Before there was BHKZ: SWB engine electrics

[304065]

Well, I decided to put everything I could find about early car engine electrical systems into a single thread. I'm going to be replacing my battery soon, and will be replacing various tired components with exact original replacements-- so thought I would do some research first.

It all begins with the wiring diagram. SWB cars have several features that set them apart. There is no BHKZ unit (All right, everybody calls it the CDI box. But Bosch-Hochspannung-Kondensator-Zündung just rolls off the tongue, no?). Secondly, they use a ballast resistor, which is always in the circuit between the ignition (on the same circuit as the fuel pump) and the coil, unless the starter is cranking, in which case it's bypassed and full battery voltage is delivered to the coil.



OK, so how does that translate into what's in the car? This is one of the hardest things for me to figure out-- the diagrams are good, but they aren't THAT good.

So it's off to whatever factory resources next. Here, the shop manual is helpful.


Image copyright Dr.-Ing h.c. F. Porsche KG. 1965

OK, now what color is everything supposed to be?

Well, for that we have to go to real-world examples. The best of which was the 37th 911 off the line, owned and lovingly restored by Alois and Estonia Ruf.



Lots of good detail here. We can clearly see the ballast resistor, with its plain black wires-- one to the coil +15 with the curls in it. Off the other tach terminal come the black to the points and the black/purple forward to the tachometer. Also, going into the bottom of the ballast resistor relay, we can see the red in the center, yellow on the right, black on the left and there must be a ground where we can't see it.

If you look in my car, which doesn't use the ballast, this is confirmed. (Pardon the oxidation!)


OK, so let's look at a few more from the Parade.





OK, do any of you SWB owners have photos of your own cars? It would be helpful in builidng a record. Also, if I've made any errors please let me know-- this is uncharted territory for me.

Mine (sad)

[304065]

Well, I was doing some work on the car over the holidays. The car has been running hot: even on a short blast up the mountains (it's a CAR, not a museum piece. . . at least until restoration. . . ) the temperature went to 210. Hence, a check of the ignition advance was required.

Now, what should the advance curve look like? The distributor part number is 0 231 159 001, it is a Bosch JFDR6. There are two possibilites for advance curves, depending on the manufacturing date code of the distributor. I went with the later one given the fact that my car is a 1966. Graphed in red are the actual values from my advance test.



Now, the 24 degrees advance (at the engine, distributor advance is half of that is not enough, the factory spec calls for 30 degrees at 6000 rpm with no load, and 32 degrees under load. Therefore, I figured the distributor was sticking. So, out it came.



That is a concours perfect towel from 1966 stuck in the hole.

On the HAZET Assistent for inspection.


Wear pattern looks ok, any comments?

Anyway, I checked the shaft for play, there was none whatsoever, both axially and radially. The "Oel" cup on the side was dry as a bone, as was the felt in the shaft above the circlip: these were liberally lubricated. I plan to ditch the Crane optical ignition and revert back to good old points and CDI, so further disassembly was held until I have the parts to continue.

I put the distributor back in the car, and managed to align it to #4 instead of #1. After a few minutes of cranking with no start, and a backfire that sounded like a black powder deer rifle I realized my mistake, reversed the rotor 180 with the Z1 mark opposite the notch in the blower, and she started right back up.

Now, with the distributor properly lubricated, I think I found one source of my overheating problem. With the distributor body turned all the way clockwise, the projection on the bottom of the distributor made contact with the distributor clamp stud, and so it wouldn't turn any further. This was approximately equal to TDC: the rotor lined up with the notch in the distributor body. Any turning of the distributor counterclockwise increased the advance setting. The range of adjustment seemed enormous: a slight turn of the distributor resulted in about 25 degrees of advance at IDLE. Obviously I ran the engine like that for only a few seconds, I'm surprised it started at all.



I think the culprit is the pickup for the Allison optical ignition-- it was moved clockwise around the distributor cup slightly-- and when moved counterclockwise, toward the rear of the car, the advance value diminished. I didn't install it, the PO did, so I have no idea whether it's the correct one for the car or not. So, out it comes, and in goes a two-piece points set to trigger a CDI ignition. Once I have the original setup I can determine whether the advance curve is correct or whether the springs have stretched, resulting in excessive advance under high rpm/load that would be causing high oil temperatures.

