BOSCH CDI discrete components troubleshooting
 

Morning everybody,
i'm Giovanni from Italy and i own a MY69 911S targa. Firstly i would make my best congratulations to all the users supporting this wonderful BBS: it's unique in the world!

I'm trying to get an help on troubleshooting the discrete components of the 3-pin BOSCH CDI unit: maybe Warren can help me on this subject. I've read many threads on this subject, but i've found nothing on how to test the components inside.

My CDI has failed. I've opened and disassembled it and i have used my digital Voltmeter to check the following:
-All the connections and wires seems to be fine.
-All the resistors values seems fine (even if sometime different from what reported in the Bosch document Warren has posted - but it is understanable this units have passed many minor changes in the layout and the components used)
-All the capacitors, but one show infinite resistance. The one not showing infinite, reads 500 ohms approx. COULD BE A BROKEN COMPONENT? I measured it without removing it from the board
-All the diodes, checked with the diode function of the multimeter, seems fine showing some 0,6 V one direction and infinite the other direction
-the RCA 2N3055 (i've found the data sheet on the web) seems fine showing 0,5 V and infinite where it should.

I don't know how to test the Zener Diodes and the SCR OT104. ANY SUGGESTION?
Furthermore i can't find the data sheet of the two transistors T4018: so i wasn't able to check them. In case should i remove them from the board to test?
In case the SCR is defective: what's its replacement? It seems not so easy to find...

Last question: testing Diodes and Transitors using the Diode function of the voltmeter is maybe not 100% sure, but, at least, if you read the Voltage you could assume, at a 1st stage, the component is fine. Is there any other way to test the component?

Let me finally apologize for this long message and for my english hope youall understand.

Thanks in advance for your help

Giovanni

I don't have a CDI but testing each component might not be usefull. Can you apply 12 volts and check each sections functionality- high voltage, trigger etc.

Hello Giovanni, and welcome to the board!

The T4018 transistors were replaced by TI 2N2905A transistors in later production.

http://212.57.231.17/datasheetarchive/Datasheets/Others/DS7059.pdf

The SCR was replaced by heavier-duty versions in later production, reducing a tendency for them to short out! Good replacements are International Rectifier 16RIA80 or 25RIA80, or Siemens CS23-8, or the -10, -12, or -16 versions, if available. The SCR should not exhibit leakage with either polarity connection to an Ohmmeter from anode to cathode, and the gate-to-cathode diode should test as a normal pn-junction.

http://www.irf.com/product-info/datasheets/data/25ria.pdf

http://www.irf.com/product-info/datasheets/data/16ria.pdf

The Zeners are probably best tested with a semiconductor curve tracer, though cursory checks with a multimeter can be useful to find obvious faults. The 1N3003B Zener is an 82 Volt 10 Watt item, and seldom fails. If it fails open, the 2N3055 transistor will follow, almost immediately, and possibly the transformer, as well, if the transistor shorts! If the Zener shorts, the transistor usually won't blow, but the transformer winding #1 will be burnt or discolored ... it depends how long the ignition switch is left on testing for spark after the CDI unit fails! The transformer can be disassembled and rewound with ordinary magnet wire.

http://212.57.231.17/datasheetarchive/Datasheets/MicrosemiCorporation/DS59317.pdf

The CDI unit can be rewired with flexible, longer [Teflon is preferred for durability and high-temperature characteristics, 20 ga or 22 ga wire] leads from the chassis to the printed circuit board ... allowing removal of the board onto an insulated bench for testing while powered up ... with an oscilloscope and multimeter. The circuit can be broken down into the following sections:

1. DC-to-DC inverter and high-Voltage supply

2. Discrete transistor trigger/logic circuit

3. High Voltage storage capacitor and switching circuit

If you have access to an oscilloscope ... a simple curve tracer can be built from misc. parts, including a transformer with a secondary winding with anything from 24 Volts to 60 volts ... here are some scans from Hewlett-Packard <b><i>Bench Briefs</i></b> and <b><i>Elelements of Electronic Instrumentation and Measurement,</i></b> 3rd Ed., by Joseph P. Carr. The homemade unit can be just as useful as commercial units such as the RCA Viz 'Circuit Tracer' and Huntron Tracker!

http://forums.pelicanparts.com/uploa...1090701545.jpg

http://forums.pelicanparts.com/uploa...1090701634.jpg

http://forums.pelicanparts.com/uploa...1090701736.jpg

http://forums.pelicanparts.com/uploa...1090701819.jpg

http://forums.pelicanparts.com/uploa...1090701901.jpg

http://forums.pelicanparts.com/uploa...1090701973.jpg

Good luck!!!

