I remember this discussion. http://forums.pelicanparts.com/porsche-911-technical-forum/203100-6-pin-sc-turbo-cdi-unit-repair-documentation-4.html#post3656628

No arguing with Loren.

At the time the CDI in my '78 Targa failed, FULLY failed, it was on its way to its new "home" in McCall ID. So I was left troubleshooting without my O'scope, logic analizer, and documented "brains" (inclusive of a USAF 52-8 "Theory of Electronics" issued to me in 1957.

The earliest indications of impending failure I believe to have been the e-brake(?) light locking up and having to disconnect the battery to "unlock", turn it off. Over a period of the next few months it would just plain quit, but would restart amost immediately. That's when I first installed a new battery which seemed to help.

And then halfway to McCall, nephew driving, it began to quit eveny few miles. After a bit of "to-and-fro" on teh phone troulbeshoorting my nephew got it the last ~100 miles by disconnecting the voltage regulator once the engine began quiting and then reconnecting it once the SOC got too low.

Oh, and running with the headlights on with the regulator connected.

And...

I suspect the transformer core does not actually saturate but the current flow in the primary reaches the resistive limit. And the zener is ALWAYS limiting the secondary peak voltage if the engine RPM is low enough to provide more than full charge the capacitor.

My nephew sent the CDI off to Loren for repair, successful repair, so I don't know what components might have failed. But my guess would be that the SCR had gotten overheated so often it would "self-fire" at a capacitor charge voltage well below the "book" value.

"book" value for the zener's internal resistance is 22 ohms so the primary voltage might well rise another 22 volts with battery supply voltage high enough. That might add another ~150 volts on top of the 490 volt regulated capacitor charge voltage, well above the SCR "book" limit.

I finally got around to pulling this thing apart to see what was left underneath the
board.

The diagrams from the Bosch book show that ZD1 has some kind of mounting underneath
the Thyristor/T1 package.

Does anyone have a clear photograph of how ZD1 is mounted? The mounting on this unit
has been removed so I guess I'm making a mounting plate to relocate ZD1 back where it
should be.

andy

Pic
 

Here's a picture of what's underneath the circuit board:

http://forums.pelicanparts.com/uploa...1327464501.jpg

No ZD1 mount as shown in the Bosch documents:

http://forums.pelicanparts.com/uploa...1327464838.jpg

Anyone have a picture of what ZD1 is mounted on?

andy

Andy,

your box is stock as far as the location of ZD1 is concerned. The early boxes used that mounting style (on top of the power transistor) while later boxes moved the diode to the housing. In those the housing was tapped with a threaded throughhole and the diode was screwed into the housing from the inside.

No reason to relocate or do anything to your box.

Cheers,
Ingo

Thanks Ingo, I see now how it should be, almost...

http://forums.pelicanparts.com/uploa...1327488425.jpg

Ignore the wire to W3, it's not connected, the perspective of the photo makes
it look so.

Does anyone have any idea why the ground from the old ZD1 is connected
to W1E?

AFAICT from the schematic the cathode of the diode should be wired to
W1E.

I suppose it is possible that this unit never worked after "rebuild"

andy

and the dim light in my head goes on.

the original zd1 was reverse polarity to the replacement part. That is why it was not mounted as the factory did it.

Any suggestions for a replacement part for 1N3003B?

andy

Today studded power zener diodes can be be ordered with the anode or cathode as the stud. Maybe that wasn't always the case. I assume when the CDI box was first desgined Bosch had to use off-the-shelf components and that caused the mounting scheme in the early boxes. Later on they had enough forcasted demand to drive suppliers to make another configuration.

Before and after
 

Before cleanup:

http://forums.pelicanparts.com/uploa...1328788309.jpg

I modified the zener by turning the threads off on the lathe and then cut new M4 threads using a die. Shortened the entire package by snipping off the top of the stud so that I now mount it in one of the holes previously occupied by a screw holding the thyristor/transistor package into the body.

I then soldered on a new wire from the cathode of the custom package diode over to the collector on T1 and replaced and resoldered all the desoldered wires.

Blasted the body earlier and now back together:

http://forums.pelicanparts.com/uploa...1328788327.jpg

Some Gibbs keeps it looking nice.

Thanks to everyone who took the time to respond, couldn't have done it without you. There's still testing to do so every chance I'll need more help in the future.

andy

Why is the case ribbed, does it really get that hot?