Stay tuned!

[at7driver]

well John,

I feel inadequate adding my pic.. but here it is.. I will be saving your research as my car goes back together.. What a great job you are doing.

[Early_S_Man]

John,

Great work, BTW! It thrills me to see how well that 39+ year-old cast-iron body '159' distributor has held up over the years! The workmanship put into that distributor is a lost art! Even the silk-screened, aluminum, anodized part number plate is a work of art!

I believe there is a typo or transcription error in your post regarding your distributor part number ... your excellent pic clearly shows 0.231.159.001 as the pn and that jives with the advance curve page you posted from the service manual!

I have found that off-the-engine tests as specified in Bosch test documents are less worrysome/troublesome than running the engine, and you can test the full limits of the specifications. Kind of difficult [and nerve-racking] to get a 911 engine to run steady at 7000 rpm ... if you get my drift! A homemade distributor machine is a doable project, complete with CDI, degree wheel, dwell/tach, and ignition scope ... that way you can get precise data points at 100 rpm [engine speed] increments, and collect transient data, too ... with a multi-channel digital storage scope, should you be so inclined to do after-the-fact data analysis of the distributor performance!

There is certainly nothing wrong with your data collected from 2000 rpm to 6000 rpm ... but it would be useful to know what the end-point data is, too! Your curve is within the limits of the factory chart! The identical advance at 2000 rpm and 3000 rpm [engine speed equivalent] is a bit of a puzzle, though ... sticking centrifugal weights, maybe, as you suspected???

Did you do any disassembly to inspect the centrifugal weights and springs [and, specifically, the phenolic wafer rubbing surface for the weights that typically dissentigrates after 20+ years]? The Stoddard Teflon replacement NLA.602.102.00 part for 356 distributors [#12 on the Stoddard page below] used to fit with just a little trimming with an X-acto No. 11 blade ...

BTW, the Stoddard parts page has improved significantly, recently ... here is a page with many useful parts for our early distributors:

http://www.*************/shop/shopdisplayproducts.asp?id=125&cat=Engine+Electrical

ps,

I applaud your plan to retrofit a Bosch CDI unit!

[Grady Clay]

John,

Warren is right on target as usual.

You have the conflicting requirements of originality and proper running. I’ll suggest you temporarily put the Bosch CDI in place of the fuel pump. Install a ’69 911E/S fuel pump bracket on the front cross member with a ‘70T Bosch fuel pump. That will work without any permanent modifications. If you want to do more, installing a return fuel line might be appropriate.

For show purposes you can install a non-functional Bendix appearing fuel pump in the original location (fuel just flows through it) and hide the CDI.

For original appearance you want to keep the coil on the relay panel. Remember, the Bosch CDI uses an ignition transformer that is shorter than the original Kettering ignition coil. You will need to make a “stock appearing” wiring harness for the CDI.


I agree with Warren about distributor testing. I’m sure you can find some old “auto-electric” shop with a Sun distributor test machine. You can take your complete distributor, CDI and ignition system and set it up with six spark plugs on the machine. They are designed for this. This way you can document the functioning from static and idle to above 6000 crank speed (3000 distributor) of the complete system.

His also allows you to rebuild your distributor and confirm the proper operation.

I would love to have use of a nice Tektronics memory scope to find an intermittent problem. Only $65K for a not top of the line instrument and you think Porsche parts are expensive.


What do you think about painting/dyeing the battery black? The white plastic is a dead giveaway. Painting it could look like the original vertical ribs. I wonder if we can find the original logo and have a decal made?

Best,
Grady

I’ll take some pictures here of the wiring/ignition for you.
G.

[304065]

David, there's nothing wrong with that picture! Thanks for posting it. It shows that you still have the original terminal block, "wye" connector, ballast resistor, resistor relay (the black one with the rounded corners) and voltage regulator. Also, the coil is interesting, it appears to be a Bosch, with the original rectangular sticker in the right place (but missing)-- but I think the top ought to be brown.