Thanks everybody for your suggestion.

I'm not in the possibility to perform the test you have suggested in the next days (i've customers here!!), but i'll come back with news as soon as possible.

Giovanni

Go ahead and replace the capacitors. Overtime, they can go bad (leak dielectric).

Warren and Jim thanks again for your support. Here by some progress:

1. tested wires and resistors: they all are fine
2. tested the 2N3055 (disconnected) using the diode function of my digital multimeter. It's fine, i think, showing 0,49 V between B and E and 0,48 V between B and C (red on B, black on E and C) and infinite all the other readings
3. on my PCB the two PNP transistor are marked ON200A ITT 930 (still case TO-39) and not T4018. I've tested them using the diode function without dissambling the two components from the board: i'm not sure results are reliable then. Anyway one is showing 0,46V between B and E and 0,67 between B and C; infinite all the other readings. This could be fine. The 2nd one is showing some 1 V between E and C (both directions) and infinite all the other readings. It looks like a defective component
4. Not able to test the SCR 'cause i can't understand which pins are gate anode and cathode. No way to find the answer on the data sheet Warren has posted. Any help?
5. MOST IMPORTANT: Warren can you post the next page (chapter 17-7) talking about how to test capacitors using a multimeter? I've realized the way i'm testing capacitors is not correct and the paragraph you have posted is not complete
6. Jim, i was thinking to replace all the capacitors (with the exception of the one by Bosh -not easy to find i suppose) for safety. In case of shopping, which type? Electrolitic, Poliester etc?
7. Diodes and Zener's will be tested soon

Unfortunately i've no access to an oscilloscope and even building a test circuit it's not an easy job. That's why i'm trying to complete the tests with a multimeter, when possible of course.

Well...step by step...hope it will be Porsche again!

Giovanni

Giovanni,

It sounds like you are making definite progress! Yes, that ITT transistor sounds bad. My suggestion would be to find a replacement that cross-references to a 2N2905 or 2N2905A.

As for the SCR ... the anode is the threaded stud, and the cathode is the larger and slightly taller terminal that is also tied to ground and the emitter of the T1 power transistor. The gate lead is shorter, and has a single wire to the printed circuit board.

As for the C11 thru C16 capacitors ... they are either mica or metal film types that are not usually affected by aging.

The Bosch main storage capacitor Cs is of an essentially obsolete type called non-polarized electrolytic, so it is a good idea to replace it. I use a mylar or polyester film type like the Panasonic E-series -- 1.5 mF at 630 Volts.

I would be very interested to hear any details about the slight component differences you mentioned that you found when testing. A picture of the circuit board would be nice, if you could post it, too.

Here are rest of the additional scanned pages on capacitor testing you asked for:

http://forums.pelicanparts.com/uploa...1091042988.jpg

http://forums.pelicanparts.com/uploa...1091043140.jpg

http://forums.pelicanparts.com/uploa...1091043276.jpg

Wow. Ask and you shall receive. Way to go, Warren.

What the hell... I thought after my last Physics class I wouldn't have to look at this stuff again. Guess I was wrong.

Doug,

It's a better read if you don't skip chapters 1 and 2:

http://forums.pelicanparts.com/showthread.php?s=&threadid=160537&highlight=bosch+ permatune

http://forums.pelicanparts.com/showthread.php?s=&threadid=173563&highlight=cdi+wa veform%2A

If you think part 1 was good ... wait until we get to the 6-pin CDI units and troubleshooting the discrete Schmidt Trigger circuit ... now where did I stash the current probe and extra IEEE-488 cables???

