Magneto Slicer

[SED]

Anyone have anything on Marelli rotating coil magnetos?

I've got a parts book from 1931 and a blog post from a Super Alce restoration (very helpful!):
http://superalcerestoration-j2maria.blogspot.com/2010/12/

The slicer had been starting and running OK, but it has this persistent backfire at trailing throttle and hunts and surges at low throttle openings.  These are classic carb mixture problems but nothing solved them.  And then it seemed to have more trouble idling...  So, after trying 3 different carbs I started to think what Chuck and Kiwi Roy say is true - 98% of carb problems are electrical  (slow learner...)

The classic magneto problem is dying capacitor (aka condenser) that fails when hot.  There are no Marelli rebuilders to be found so like so many of my projects I start thinking I can do it myself      

Problem is I don't know what they look like on the inside.  My buddy Carlo decides to loan me his non-working old Marelli mag that is probably 80-90 years old for me to experiment on.  "Hey, watch this!" 

Note the broken spark lead and the loose horseshoe magnet.




Thought I'd post this essay for others to see what it looks like inside.  (I'd post on the Google singles site, but it doesn't handle photos and text as nicely as WildGuzzi). 

Remove points by pulling the center bolt.   Points are on a keyed taper that must register correctly before tightening the bolt.  The center bolt is the electrical connection to the insulated point.  The carbon brush (7mm diameter) is the ground contact (along with the taper with a path through the bearings).




The spark lead-out must be removed before pulling the armature.  It's counter-intuitive, but the fastening screw is at the bottom.  Looks like someone tried to pry it loose and broke the Bakelite fitting.




Pull grounding brush (6mm diameter).




Put a "keeper" on the horseshoe magnet that has a cross-sectional area the same as the x-section of the magnet.  This means the C-clamp I used was too small    Old steel magnets will instantly lose most of their magnetism without the keeper so it has to be put on before removal, kept in place and removed only after the magnet is reinstalled.  These old magnets will also lose magnetism over time, so the need periodic remagnetizing.  Apparently better steel magnets came out in the 1920s, to be replaced in the 1930s by Alnico magnets that keep their magnetism much better. 
All this gibberish gleaned from the interwebs.   

The mag had been apart before without a keeper and the magnet seemed very weak so this "keeper" business is mostly for practice.

Note that under the horseshoe magnet there appears to be an Alnico magnet cast into the body.  Note also that if the C-clamp is to be used a "keeper" it must be in direct contact with the steel of the magnet - not on the aluminum body.




More on magnet keepers - contact between the magnet and keeper allows the magnetic flux lines to pass through the keeper.  This is why it needs a x-section similar to the magnet itself.  Air gaps (aluminum is non-magnetic so it is considered an air gap) minimize the magnetic flux and so weaken the magnet.  I've read that a keeper is needed when pulling the armature from a mag - probably because the steel core continues the magnetic flux.  Rebuilders always remagnetize the compete assembled magneto after a rebuild.   

So either correctly use the right size "keepers" or have the unit remagnetized when done.

I don't know this stuff - I just read it and it makes my head hurt.   

More soon...

[SED]

The best source of magneto rebuild information for rotating coil mags is a thread posted by MagnetoMan on the BritBike forum:
http://www.britbike.com/forums/ubbthreads.php/topics/446733/restoring-a-rotating-armature-magneto

It pertains to an old Bosch magneto and newer Lucas magnetos and the pictures and text made it the best resource for this work.

The big failure points are the condenser/capacitor, the insulation on the coils and the loss of magnetism.  This post recommends a good quality capacitor with the right features to replace the originals.
http://www.britbike.com/forums/ubbthreads.php/topics/508762/re-restoring-a-rotating-armature-magneto#Post508762
These guys recommend a different capacitor and show how to install them:
http://www.themagnetoguys.co.uk/condensers-practical

This blog shows how coils are rewound. Some rewinders apparently skip insulating layers and don't vacuum impregnate the resin which seems to be critical for longevity.
http://www.brufnut.de/WORKSHOP/LUCAS/lucas.htm

These two show how to build and use a magneto remagnetizer:
https://www.gasenginemagazine.com/equipment/make-your-own-magnet-charger
https://www.gasenginemagazine.com/gas-engines/make-your-own-magnet-charger-part-2

This link will get you a PDF of the 1931 Marelli magneto parts catalog:
http://www.rpw.it/Files/Marelli_CPR_1931.pdf

[Murray]

Before you get too carried away it might be worth seeing if there is a bug smasher airport near you with associated support services, magneto's are still in regular use you'll probably find someone that knows their way around them.

[SED]

Time to pull the armature.  The Bakelite slip-ring (on right in 2nd picture) is very brittle and easily broken.  It is recommended to replace the armature with a large piece of iron as a keeper.  (or remagnetize)

   


Once out I tested the coils using the method described here (3rd picture):
http://superalcerestoration-j2maria.blogspot.com/2010/12/
And it generated a spark though it seemed to jump from the armature to the lead so maybe I had connections reversed.