One of the wires from the ballast resistor has been spliced, but that can be easily fixed by opening up the resistor box, desoldering the correct wire, and re-soldering in another black one. Other than that, everything looks like it can easily be saved!

Warren, thank you for the part number correction, I have revised the post above accordingly.

I needed to get the car moving by sundown so unfortunately couldn't disassemble the distributor. But the diagram you posted in another thread, distributor lube...

along with the instructions for removing the circlip, will be a great help. I have been looking for a diagram of the early distributor showing the "two-piece" points set but so far haven't been able to come up with anything. However, Glenn Ring, VW enthusiast, has an excellent web site with pictures of his restoration and collection of early VW distributors!

http://www.glenn-ring.com/010/

I'll look around for a shop with a distributor machine. It is quite disconcerting to stand over the running fan with the engine at 7000 rpm, earplugs notwithstanding!

Grady, thanks as ever for the advice. I was thinking of mounting a CDI somewhere in the engine compartment, perhaps up on the firewall, where it would be concealed during everyday use by the air cleaner. Assuming proper shielding of the leads, the run could be longer, allowing me to mount the box elsewhere in the car, although I suppose I could forget about ever using the radio, with an 18Khz 460-volt pulse running through the car!

That fuel pump needs to be investigated next. It is running direct to the Webers with no regulator inline. So I need to get an inline pressure gauge for diagnosis purposes. The instructions for cleaning the Weber float bowls you posted on another thread will be useful in the next few months, when I will be able to remove the carbs for a proper inspection and possible rebuild.

I agree with both of your suggestions about a scope. Fluke makes an automotive "Scope-meter" that isn't TOO expensive and might be available used. Also, the PICO data logger folks make an inexpensive, PC-based, USB oscilloscope that appears suitable for automotive diagnosis purposes. Perhaps next Christmas?

I do agree that I need some more sampling points for greater accuracy. After I get things back to original I'll check the advance curve again with many more data points. Thanks for the suggestions.

Grady, p.s. The Sonnenschein's battery is light gray, it could definitely be painted black, though, and an original-looking varta circular decal added-- points for effort might even be awarded!

[304065]

Some Bosch numbers. . .

Shim kit. . . 1 237 010 007

Insulator kit . . .1 237 011 500

912 hardware kit with correct screws . . .1 237 010 012 (not strictly required for this distributo- but I wanted to be assured of correct hardware)

Two-piece points . . .1 237 013 060-850

Condenser (with light green wire and ring terminal) . . .1 237 330 067

From UP-FIXEN der Porsche, Volume III. Total credit to PCA.

[304065]

OK, I have unified the diagrams and eliminated the nonessential parts. Of course these can be put back for show.

[304065]

Okay, I didn't have any blue/yellow wire handy, so I just used blue! It's only a prototype for testing anyway. . . I'll go back and change the diagram.

Grady, Warren, I managed to make everything fit on the original '66 electrical panel. This panel is made of a sheet of aluminum which has been nibbled, and steel U-nuts. I don't know whether the U-nuts were cad-plated or not, I have seen a couple examples where they were, but the factory manual shows the entire plate dark, as one would expect if it were painted with a thin coat of black as a unit.

Anyway, I stripped the panel and re-painted it with Duplicolor Semi-Gloss black, which is the "right paint" for most of this stuff, just the right degree of gloss, and John Paterek says it's the same finish as on the engine tin.

Now, you remember that for street non-concours purposes, I will be operating without a ballast resistor or resistor relay. This freed up a lot of space on the panel to mount the fuel pump in the original location, the BHKZ unit, which is held to the panel with some oversize washers that "clamp" it on, and the CDI-compatible Bosch blue coil, which had to be moved toward the rear of the car by loosening its clamp a little. Now, that left little room for the large Bosch VR, so I used one of my Transpo adjustable solid-state VR's, which fit perfectly. The harness was a little short to go around the BHKZ and up to the VR plug, but I plan to fabricate a new one anyway as soon as I can source the right connector-- the old wires are work-hardened and have the inevitable overheat marks from 40 years of service!