Warren you must think you have died and gone to heaven right now!;)

I can't believe this stuff!! A grounded extension cord confuses the he!! out of me. Old dog, new tricks,,,,sigh

"The Bosch main storage capacitor Cs is of an essentially obsolete type called non-polarized electrolytic, so it is a good idea to replace it. "

I don't think so! Electrolytic capacitors are used for filters and not for
switching/commutating high currents, i.e. They lack the very low ESR
values required for energy transfer as used in a CDI. The peak current
of the capacitor exceeds 8 amps! Many of the components being used
appear to be standard (off-the-shelf) parts, but are really "selected"
parts with higher ratings because of the design or the desired reliability,
e.g. Just because a VW engine looks like a 912, most good techs wouldn't
interchange parts because the parts look the same.

Check out the Technical page on this web site ( www.systemsc.com )
under Ignition Systems for some theory on how a CDI functions.

The troubleshooting approach being taken usually results not finding the
problem, but only causing another problem. When the component count
exceeds a couple of components, a test circuit is required with a dynamic
test of the device using a meter/scope with proper inputs. Electronic circuits
can not be analyzed properly as would a transmission that didn't shift
properly by totally disassembling it and inspecting each component.

Warren

i've no words...really.

Some days ago i presume i know many thinks about my 911...but there's an hidden world to be discovered! If i 've time i'd like to read all the 3D in this forum!

I think going deeper in technical stuff and DIY approach are still a big pleasure for enthusiast! Being successfull on this job it's a real demonstration of what a forum can do at 1000s miles distance: this's globalization...

Tonight i 'll check capacitors and SCR and let you all know. A question is still open on my mind: testing component still assembled on the PCB does make sense?

The three sectors you've listed can be easily checked? My first idea was to supply a 12V to the CDI without loads and to listen to the normal switching noise once i've replaced the broken components and then check the unit in the car. Wrong?

About my CDI i can say that, looking at the Bosh document Warren has already posted, there're some differences on the layout and on the components that are used. I'll try to post some pictures and to list all the components. My CDI appears to be in original -never opened- conditions, though i can't be sure at all.

Another question: the CDI on the '75 to '77 Turbo 3000 is the same? According to the PET from Porsche they have a different part number...

Giovanni

If the capacitors are a problem then couldn't one utilize spray component cooler to isolate the problem?

OK
i'll try to highlight the differences found on my CDI when compared with the schematics posted by Warren.

Capacitors: they all are the same
Resistors: R3 reads 33 ohm (instead of 120) and there's one resistor more (total of 11 R)
Transistor: T1 is the same; T2 and T3 are ON200A ITT 930
SCR: the same
Diodes and Zeners: here i have some troubles. There're 2 standard (black) diodes (one is marked 1/1000); 3 have a transparent case (bosch?; 3 have a spherical shape; the last one looks like a small metal cylinder (assuming it is really a diode). There's another Zener (not in the PCB) marked as ZX82

The PCB number is: 1 228 310 014

The components layout on the PCB is really different, while the mechanical drawing of the PCB is approximately, but not completely, identical.

Transformer, Cs capacitor and R1+R3 resistor layout looks the same, while the 3 components (SCR, Zener ZX 82 and 2N3055) are all mounted on the metal plate fixed to the case through 3 screws.

I will try to post some pictures.

Now some progress. I've tested the SCR using the diode function of the digital multimeter: it shows 0,27 V between G and C (red on G black on C) and infinite all the other readings. What do you think?

About capacitors: i' ve a lot of concerns...
I've set the 20M Ohm range on my multimeter and shorted all the C for discharging. C13 to C16 looks fine. They all show 0 (short circuit), then the reading increase (like C charging) till 1 (open circuit). C11 and C12 have a different behaviour: most of the time they show 0 at the beginning, increase a little bit and stop: they never reach 1. Sometime reading is different: let's say it's not ripetitive.
Honestly i didn't take care about the Red and Black orientaton touching the capacitors...I'm thinking to replace all of them.

I need to test Cs, diodes and Zener's.

So far i wish you all a pleasant we!

Giovanni