Using a test light I was able to confirm a few connections.  The wire to the left in the first picture connects the points to the coil ("interrupter" in the old books) while the right wire connects the other end of the coil to ground through the armature.  The 2nd picture shows a Bakelite boss that is part of the slip-ring.  Pulling the slip-ring (after the bearing) exposes a copper needle that carries current for the spark plug from the coil to slip ring contact.  Everything I read said the needle is very fragile and warned to protect it.  (none of this was Marelli specific info, but the Marelli is similar to the Bosch and Lucas...)

   



The spark test gave me the courage to dig out the condenser.  Scribe marks between the end cap and armature so it can be reassembled it correctly.  (some quality of electrickery is maximized by making these armature cores and generator cases etc. from many thin steel plates insulated from each other so scribing or denting the core compromises this quality a small amount) The condenser lives in the brass end-cap on the drive end so swallow hard and remove 4 screws.  Unsolder the 4 wires and find the insulated and grounded sides of the condenser and mark the end cap, armature and coil.

           


Red arrow points to insulation on the + of condenser and black arrow to -:




This is the "safety gap" where a spark will go to ground if the magneto fires if not connected to a grounded spark plug. If it were not there the coil would spark internally destroying the windings.  More here:  https://occhiolungo.wordpress.com/2012/02/02/how-to-more-magneto-rebuild-tips/





Pull condenser and check out where the new capacitors must fit.





Oops - wrong capacitor!    These are the recommended capacitors:
https://www.digikey.com/product-detail/en/vishay-bc-components/BFC238320823/BC1883-ND/502663

   

[SED]

Before you get too carried away it might be worth seeing if there is a bug smasher airport near you with associated support services, magneto's are still in regular use you'll probably find someone that knows their way around them.

Thanks!  Will ask around for remagnetization.  Not sure about coil winding.  Most of the aircraft mag sites I've found are showing rotating magnet magnetos so not sure if they can do armature coil winding - need to ask!  Already in too far - but this is mostly a learning exercise getting ready for the main event on the GTV.

[SED]

Two capacitors must be connected in parallel and then connected to the 4 wires from the coils.  The safety gap must be maintained and the replacement should be removable rather than glued in with epoxy.

Decided to put the capacitors in a phenolic cradle and use the original brass connectors to solder to the coil wires.  Got some machinable phenolic sheet from McMaster-Carr and cut to size.  I don't have a mill so tried many different designs using the tools I have.

      



22 awg wire is about the same size as the terminals on the new capacitors.  Wrapping it around a very slender nail gives a couple of loops that can be slipped over the terminal and crimped to hold it for soldering.  These wires were soldered to a new + connector and the original - connector.  Used pop-rivets for the + connector so that they didn't ground against the end-cap.

     



This may be where the original condenser failed.





The - connector ground strap was angled up slightly where the center screw goes through.  After it was epoxied together there wasn't much I could do without tearing it all apart and starting over so I removed the thick copper washer and it fits fine.   The witness mark on the coil shows that clearance was tight with the original condenser also.

   



 

[dsrdave]

I had my super alce mag rebuilt by a shop in Florida.  Great job and a fair price. 

[dsrdave]

 Just remembered the name.  Ed strain in Pinellas park.

[SED]

While cogitating about how to install the capacitors I repaired the spark plug leadout using the technique shown here.
http://www.themagnetoguys.co.uk/araldite-repairs





The carbon brush had a weak spring and was a little floppy in the holder.  The wire is soldered direct to the brush holder.  Not sure how it was done originally, but I was able to feed the spark plug wire down through the leadout and pull the wire through far enough to solder the wire to the holder.  Like this only through the leadout.

   


Bought a 1/4" carbon brush to replace the 6mm grounding brush.  It fit with the help of fine sandpaper. 

I reassembled the mag, connected a spark plug and the result was shocking!    Spinning it with my fingers gave me a jolt!  Spinning with fingers fires a plug with an 0.018" gap.  Spinning it on the lathe at less than 200 rpm gives consistent sparks.

   


This is just a learning activity and it hasn't been properly rebuilt or tested. It needs 2x E15 magneto bearings (15 x 35 x 8) and a shim under the endcap.  It should be remagnetized!  But now I have some confidence that I can replace the condenser in the GTV mag if I need to. 
 

[SED]

Just remembered the name.  Ed strain in Pinellas park.

Thanks for that Dave.  Good to know some resources.