Next, I finished the distributor, which needed a repaint, and I also changed out the various washers and shims inside. I'll be posting another question I have about advance weights, I am only 99% certain I have the correct orientation for them, but I have to draw a diagram as my assembly pics aren't handy.

Warren, the old phenolic rubbing plate was in NOS shape! I couldn't believe it. I still think teflon would be a better material, however.



Now, I was going through some spare parts from the PO, and look what I came across! The original coil and a brown distributor cap! Look at the semicircular notch on the cap, that is intended to clear the circular insulator on the side of the distributor. When I restored mine, I slightly ground down the insulator so it would clear the cap. This enables me to use a modern cap for street driving.



Anyway, I'll let you know how it works when it's fired up! I'm looking forward to the test.

[Fidalgo911S]

I noticed you guys are talking about "159" disributors. I have my out while the top end is being re-built so went to look before I ordered points. The number is "0 231 121 006 JFR 6".

No points are listed here at Pelican for this distributor. Maybe that unnamed place in Ohio has them. Am I screwed? Or will the other two-piece points work?

[304065]

Fidalgo,

Oddly, you have a "third style" distributor that was used on the later SWB cars. If you look in the factory shop manual on the page with the ignition curves, you will see that there are three advance curves-- two for the 001 (early and late) and then the 006. Anyway, I would think that the two-piece points from the 001 would fit fine.

http://www.pelicanparts.com/cgi-bin/ksearch/PEL_search.cgi?command=show_part_page&please_wait=N&make=POR&model=911E&section=ELignt&page=2&bookmark=7&part_number=1-237-013-060-M14

[Fidalgo911S]

John,

Here's a picture of my partial reassembly. The motor is out going through a top end rebuild, so the alternator connections are missing.



Nate

[304065]

Fidalgo, that looks great. Very original! Are those metric adapters for the banjo fittings for the fuel pump? I think (although the Ruf car contradicts this) that the coil sticker is supposed to be rectangular, all I can find are oval ones for 356's. But we can easily make one these days with computers!

You and I are headed the same direction-- I'm going to stash a CDI probably under the passenger floorboard and extend the wires back. For testing I've got it in the rear, see below.

So I did some experimentation with my refurbished distributor. I put everything back in the car and connected all the wires for a test. I used my multimeter to set the distributor so that the points were just beginning to open on #1 cylinder at TDC. That's the "basic setting" which gets you in the ballpark.

My results appear below. I know, I still need more data points, particularly in the 2000-4000 rpm range to see when the second weight flips out. My ears are still ringing from the 6000 rpm test, do NOT try this at home.

Now, this is a digitized version of the factory curve for the 159 001. The thick black lines represent the factory low and high limits. The red line is my measured setting. And the green line is the factory high limit assuming a basic setting of five degrees, minus three degrees for installation variables.

Now, Roger W. Chaney, in volume III of "Up-Fixen Der Porsche," on page 87, says

However, an exact 2:1 ratio does not exist between the installed and uninstalled advance curves. In rpm the ratio is mechanically fixed and is true, but the advance degrees are not in a true ratio due to "installation varable"; i.e., engine vibration, temperature, initial or "static" setting and variances in manufacturing tolerances between distributors. A good rule in going from uninstalled values is double uninstalled values, add basic setting, then subtract 3 degrees from the new curve values.

Anyway, looks like too much advance, right? I have to think that the springs have fatigued over the years (40). I either need to get different springs or try bending the tabs internally to limit total weight, but I want to do some more experiementation before doing anything.

By the way, driving the car with the Bosch CDI was AWESOME. It was like a completely different car. The engine would idle as low as 400 rpm (not ideal, I know, but I was working on the carbs and it would occasionally slip down) and the plugs remained clear. I haven't gotten the tach working yet, but that is a simple matter. Anyway, I highly recommend conversion to the Bosch CDI if anyone still uses Kettering-ignition (both of you guys!)

[Fidalgo911S]

Good info John. In fact I need to go over and re-read it. Your original coil looks exactly like mine. If you look closely there is a dark green rectangular "Bosch" sticker on top of the oval sticker. I only have about 1/3 of my rectangular sticker remaining (being this fastidious makes me feel like an archaeologist).

