Here is how I tested the MCR4-E Marelli magneto:
Testing the magnet: I turned the mag over by hand and should have felt a tug as the armature crossed over the maximum flux point - in fact, if you let go of the spindle it should rotate back. There was virtually no resistance, a sure sign that the magnet needed re-magnetizing.
Testing the internal armature timing: The internal timing of the mag should synchronize with the ignition points timing. This means that the points should open almost right after maximum flux, the point where the voltage in the primary windings is at the maximum, on full advance. (There are two maximum flux points 180 degrees apart but on a single only one is used).
Testing the pick-up: I also checked that the Bakelite pick-up was in full contact with the brass slip ring. Some mags have just a brass segment in the slip ring instead of the ring being entirely in brass - when the points open, this segment should be in contact with the Bakelite pick-up in both the full advance and full retard positions. I cleaned the pick-up contact and spring and made sure the spring had enough pressure to push the tip against the ring. There were no cracks in the pick-up body that could cause sparks to "leak" out and not reach the plug.
Testing for shorts: The next checked for shorts in the primary and secondary windings. I first put a G clamp around the horseshoe shaped magnet on the mag body to retain the magnetism when I withdrew the armature. If you don't put a "keeper" around the magnet, it almost instantly gets demagnetized!
To test for shorts wind some stiff copper wire around the slip ring with one end about 1/8" from the metal portion of the armature. This replicates the spark plug gap. I replaced the center points retaining screw. I then took a standard 12v motorcycle battery and connected the positive terminal to the center screw. I then touched the wire from the negative terminal to the armature body and got a healthy spark jumping from the end of t he copper wire wound around the slip ring to the armature body. This showed that both the primary and secondary windings had no shorts. Most of the time the thicker primary winding wire should not short - if there is no spark it usually is a short in the thin secondary winding wire. If there is a short, rewinding by one of the mag reconditioning specialists is the only solution.
Testing the condenser: I examined the faces of the points - they were smooth with no pitting, a sign of a duff condenser which is hidden inside the armature. Another test is running the bike at night with the points cover off - any visible sparks and it is a sign the condenser is failing. The points, the contact breaker plate and the points earth brush and spring behind the plate were all in good condition as were the armature bearings.
To be safe, and to have the mag re-magnetized, I sent it to Mark's Magneto Service, (860-537-0376) at 321 McDonald Road, Colchester, CT 06415. He warned me that he had a three to four month wait, but winter's here and this is the best time to check it. Three months later it came back in excellent working condition. Despite the wait, I highly recommend Mark.
Testing the spark
I made a three point spark tester as shown in the photo on the right. Using a hold saw, I cut a circle in a piece of wood and inserted two metal screws from either end so that the tips were opposite each other and 5 mm apart. An earth wire is attached to one screw head (LH in photo) and clamped to any earth point on the bike. I turned down the head of the other screw and threaded on a spark plug cap adapter to which I attached the HT lead (RH in photo). A spark should jump the minimum 5mm gap when the bike is kicked over indicating an HT voltage of around 8,000v. To jump 6mm, 10,000 volts are required and to jump 8mm, 15,000 volts. The use of a screw allows for the gap to be changed. If it does not jump 5mm, something is wrong. The third contact, a metal nail, is to simulate the ionization that takes place around plug electrodes. This nail is offset about 0.40 to 0.50mm from the HT (RH) screw tip and is free standing in the wood. Of course, this set-up does not simulate the pressure build up in the cylinder head which might overwhelm a weak spark; however, it usually is not a problem if the spark jumps 6 to 8mm. Test with both full advance and full retard. Radco in the Vintage Motorcyclists' Workshop quotes a test where the mag should spark for 12 hours with a 5.5mm gap with the mag run at 3,000 rpm, at 85 rpm at full advance and 200 rpm at full retard.
Setting Ignition Timing
First, make sure that the magneto's internal timing (point of max magnetic flux) is synchronized with the opening of the points. The points should crack open just after the point at which you feel the most resistance to turning the magneto. Second, once the points open fully, set the max gap between 0.30mm and 0.40mm. This is most conveniently done when the mag is off the bike; if still on, a feeler gauge bent at 90deg., will help accurate gap setting.
Now, to time the opening of the points with the engine. If you have a manual advance and retard for the ignition (as I did) set the lever to full advance (pointed to the front of the bike). Turn the flywheel clockwise (engine rotates backwards) with the spark plug out and note when the intake valve opens. Keep turning till intake closes and then continue to turn clockwise till the TDC arrow on the flywheel aligns with the TDC marker on the crankcase. Now, rotate the flywheel anti-clockwise 105mm measured on the flywheel rim. Mark this position on the flywheel with a black marker for use later with a strobe light. Now, mount the timing gear on the magneto shaft taper with the gear retaining key in its slot to hold the gear in position. The points should start to open at this point. A powerful light probe wedged into the points cam will show a hairline of light when the points crack open. Or, use very thin tissue paper. If it does not open, rotate the timing gear till the points start to open 105mm BTDC on full advance. If you have an automatic advance / retard mechanism, set it to full retard. The points should open 30 mm BTDC. Either way, with the strobe light you can test the full range from 105mm full advance to 30mm full retard.
Once done check the spark strength using the three point spark tester shown above. Start the engine and using a strobe light make sure that on full advance the black mark scribed on the flywheel 105mm BTDC synchronizes with the TDC arrow on the crankcase and similarly, on full retard, the 30mm mark synchronizes with the arrow.
I usually start with the ignition almost fully retarded and once it fires, move it to full advance and don't bother manually adjusting when riding. Too many things to control in addition to my hand shift!!
Incidentally, the correct way to kick start a 500cc single (taken from Rick Parkinton of Classic Bike) is to first kick it over till you feel compression close to TDC. Then, lifting the exhaust valve lifter, gently turn the bike over till a little after TDC- the flywheel should let the piston just drop a little (if it turn over too much over TDC, start again). Then flexing the knee, jab sharply downwards, making sure your leg is not rigid or the kickback could be painful.