Calling CDI/spark box experts

[reachme]

Yes I could and probably need to replace my 3 pin CDI box (in a 71 T) but I'd really like to find out what is wrong and if I could open it up and fix it myself. Car sat for a few years before I bought it and I had it "working" up until a few weeks ago but I had fuel issues I had to clear up first so just back to starting it again.

I'm getting a good 13.5v to the box, it's generating a tone, connection to distributor is good to trigger it, ground is good. When I crank the engine I read I'm supposed to get 500v going to the positive on the coil but I only see about 250v on the ohmmeter and it drops a bit with each crank.

I'm thinking that is not enough for the coil to then create a spark.

Does anyone know what DC voltage I should expect to see on an ohmmeter for this box when cranking? If this is too low does anyone know what I could replace in the box or what I can test further? Does it depend on the coil it's matched to (pic below)? No I don't have a oscilloscope and the price of entry is kind of steep. I'm really looking to learn here so thanks for all responses.






Here is the bottom of the coil

[mysocal911]

Originally Posted by reachme

Yes I could and probably need to replace my 3 pin CDI box (in a 71 T) but I'd really like to find out what is wrong and if I could open it up and fix it myself. Car sat for a few years before I bought it and I had it "working" up until a few weeks ago but I had fuel issues I had to clear up first so just back to starting it again.

I'm getting a good 13.5v to the box, it's generating a tone, connection to distributor is good to trigger it, ground is good. When I crank the engine I read I'm supposed to get 500v going to the positive on the coil but I only see about 250v on the ohmmeter and it drops a bit with each crank.

I'm thinking that is not enough for the coil to then create a spark.

Does anyone know what DC voltage I should expect to see on an ohmmeter for this box when cranking? If this is too low does anyone know what I could replace in the box or what I can test
further? Does it depend on the coil it's matched to (pic below)? No I don't have a oscilloscope and the price of entry is kind of steep. I'm really looking to learn here so thanks for all responses.

There are many threads here on Pelican on this topic. Use the search function.
If you need to have it rebuilt, Pelican Parts provides this service.

By the way:
You should not be able to measure voltage on the coil. The output CDI signal is a pulse.

[timmy2]

Where are the original braided grounds that run in the CDI harness to the base of the coil mount and to the distributor?
The greenish wire to the ground terminal doesn't look correct.
Bad grounding can cause lots of problems.

[mysocal911]

Originally Posted by timmy2

Where are the original braided grounds that run in the CDI harness to the base of the coil mount and to the distributor?
The greenish wire to the ground terminal doesn't look correct.
Bad grounding can cause lots of problems.

Good point. It appears, though, that the braided strap is providing the necessary
ground, as the CDI is buzzing. Not having the shielding from the distributor can
cause missing/poor running. It should still make a spark each time the points
open.

[timmy2]

I was wondering where the green ground wire was going. Hopefully to a location near the coil.
And why the originals weren't in place.
What most people don't know is the black and the blue/yellow wires are each run inside the separate grounding braids in the sheathing, to act as a shield as well as the ground.

The '71 CDI harness should look like this one with two solid ground locations. One at the coil and one at the dizzy.

[mysocal911]

Originally Posted by timmy2

I was wondering where the green ground wire was going. Hopefully to a location near the coil.
And why the originals weren't in place.
The '71 CDI harness should look like this one with two solid ground locations. One at the coil and one at the dizzy.

Nice cable! Too bad he doesn't have your cable. That would eliminate at least one unknown.

[ischmitz]

As others said the output of the CDI box is not a simple DC voltage to the coil. The coil's primary winding is providing a path to GND to charge up the internal main capacitor of the CDI box. The coil acts as a transformer. When the CDI box releases a spark the other side of that internal capacitor is momentarily shorted to GND and that produces a current pulse through the primary of the coil. This pulse induces a secondary pulse that that produces the spark.

(It's over-simplified and Loren will pick at every letter, but never mind.....)

