Magneto Magnet strength ?

[cmfalco]

Pole pieces shaped to match the magneto being magnetized are needed to minimize the amount of magnet flux that escapes without passing through the magneto.

Attached to my electromagnet when I photographed it are pole pieces for a Lucas KVF. Next to it are the ones for a Magdyno, with the small piece at the right to be mated to a flat pole piece. Although not required, this brings the field closer to the base of the magneto, so why not have it? SEDoan's post shows he has made the same extra piece but, unlike his, mine has two holes for the pegs that are on Gold Star Magdynos rather than bolt holes. The final pair are two flat pole pieces for magnetos with external horseshoe magnets, like a Bosch ZEV.

The next photograph shows my "power supply."

The voltage from the wall varies somewhat with the season and time of day, with it lowest on afternoons in the summer when all the A/C units in the neighborhood are operating. But, it's nominally 240 V and single phase. That voltage is supplied via a 50 A circuit breaker to the "disconnect box" at the top of the photograph, containing a large switch and 30 A fuses for each leg. From there it goes to a 20 A Variac, the side of which is on the right of the photograph. From the Variac it goes to a 90 A full-wave rectifier in the box at the bottom. It's not that a 90 A rectifier is needed, but it was only a few pennies more than one of lower capacity. The output of the rectifier goes to a 20 A DC ammeter, and from there to the electromagnet. Although the unloaded output from a full-wave rectifier isn't quite DC, the inductance of the electromagnet eliminates the oscillations.

On days when the full 240 V is available, the maximum current is 240V/13.04Ω = 18.4 A, which is within the range of all the components, and the resulting field strength is 84,074 A-turns. In operation I take roughly 5 sec. to ramp the current up to its full value using the Variac, and a similar time to ramp it back down to zero. Although that is sufficient to fully magnetize a magneto, I take the opportunity to repeat the process for good luck.

If anyone is interested in cost, I just checked, and suitable Variacs are available on eBay for approximately $60 delivered. A disconnect box is around $20, and 50 A rectifiers are only a few dollars, so my "power supply" could be assembled today for around $100. Someone would have to investigate the current cost of the Armco iron and copper wire needed for the electromagnet itself if they wanted to reproduce it. But, if I take a wild guess based on the price of copper and iron that the materials in mine would cost as much as $500 delivered, that would be ~$600 total -- plus a lot of time machining -- to reproduce my entire "system."

I've only remagnetized maybe a dozen magnetos since making my magnetizer over a decade ago. That works out to $60 each if I value the time I spend machining at zero, which is reasonable since I value all my time spent working on motorcycles at zero. Actually, I enjoy working on motorcycles, so should count that time as money earned. But, I digress. While $60 might sound like a lot, as far as I can determine, Joe Hunt Magnetos recharges magnetos for $25 plus shipping. Since the cost of shipping is quite high, that's probably around $75 total.

It's interesting that my magnetizer has paid for itself, which wasn't a consideration when I built it. And that's without taking into account the convenience of having a fully-magnetized magneto within a few minutes after I finish rebuilding it, or the potential loss or damage in shipping of a hard-to-replace one like the KSA1 on my 1928 Ariel.

[SEDoan]

Great details on the magnetizer power - thank you! BTW, when I priced soft iron from Ed Fagan metals soft iron was about $10/pound and magnet wire was over $10/ pound. It's the primary reason I tried the capacitor type. If you know cheaper sources please share them.

[SEDoan]

I will show how the magneto charger was built but it will take a few posts, partly because I need to sit down and draw out a wiring diagram for the final version built.

Paul Greet, an electrical engineer from New Zealand designed this charger and was kind enough to share his design and advice with me and a few local friends. I apologize if there are any mistakes in my version, they are not Paul's errors.

Warning - the large capacitors store enough charge to kill a person. Apparently they can also explode so they are built into a wooden box to keep me from touching them and to catch any shrapnel. The capacitors make this charger more dangerous than the Lucas type.

The electromagnet is made from an old transformer that weighs 43lbs. Transformers are built around a laminated soft iron core. The windings should be good so they need to be checked for continuity and grounds. The outside of the frame also needs to be large enough to fit a magneto between them. You also want many windings - transformers for very high currents seem to have fewer windings of heavier gauge wire (like the large one in the picture below). Picture below shows 200lbs and 43lbs transformers.