Anyway, I'd like to set up a CDI system. The car came with a "Sydmer Fireball" ignition system connected to it. I imagine it was a 60's setup since my old '67S had one also. I took that out and will be collecting the necessary pieces.

The fittings to the fuel pump are metric - solid brass too. They shine up well. More to follow as I finish up the restoration.

[304065]

Fidalgo,

Nice hoses and hose clamps. I'm doing that this weekend. What size hose and where did you get it?

I agree with you about the stickers. It looks like Bosch put a rectangular sticker over the top of the oval one on both of our coils. If I can find a good image I'll take a stab at creating a graphic that can be printed, in the manner that DarrylD reproduced the NLA alternator strap decals.

Speaking of alternator straps, do you have any rust on yours at about the 11:00 position, viewed from the rear? Mine has a big rust spot which just happens to be right below the edge of the engine lid opening. No wonder they covered it with a foil sticker and redesigned the decklid.

You have an American Bendix fuel pump with metric adapters sandwiched in to allow the metric banjo fittings to work. I remember your thread on fuel pumps, I will be putting what I can find in there.

Anyway, thank you for your contributions to the thread, it is great to have someone to compare notes with.

[Fidalgo911S]

I checked and found no rust on the alternator strap. PM'd you on the hose and clamps. Taking the car to the paint shop on Wednesday!

[mjshira]

impressive thread. nice work John. I was going to say "the new Warren" and then I noticed his post. Grady too. I need a SWB car, but then, Scruffy still needs more little bits to finish.

I'll just keep reading your posts and learning for now :-D

[Early_S_Man]

John,

Glad to hear that the CDI worked out so well!!!

I'm not sure that I believe the 40-year old springs have weakened!!! The German springs in my engine have impressed me with their lack of 'lost spring tension!' It seems to me that the cause of your excess advance is wear on the tabs that the springs are attached to ... your data shows too much advance from idle, initially, which leads me to believe that the tabs have wear that gives too much 'slack' in the spring and allows too much advance as the rpms rise as well.

My suggestion is to examine the tabs very carefully, and if wear is noticeable by the springs being loose at rest, and notching of the tabs ... I suggest building up a small amount of nickel-silver brazing rod [the kind used for 'hard-facing' repairs on agricultural plow blades, as well as brazing chrome-moly frames of tube-frame race cars ... on the outside of the tabs. The repaired tabs can be filed down [with great difficulty] if the build-up is excessive [as evidenced by a too-gentle advance curve afterwards] ... and you may get real familiar with the '159' disassembly procedure in the iterative process of fine-tuning the advance curve!

Good luck!

[304065]

Warren, thanks as ever for the advice. I'll check the tabs when I open the distributor to replace the phenolic plate with teflon!

There is a manufacturing date code stamped on the side of the distributor, in very small letters. I didn't have the camera handy to take a photo, but it's there, and I believe it was earlier than "609," I think it had a five as the leading digit. That would imply that the distributor is earlier than I thought, which would give it a different advance curve. Edit: IF it's accurate that the code is year, then manufacturing week, that would imply that the "609" changeover was in the first week of March, 1966. My gauges and temp sender are stamped 2.66.



Now, the green trace is as above, but assuming five degrees initial advance. Now, when I took my measurements, I used a basic advance of TDC. So I still have too much advance in the curve. More measurments are required: I really need to fab up a distributor machine.

[Early_S_Man]

John,

Thinking back ... there is also a possibility that wear of the pivot pin or the hole itself in the weights is also a factor in too much initial advance and 'creeping' upward of the curve as rpms rise.

The primary reason I mentioned the tabs is that I have read magazine articles suggesting bending the tabs ... and that seems very imprecise and crude to me ... not to mention the possible side effect of breaking off the tab! I just wanted to offer a safer alternative with a bit more precision to try!

For wear in the weight holes ... drilling or reaming to a larger size and fitting a Nylon or Teflon bushing around the pin seems like a reasonable fix.