Not sure how you measure a voltage at the output with the coil. If the coil was disconnected it is because the internal capacitor has a non-zero leakage current (it partially acts as a resistor) and your voltmeter is high impedance and picks that up. It doesn't mean the box is necessarily bad. With the coil connected you should see nearly zero due to the low impedance of the primary coil winding.

Switching gears now: What is the problem you are trying to address? As a first step in troubleshooting you need to let us know a problem statement.

Next, What makes you think the CDI box is causing the problem? Valid tests are swapping the box with a known good one to see if there is a change or using this box in question in a running car to see if a condition transfers with the box. If the CDI box has been identified as root cause to a problem here are some general recommendations:

- Make sure you have proper GND and then you can do some basic tests: As a first test you can manually trigger a spark by closing the points and observe if a spark plug connected to the center electrode of the coil will produce a nice blue spark.

- More sophisticated tests require measuring the output of the DC/DC converter in idle and under load and that requires some more advanced test equipment. The spark voltage can be measured with an adjustable spark gap.

Ingo

[reachme]

THANKS ALL for the replies. I am currently trying to translate some of ischmitz excellent response to english. .

The car is not getting any spark. I have read all threads I could find, I grounded spark plug and connected directly to the coil output and no spark (3 diff plugs) and yes this did work a few weeks ago. I didnt think to change gap to test for lower voltage-i'll try that. I got continuity between both primary and secondary on the coil but resistance # was not what Pelican threads said it (should) be but there was disagreement and since this worked a few weeks ago Im inclined to not suspect that initially.
I checked for wire shorts between CDI and coil, distrib and CDI, power to CDI and gnd (not CDI ground connector coming out the front-I thought that went to the tach. That was a good point-ill check that.)
Not sure what that second wire is connected to coil positive, or where it goes (or second wire off distrib)
The issue as I see it is somewhere between a whistling CDI and a non sparking coil. I will test
1. changing plug gap directly off coil
2. CDI ground

I know I can replace to test but if someone could point me to those more sophisticated tests I would appreciate it. Or if someone lives close I can try the replacement test.

thanks all!


Sent from my iPhone using Tapatalk

[ischmitz]

Maybe add a location to your signature.

Sorry for maybe being confusing with my previous explanations. Bottom line the output isn't to be tested easily. Further an adjustable gap is a tool that looks like a screw mounted to a wooden board. It's used instead of a spark plug. It is used to increase the gap to several mm. As rule of thumb for every 10,000 volt you can have 10mm of gap for a spark to form. This depends on the relative humidity but is one way to get an idea of the level of HV being produced. There is no need to reduce the gap of a spark plug. Even an impaired box will produce a spark with a gap of several mm.

The wires of the primary could be for an aftermarket tachometer. The distributor should have one wire to GND and one as output to the CDI (trigger)

And thanks for clarifying the issue: engine doesn't start due to no spark.

Your test with the spark plug connected directly to the coil excludes the cap and rotor from being root cause. Next, disconnect the CDI and measure with an Ohm meter inside the CDI harness connector and the other lead connected to engine GND in lowest setting. One outside pin needs to show the resistance of the primary of the coil and the other is open or zero depending on whether the points are closed. The center pin will show VBatt with ignition on (we know that works since the box whistles).

If the above tests work out it could be a damaged trigger circuit inside the CDI or a bad coil. How different are the measurements of the coil from what they should be?

How is your alternator and regulator working and what is/was the voltage at the battery with the engine running. A bad regulator or bad battery is often damaging the CDI box. If you had any issues with overcharging/boiling battery I'd say your CDI box is damaged.

Ingo

[tirwin]

You should put your location in your profile so people know where you live. It appears under your join date in the left column when you add it.