The specs for the smaller transformer used for the magnet.

The same guy that had the transformer also had some large capacitors that were past their shelf life. The charge is stored by two of these capacitors in series.

Cutting one side out of the frame of the transformer leaves you with copper coils around a U-shaped core. This picture is from a video that shows the resulting magnet pulling the steel rebar when connected to a 12 volt automotive battery charger. I used this setup to determine the polarity of the coils and experiment with connections between the coils to see which connections would produce the strongest magnet.

[SEDoan]

BTW - I got the old capacitors and transformer for $60.

Wiring two of the large capacitors in series requires 800 volts to bring them to full charge. This is accomplished by a voltage multiplier built from diodes and small capacitors. It uses a capacitor paired with a diode for every voltage increase. Because we have 120 volt AC house current in North America I built a 5X multiplier. It is capable of getting 840 volts DC from 120 volt AC house current and takes a little over 60 seconds to charge the capacitors to 750+ volts. Paul's original design used a voltage tripler for New Zealand's 240 volt AC.
Picture below shows 120 volt AC input black (hot) and white (neutral) wires are coming in from the right side. Red and black wires at the bottom are DC output to capacitors. The black piece between the black and red wires is an inrush current limiter - it will pop the circuit breaker without it. Second picture shows the DC output voltage on the meter.

Here's a wiring diagram - please double check that I've drawn the diodes pointed correctly!

Underside of the board - there must be a better way to do this but a friend who prototypes electrical components says this is OK.

Note: some of the small capacitors on the voltage multiplier board don't seem to discharge on their own, even after using the safety discharge for the main caps. I've started discharging them with a short piece of wire every time I pull the lid off. Or check with a volt meter.

Part numbers from Digikey:
small capacitors - order extra I've blown 3 or 4. And they are small, but still give a hell of a jolt when you touch them!
500HXW22MEFR12.5X20
1189-500HXW22MEFR12.5X20-ND
CAP ALUM 22UF 20% 500V RADIAL

small diodes for voltage multiplier. Order extra - I burned out 4 when accidently charging and discharging at the same time!
The diodes in the picture have heat sinks on them. I thought they would clip onto the diodes but they are too large. If you order the correct size they should clip on without the screw and nut shown in the picture.
VS-E5TH3006-M3
112-VS-E5TH3006-M3-ND
DIODE GEN PURP 600V 30A TO220AC

In-rush current limiter for 120 volts. You may need different specs if you are using a 240 volt power source.
B57127P100M301
495-7209-ND
ICL 10 OHM 20% 15A 30.5MM

blank circuit board
EXN-23406-PCB
377-2629-ND
BREADBOARD GENERAL PURPOSE NPTH

Silicon Controlled Rectifier - this is the diode that acts as a switch to discharge the caps through the electromagnet
VS-81RIA120PBF
VS-81RIA120PBF-ND
SCR 1.2KV 125A TO209AC

Safety discharge resistor
HS50 100R J
HS50100RJ-ND
RES CHAS MNT 100 OHM 5% 50W

Also you will want switches. I am using a momentary push button and two DPDT switches all rated to 600 volts.

And you will want a volt meter. A friend had a 250 volt AC meter that he modified with a small resister that I calibrated with the multimeter.

Wire - I am using 14 gauge wire between the voltage multiplier and capacitors and 12 gauge between the capacitors and the transformer coils. The wire between the capacitors and diodes on the voltage multiplier is finer.

[fpassmore]

Absolutely, Nev is right!!! I used to give courses in workshop safety repairing and assembling high-efficiency permanent magnet motors and they were rated as highly hazardous
Frank

[Roger Gwynn]

Thank both Shawn and CM, I now have another project on the list.

[SEDoan]

Here is the wiring diagram for everything except the voltage multiplier. Apologies for the mix of conventional symbols and mine. Hopefully it makes sense.

If you want to geek out on the circuit you can use a tool that an electrical engineering friend shoed me - LT Spice.
https://www.analog.com/en/resources/des ... lator.html

Here is a charge simulation showing that the voltage multiplier would reach almost 800v in 60 seconds. In reality it is a little slower than this and the capacitors leak some so difficult to build up more than 770 volts no matter how long you charge them.

And here is one of our early iterations of a discharge simulation. R1 is the electromagnet, S1 is the SCR. The discharge duration is a little more than 1/100th of a second (voltage drops from 900v to 0v) but the amperage peaks at about 500 amps but only for about 1 millisecond.