I'm still hoping to find a high quality coil winding service and some more info on these Marelli mags.  I even wrote the factory, but no joy.   
https://www.magnetimarelli.com/company/history

[Chuck in Indiana]

Excellent work, Sean..
You'll probably be out of luck at the airport. Yes, they rebuild magnetos, but it is a take out old part, install new part kind of thing.
Mouser
2018-05-29_09-38-45 by Charles Stottlemyer, on Flickr
Gave me fits for a while with hard starting. Needed to be very rich to start. Changed carbs. I even put in a primer system. Finally  the light came on and I sent out the magneto to the Kid's shop for a total rebuild. Started first blade.
Most carburation problems are electrical..

[acogoff]

     The guys who monkey around with 2 cylinder John Deere tractors and stationary engines would be the ones to ask about remagnetizing. They will know who can do it.

[SED]

Chuck that is a great looking Mouser - thanks for the story and encouragement.

Acogoff - those are exactly the guys that publish how to build magneto rechargers - Gas Engine Magazine (lots of good info there):
https://www.gasenginemagazine.com/gas-engines/make-your-own-magnet-charger-part-2

There is an old tractor club that meets in the north county so should check them out.

Hope you've found it interesting. 

And if anyone knows where to get a Marelli manual for an MLA mag I'm still looking.

[Chuck in Indiana]

 

Chuck that is a great looking Mouser

Thanks. Mouser is the second machine that taught me that weak ignition needs a rich mixture to fire. The first was my Ducati Darmah, which I owned when I was too stupid to figure that out. So was the well known dealer, though..

[SED]

Chuck, your experience with weak ignition needing rich mixture is getting me enthused to try this with the mag in the bike. 

Being a weekend I was able to make a 0.010" shim so the bearings aren't tight.





And a 3 point spark tester.  One side is grounded to the mag, the other side is the sparkplug lead and the 3rd point is a nail that is not connected to the mag.  The nail is 0.40-0.50mm from the spark plug point.  Apparently it has something to do with ionization at the spark plug point probably to wick away ionization so each spark is just as difficult for the mag to fire. 





Reportedly a spark should jump a 5.5mm gap (in air) at kickstart speed.  A test at 5.5mm didn't spark so I double checked the pictures of the safety gap and filed it wider - to about 2mm.  (I don't have a spec for this). Now it produces consistent sparks across a 5.5mm gap at less than 180 rpm (360 crankshaft rpm).

   


Upping the rpm (to 2500???) produces a brighter spark.





Then there is the high speed test where the gap can be opened up to 8mm.  8mm gap = 15,000 - 24,000 volts (depending on your internet source...). 





Amazing considering the mag still needs to be remagnetized.  Need to find those tractor folk...

[Chuck in Indiana]

  Pretty fancy test fixture..

[SED]

Pretty fancy test fixture..

 

New magneto bearings arrived today.

[SED]

Chipping out old copper lined leather seal so the puller will fit under the race.  Removing bearing race.  Making a new leather seal. 

      


Install new bearings and check armature alignment - found it's better to spin armature on bearing races than detents at the ends of the shaft.

     


Steampunk sparker...

   

[Chuck in Indiana]

Fun stuff..

[SED]

A retired guy a few miles out of town does a small business rebuilding mags - mostly for tractors, old stationary engines and logging equipment.  He has a remagnetizer like this one:




A few seconds of magnetizing and the mag sparks more easily and consistently.  8mm at 875 rpm, 5mm at 167 rpm.  It will even throw a 5mm spark twisting the shaft with your fingers!


 

These things aint cheap! 

[Chuck in Indiana]

Looks like you're good to go.  Attaboy..

[SED]

Success with the first mag has lead me to tear into another one.  The giant learning experience continues...

The reason I pulled the mag apart was to learn how the Marelli rotating coil mags work so I could pull apart the one on the bacon slicer.  The reason I posted the info on WildGuzzi is that I was hoping to scare up some more info by posting info that is not available anywhere else. And there are lot of knowledgeable people on here that might be able to help!


      



The MLA is on the left the LAN is on the right. 
Differences:
MLA has single outlet for high tension lead- cap covers the nearest outlet on the LAN.  Cable stop support extended on LAN.
MLA has a narrower base than LAN  65mm vs 75mm
MLA uses a bolt rather than nut to hold on drive gear
Points are slightly different (MLA points on right)
MLA (on left) has shorter slot for timing advance.


            



A quick spin on the lathe with the 3-point tester showed that the MLA missed about 1/3 the time when spinning at 167rpm.  Pulling the cover and armature showed that it had water sitting in it for quite a while!!!  Amazing that it still worked.
 
rusty magnets and armature, corroded contacts, carbon brushes stuck due to corrosion, rusty springs and gunk.

         






The brass contact (top center) was missing in the LAN leadout which required that the spark plug wire be soldered to the brush holder.  Good little lathe project!




