygnus58]

if they are connected as you say, then they are certainly the "last" device in the output chain and common emitters makes sense to me. Sounds like positive thru injector, and current sunk to ground thru these enigmatic devices. And as Lawson says Darlington NPN, you should really try to get ahold of something and tack it in. The problem is, sometimes when semis are direct-coupled, more than one can be destroyed by a short.
Still wish I knew what number he crossed - I must have missed something.

why is it every time i post something on this subject , i feel more stupid each time?

 

[mr_W]

why is it every time i post something on this subject , i feel more stupid each time?

Why nobody knows anything about these after 20 years of existence? Enigmatic devices. I'm tracking the leads, and hopefully will have my eyes still working after I finish

 

[5-90]

I've been watching this thread - just haven't been able to really jump in yet.

I've tracked down the last patent holder (I think) to Actia Corp. in Elkhart, IN - but I've not gotten anything out of them yet. I may have to stop by next time I'm in the area...

Mr_W, any chance you can read that Motorola number a couple of times, write it down, and post it? My eyes aren't what they once were, and I'm not sure I'd get a good read off of that picture. If you get me a number, I can start washing it through cross-reference guides I've got access to and see what I can dig up.

Oddly enough, injector drivers are really the only "failure" I've heard of in RENIX, save one unit (that's the one that Boatwrench sent me in trade for a good one. I want to do FA on it later.) I've heard of two failed injector drivers (now three,) and I've got one of the old units stripped of its case and out in the shop somewhere.

The problem with knowing anything significant about these things is simple - it's a non-standard pre-OBD system that was designed by an overseas (to us) automaker and built by a company that does avionics (Bendix/King.) It was only used for four years - 1987-1990 - and the company that used it over here is defunct (AMC, purchased by ChryCo in 1988. ChryCo was purchased by Daimler-Benz around 1993, and has recently changed hands again...) So, it's work just to track anything down. I'm working on it tho - I'm rather fond of the RENIX setup, and I prefer it to OBD used in later years.

See if you can give me an accurate read on that Motorola number, and I'll see what I can do with it. I don't guarantee results, but I'll do what I can (if I can at least identify the part and get specifications, we can come up with a workaround...)

 

[lawsoncl]

Still wish I knew what number he crossed - I must have missed something.

I just read the 5849 off the picture, plugged it into the NTE cross reference. The second of three possible matches was a NPN Darlington with specs that look about right for this application. The other two were a diode and rf connector.

I'm pretty sure it would have to be a Darlington or a high gain (Hfe) transistos. I doubt any of the chips would have enough drive current and I didn't see any other smaller transistors on the board.

 

[mr_W]

Motorola numbers are
89
15
5849-11

Now one of them has number 35 instead of 15, an that is the one is also used as driver. Same as others.

I've desoldered everything from that part of PCB just to be able to track everything, and eventually make some schematics.

Addition:

S 114-8 (other 7) are PNP, so it could be that they are used in pair with 5849

 

[lawsoncl]

There is something strange that I fail to understand.

If this is NPN transistor (which it is, I've checked with multimeter), how comes that collector is connected to injector output?? That would be true if jeep didn't have positive side switching (injectors are all permanently grounded).
In fact, all 6 emiters are connected to same point.

That does seem odd. NPN would imply that the load is "upstream" of the transistor. I do seem to recall that the injectors had a common ground. Here's some circuit examples for the less electronics inclined. http://www.kpsec.freeuk.com/trancirc.htm

You sure you got the right transistor and this isn't one thats providing the ground path for something else like a relay somewhere?

 

[lawsoncl]

http://www.diy-efi.org/efi332/other.htm
Damn - not your part number. On to try to see what is is though..

You refrring to the injector driver?
http://www.datasheet4u.com/download.php?id=554794
http://www.datasheet4u.com/download.php?id=554794

 

[mr_W]

Now I am 99% sure that we got it wrong about figuring out NPN vs PNP. I did the same thing as cygnus suggested previously, and what happened to match my knowledge, and also what I found on other webpage. But it must be wrong with my multimeter.

They must be PNP type. I've tracked Injector Feed input, which gets +12 from outside. It goes to diode D6 (at heatsink), passes it, and then directly hit EMITER of Q1, Q2, Q3, Q4, Q5 and Q6. Their collectors are connected to A1-A4 and B1-B2.

 

[5-90]

"89-15" sounds like a date code - 15th week of 1989 (common marking on semiconductors.) What year was the rig again? I don't see it on the recap I'm getting with the "reply" screen - but mid-1989 to EOY1990 would fit under that. "89-35" is 35th week of 1989, so make that a late-1989 or 1990.

IIRC, the injectors are "hot-switched" with a common ground (I'm fairly sure of that - I put about eighteen hours, all up, into tracking a floating ground and rebuilding the injector subharness in my 1987 several years ago...) just as an FYI.

 

[mr_W]

They are PNP. No way otherwise.
And suddenly everything gets logical.

They all get power over D6, B is connected over SMD transistors on "mini boards" CM10A and CM10B.

 

[cygnus58]

Glad you said that first. My meter only deflected on the junctions with Red lead common to base. That's what I thought and you are right, that explains the common emitters.

Yes, Lawsoncl what number did you enter for the cross to the Darlington?

 

[mr_W]

So we need to find equivalent PNP transistor for couple of amps, but it must be darlington. hFE reading on my multimeter showed 215!

Note that I didn't use this hFE feature ever before, but I did check with common BC337 right now, and it had 422.

 

[lawsoncl]

So my swag of using NTE to cross the 5849 number was bogus, since it came up with a NPN. Do you have a meter that shows Hfe, that you could check the other good transistors and figure out if its a Darlington?

I was thinking about solenoid circuits and reading the pdf on the injector controller the EFI project was using. Its common to hit full voltage to open the solenoid and then a lower voltage to hold it. Faster response and lower heating of the solenoid. In that case, it might be less of a high-gain on/off switching and the transistor gain might be important.

Edit: just read your last post. An HFe of 215 doesn't sound like a Darlington. Aren't they usually 1000 or higher?

 

[cygnus58]

since you have them out of circuit mr_w. , compare junction readings with a known NON-darlington. they read quite differently. One more way to get to the bottom maybe. I believe the E-B read is higher (two junctions in series)

 

[mr_W]

lawsoncl,
the thing with PnH (peak and hold) system you described is that controller is more complicated, and while it has better performance, reliability is lower, and they are abandoning that system in favor of old and proved.

On the note of hfe, I am going trough catalogue of my local electronics shop, and looking at datasheets. Non darlingtons of that power (10amps for example - however I don't think there is more than 1-2amps going trough them) have hfe below 100. On the other side, darlingtons have 750 and up. I'm somewhere between.

However, it could be that hFE varies with operating factors, like current etc.. (higher the current, lower is hFE), and as multimeter uses very low current, readings could be off.

Now, what I have on my mind is BD911 (they have them, hFE is in 150 range), and it is non-darlington. Cost is approx $1.

Or should I find something that handles higher voltage? I'm thinking about that flyback from injector coil. Spike can be really high.

 

[mr_W]

Just for reference, before I put all this together again