Here is a MO1 shell that I reversed the magnetism on at 400 volts. Notice the compass needle. BTW, it's easy to reverse your compass polarity so go outside and check that it still points north!

And here is Paul's original circuit diagram for 230v AC. He warned us that it was simplified. Note that it also includes a power supply to trigger the SCR. I was given some suspect advice (not from Paul) and accidently burned up my trigger power supply (an old computer charger) and the SCR at the same time so now I just trigger the SCR with a 9v battery.

[SEDoan]

First I built a box for the electromagnet with holes for the poles. This also shows an experimental setup with 4 capacitors in series. At this time I only had a 12v battery charger to charge them so it was still pretty safe.

This shows the capacitor box. The cloth covered wire are the transformer taps, the bright green wire connects the safety discharge resistor, and the finer dark green wire goes to the volt meter. The larger red and black wires come from the DPDT switch to charge the caps. Note how the positive side of the SCR is bolted to an aluminum angle bracket and the negative cable of the SCR is bolted to the negative post on the second capacitor.

Note: I've been told that capacitors don't do well when charged in series because one cap usually fills up faster than the other causing the full cap to overcharge and begin to leak. A friend and I tried a simple circuit to charge in parallel, then switch to series for discharge but it didn't work - though I did hear a nice pop in the switch. For whatever reason, perhaps the 50v surge tolerance, these two capacitors have been working well together. The other two capacitors failed, perhaps because they were not as well matched, and now leak so fast they can't be charged more than about 150 volts.

Caution: These capacitors store a LARGE charge. Before I had the safety discharge connected I discharged some remaining charge with a wire (no resistor) and it was like a gunshot!

Here is a picture showing the charge voltage of the capacitors 762v. Voltages much higher than this are not possible because the capacitors lose the greater charge too quickly. Also note the SCR trigger wires next to the 9 volt battery. The SCR only requires a few volts to discharge so I just touch the 9 volt battery and the caps discharge through the magnet coil.

This is the current configuration of the switch box which also contains the voltage multiplier. Turn the power on (right side switch), turn the charge switch on (left side DPDT switch) and press and hold the push button for about 1 minute. Voltage rises quickly to about 400 volts, but then slows as the capacitors reach capacity. The switches on the bottom row are not used right now - they were for failed experiments with power supplies and parallel cap charging. To trigger the caps and charge a magnet the left side DPDT switch can be put in the center so the voltage multiplier is completely disconnected from the circuit before discharging the caps. In practice it seems to be enough to be disconnected by the momentary pushbutton. You will blow all the diodes if the caps are triggered with the multiplier connected. Also I had to carve out a well for the mag to lay in so that it would sit between the poles of the magnet.

When charging the magnet the magneto should be completely assembled and the polarity of the magnet determined. The MO1 magnet is cast into the thicker side of the mag body so one pole is upper side and other is on the same side below it. Opposites attract so it should be placed with the north pole of the magneto in contact with the south pole of the electromagnet and vice versa. At full advance the points are supposed to open where the greatest magnetic flux occurs. I would think this position (points opening) would be the ideal position but I have read that the armature should be rotated 90 degrees backward from this - though I don't see it mentioned in the Lucas instructions for the MO1. Any air gap weakens the magnetic field dramatically so it is important to have shaped blocks. Lucas instructions suggest they can be made of mild steel.

[nevhunter]

When you are successful you'll feel the magnetic force has become stronger and the points opening position must relate to where the magnetism reverses. It's possible with mixed parts to get it wrong. On a Vee twin there's also a correct polarity to have the sparks as even as possible in intensity. Nev

[cmfalco]

BTW, it's easy to reverse your compass polarity so go outside and check that it still points north!

I hope it doesn't confuse the situation, but the pole of the earth that is near Finland is actually magnetically south according to our electrical convention. This results from the decision that current flows from a positive terminal to a negative before it was discovered that current is carried by electrons that flow from the negative terminal to the positive. What this means is the "north" end of a compass needle actually is a little N magnet, since it points toward the pole above Finland that is magnetically south. Where a Lucas manual shows a "N" and "S" pole, those are to be taken as the poles that will attract the ends of a compass needle that would point toward Finland and Antarctica. Are you confused yet?...