Pulled armature bearings and end cap to reveal an aftermarket condenser.  There appears to be no provision for a spark safety-gap which means a fouled plug or loose plug wire will cause internal coil arcing and quickly short the windings.  0.125uF is low end of acceptable range when new (Lucas published 0.15-0.18uF).  Experts say most ignition capacitors have wax paper dielectric that degrades in 20 years or so  - even just sitting - so this one is on it's way out.

              

[SED]

Got a box to sort out magneto parts.




Cleaned out the inside of rust and grease.




Most of the energy went into fitting the new capacitors.  A friend gave me a meter that measures capacitance.  Two 0.082uF capacitors in parallel measure 0.163uF - perfect!  Fit into the end cap and soldered into place.




Next- bearings and final assembly.

[Chuck in Indiana]

Really good stuff here..

[SED]

Well...   

I was unable to get a good conductivity measure on the secondary coil of the armature (5MOhms) and too inexperienced to know that I'd done it right. But it sparked on the bike and when 12v was applied to the primary so I reassembled it and tested it.  It will throw a spark 5mm, at 167rpm, but not consistently.  Speeding up to 350rpm and it still misses a little.  This mag is not nearly as good as the LAN.

I think the problem is described by Bob Kizer on the PODtronics (Prince Of Darkness) website:
     " I've seen mags taken off a running motorcycle that tested "open". (Starts okay when cold, but not when hot.) So why did the bike run?? The secondary is wound with a very fine wire about the thickness of a human hair. As a matter of fact, about one mile of it. If an open circuit develops in the secondary the high voltage can and will jump the gap, and the bike continues to run. Slowly but surely the ends of the fine wire melt and the gap gets bigger and bigger. This arcing inside the secondary will also leave a carbon track to assist the jumping of the spark. While the carbon track is not a perfect conductor, it's good enough to conduct the high voltage spark when the mag is cold. After it heats up, the bike misfires.  Let it cool down and it will start okay.  Then the cycle repeats itself. "

The cause is likely that no safety gap was provided when the aftermarket capacitor was installed.  Poor workmanship...
Anyone know who rewinds magneto armatures?  This forum is such a wealth of experience and info someone likely has a connection.

Thanks!
Shawn

[Chuck in Indiana]

I had that problem with an antique airplane one time. It had 2 mags (of course) and would start and run fine. Still ran fine until it was shut down. Then.. it wouldn't restart. It only had an impulse coupler on one mag, and started on it only. Naturally, that mag was the issue, and it sounds like that exact scenario. Thanks for the explanation.

[SED]

Mark's Magneto Service 860-537-0376 in Colchester, CT was suggested or recommended by several sources. (they don't have a website)  I  called them and they had a rewound coil for a Marelli MP4 sitting on their shelf.  Ebay listings show that the MP4 fits old Fiat and Lancia 4 cylinder cars - it has a distributor.  I sent the armature for my MLA and Mark called and said that the coils are the same size, but that the lead wires are longer on the MLA - the LAN also has the long leads. 

Mark said the primary has 600 turns, while the secondary has 36-40 layers (about a mile) of very fine (41 gauge) wire.  He explained that the 2 positive leads to the capacitor are the end of the primary coil and the start of the secondary.  If I understood correctly the circuit should look like the one below.  This matches my recollection of the continuity tests I made, but can't be certain as my notes are not detailed enough!




There was no circuit diagram available so I struggled to test the windings.  Maybe this will save someone some time:

Secondary test: connect one lead of an Ohm meter to the sparkplug wire (or directly to the slip-ring if you can reach it) and the other to the magneto body.  A reading of about 5 KOhms is OK (LAN=4.75 KOhms).  This test requires continuity through carbon brushes at the slip-ring and at the armature ground brush so a higher reading may be dirty or stuck brushes.

Primary test: connect one lead of an ohm meter to the insulated point (points open) and the other to the magneto body.  A reading of about 0.7 Ohms is OK (LAN = 0.7 Ohms).  Continuity is made though the armature grounding brush so a high reading may be a dirty or stuck brush.

Mark said the original Marelli winding would have been cloth covered and thought this one must have been rewound in the '50s. He said that their rewound coils are impregnated with epoxy under pressure to drive the epoxy to the core and then a vacuum pulled to draw out any air bubbles.  Mark's Magneto Service rewinds coils for other magneto rebuilders and apparently rewind about 5,000 coils a year!  A busy place with quite a backlog. 

[SED]

Chuck, I just noticed something clever about your Mouser -   
are those incisors or canines?! 

[Chuck in Indiana]

Actually, it's the logo of the aircraft designer back in the day. It's a stylized M for Monett. This picture is of his original Sonerai II, a stretched race plane.



The "real" Mouser is on the other side.  I don't think I have a picture, though. I'll take one when I'm out working on the loop (Guzzi content) today.

[SED]

Got the armature back yesterday and set it up on the lathe.  Consistent bright sparks!  now just have to get engine back together...