[reachme]

really good point. I used to have it, will put it back in when back at the computer.. I'm 1/2 hour west of Boston


Sent from my iPhone using Tapatalk

[mysocal911]

1. A spark occurs on the three pin CDI when the points open.
Make sure that the CDI input from the distributor is at ground when
the points are closed and at about 12 volts when the points are open.
2. A bad coil wire could be your problem, so you need to check its resistance;
zero ohms if non-resistive wire.
3. The coil primary should have about .5 ohms. The secondary should have about
700 ohms. That's for the stock Bosch small black coil.
4. Eliminate all wires on the coil except the one from the CDI and the ground
wire on the other pin.
5. Find another coil to use as a tester. An old VW coil will work for test purposes.
You can buy a basic VW coil for testing at any auto parts store for $30 - $40.

[reachme]

1. A spark occurs on the three pin CDI when the points open. Make sure that the CDI input from the distributor is at ground when the points are closed and at about 12 volts when the points are open.
I am getting 12.66 from the CDI to the distributor trigger wire and I understood that the distributor grounds that to trigger a spark so to make sure the triggering was working, I put a whistling voltage probe inline between the distributor signal, and the CDI wire and as I manually turned the crankshaft it went on and off indicating correct triggering to the CDI box.


2. A bad coil wire could be your problem, so you need to check its resistance; zero ohms if non-resistive wire. Good point. I checked it, zero ohms, its fine

3. The coil primary should have about .5 ohms. I got .4, up to .5 once so I mark that as within spec. The secondary should have about 700 ohms. That’s for the stock Bosch small black coil. hmm, I get 117.3 consistently. That's way off. What causes a sealed wire to go bad? Looks like I have a coil out of spec.

4. Eliminate all wires on the coil except the one from the CDI and the ground
wire on the other pin. did and retested. Also the ground-its fine, tested to make sure. still no spark from the coil directly to a grounded plug.

5. Find another coil to use as a tester. An old VW coil will work for test purposes. You can buy a basic VW coil for testing at any auto parts store for $30 - $40. I have a "might be bad" coil from a 930 that was paired with a Permatune. It was pulled out during troubleshooting and measures 1.5 for primary and 11.72 on outside, can I use that? I tried it, no spark.

So the coil I have been using seems to be out of spec. if the numbers are correct. Not sure if testing with the turbo coil would work even if I did know it was good so I need to get a coil.
The CDI box is receiving trigger from distributor and wire is good. CDI is whistling (so receiving good power and wire is solid) and third connection from CDI goes to coil, wire is clear and I isolated it to the coil but still no spark from coil.
I don't know of any way to test the CDI output to the coil other then using a known good coil so that is my next step. In the meantime I'll try to borrow an "up to 83" coil so I can test and make sure the CDI is working.

thanks all, bad coil





thanks all, I'll keep you updated

[ischmitz]

You can make connections individually to the three-pin connector with three flying wires. Use spade male and female crimp connectors and make 4 5-inch extension cables. Then try the following and be really careful not to get shocked:

1. Connect a spark plug to the center electrode of the coil and ground the plug.

2. Connect the center wire from the harness to the center connector B of the box and the GND wire to the box GND tab. The box should now whistle with ignition on.

3. Connect the wire that goes to the coil primary to the A terminal of the box - you should hear a momentary quick change in whistle note (indicating the internal capacitor is charging up through the coil primary. Be careful here as that output from the box can carry dangerous voltage levels if the internal capacitor is marginal or short.

4. Lastly, take the last flying wire and connect it to the trigger input C of the box and short the other end to the housing of the CDI box a couple of times - this should release sparks. If not I'd say your CDI box is bad.

Ingo

[mysocal911]

Originally Posted by ischmitz

4. Lastly, take the last flying wire and connect it to the trigger input C of the box and short the other end to the housing of the CDI box a couple of times - this should release sparks. If not I'd say your CDI box is bad.

Ingo

Assuming the coil is good!

[reachme]

Just a followup to close this out.

Dave, your last words were key. I replaced the coil with a new one from our host and using the flying wire method I have spark! I'll put it all together this weekend and have her running again.

Yes I did open the CDI box to see if I could find something obvious. I didn't. There are multiple layers of boards in there with wires jumping between and when my effort to get to the heart of the issue became asymptotic I needed to stop trying to learn and swap a part.

thanks all for the troubleshooting help everyone.