Rare earth update

[browning]

I have been playing with flywheels and magnets the last few days. The previous thread was getting lengthy so I decided to post some results separately. I had two flywheels with freshly recharged original magnets and another on which I installed both neodynium and samarium cobalt rare earth magnets. I installed all of the flywheels on my hcct, operating at approximately 190 rpm, and adjusted the air gap as close to .035” as I could get. I installed a carefully adjusted coil in the hcct which tested “excellent” on my ecct for a load. My AC voltmeter has a 75 volt scale which isn’t perfect for the accuracy of the test but OK for comparison I think. Results as follows:

Original flywheel #1. 2 volts
Original flywheel #2. 2.5 volts
3/8” neodynium magnets. 5.5 volts
1/2” samarium cobalt magnets 11 volts

The last test exhibited considerable multi-sparking even though the coil previously tested good/excellent on the ecct. The rare earth magnets were installed on the flywheel within carriers made of medium density fiberboard which is all right for tinkering but the next iteration will be made from 6061 aluminum and will be installed in my next engine. I will post more results as I find them.

[speedytinc]

Excellent info.
My question is what is the state of magnetism of the 2 original flywheels. For reference. What weight will they hold. The customary 2# iron piston, a 4# sledge hammer or something in between? I look forward to your in car results. Thankyou.

[browning]

I recharged both of the original flywheel magnets a while ago so I can’t verify the weight they would hold but I have a very heavy industrial magnetizer and always give the magnets all they will take. Neither flywheel has been installed in an engine so I am confident that they represent any fresh original recharged set. I can’t explain the apparent 1/2 volt difference between them but their similar output seems to indicate that their condition is as I have represented them.

[MKossor]

Not exactly clear on the on the objective in supping up the Magneto output, perhaps just having fun.

A couple thoughts for your consideration. I would fully expect a coil driven with such high output and abnormally slow rise time on 190 RPM (that is a rise time of nearly 0.010 seconds) to produce multiple sparks. That's because the nominal time to fire spark is 0.002 seconds and recovery time os 0.0015 seconds means there is plenty of magneto voltage still present to initiate and fire a 2nd (and maybe 3rd) spark.

Another thing to consider, if the first spark fires at the proper time of the proper energy, full and complete combustion occurs. A second spark firing really is of no detriment or benefit.

Lastly, generating higher magneto output than necessary to fully charge and fire a model T coil with more than 0.050J of spark energy typically produced is just putting additional load on the engine! The moving high power magnetic field of the magnets induces eddy currents in the Iron engine block (energy dissipates as heat) as well as higher than necessary coil current sapping horse power from the engine without clear benefit as far as performance is concerned in my view.

Probably makes some really neat sparks with really loud snaps though if just having fun

[speedytinc]

My thinking is keeping the magneto power source, but having a lighter flywheel & one that wont/cant fling magnets could be advantageous.
I understand too much magnetism makes for power loss. So why not use smaller magnets to not over produce excess energy?
Shoot for a strong 35 volts.

[TXGOAT2]

My interest in upping magneto output is centered on producing more magneto output at hand cranking speeds. The benefits of accomplishing that are obvious. I believe it would also be of benefit in cars that use magneto output for lighting. A magneto with more durable magnets and power to spare at low crankshaft speeds could be expected to provide trouble-free service for a very long time. Regulation would likely be needed. Regulating current output would not eliminate magnetic losses due to eddy currents, but it would probably offset them to some degree, and reduction of losses due to the pumping action of the stock assembly rotating in oil would offer further offsets of parasitic losses.

[browning]

My fascination with the magneto is based primarily on the "why not" factor. I find some sense of joy from doing things that are outside the box. The benefit of the rare earth project would be, in my mind, that the samarium magnets are almost bulletproof with respect to heat and the air gap in the magneto is much less critical at assembly or in the case of crankshaft end play. As far as parasitic losses it should be fair to assume that the rotating mass of the flywheel with steel magnets would be much higher than with aluminum carriers and the balance would be much easier to achieve with machined lightweight parts versus sheared steel magnets. The output of the rare earth approach can easily be modified to suit the application by using smaller magnets or larger air gaps as needed. The side benefit of samarium cobalt magnets is that they are much less affected by interfering magnetic fields i.e. accidental voltage applied to the mag coil. What I am seeing is an increase of four or five fold over stock magnets which is probably overkill but easily modified. These magnets look to be affordable and install in the same way as the originals, in fact they could be black anodized and hardly distinguishable from the original steel magnets. With the last test I was surprised that the hcct was still sparking down to the last few revolutions when the power was disengaged. Henry might not have bought it but I'll bet he would of at least smiled.

[TXGOAT2]

Henry might very well have bought it, had any such thing been available. He spent money when he had reason to believe it would provide practical benefit. He was an early adopter of vanadium alloy steel, among other innovations.

[browning]

So here is plan B for the improved mag project. If it won't work you might tell me so before I spend any more time or money on this cranial flatulence. We recently had to replace the magneto coil and voltage regulator on one of our commercial front cut mowers with a 27 hp Kohler engine and when we tore it apart I was surprised by the similarity to the Model T system. The Kohler was much smaller and had 18 poles vs. the Ford 16 but the windings were much lighter gauge copper wire and the rated output was approx. 15-20 amps dc at 14 volts or so. The solid state regulator was about half the size of a pack of cigarettes and cost less than fifteen dollars which got me thinking. Apparently output frequency isn't critical since the Kohler would range from 3600 to about 33000 cps and from what I could tell the mag output voltage would run up to 40 vac or so. The obvious question was whether a Model T mag could supply the current necessary to charge a 12 volt battery through the cheap regulator. I can't see any weak points in the system with the possible exceptions of the mag post and the logic of the ac mag output and the dc regulator sharing a common ground. Seems worth a try to me. The only telltale would be the dc output wire from the regulator (well hidden) to the terminal block. I envision the generator, disabled cutout, and wire staying in place but removing the drive gear to quiet the thing down and retrieve the lost energy. Rewiring the starter, changing the battery and light bulbs would complete the modification. The plan now is to mount a modified flywheel/magneto into an engine lathe so the spindle speed can be varied and to carefully mount the mag coil onto the lathe saddle so the air gap can be infinitely changed to calculate the most appropriate installed gap. The regulator, ammeter, and battery could be installed on the lathe setup as well. The choke coil you guys mentioned might me necessary to prevent blowing out the regulator at higher engine speeds but I still don't understand that element. Am I as crazy as my wife thinks I am?

[Scott_Conger]

You obviously have the interest, intellect, curiosity and apparently the equipment to do this, so why not? There is no telling what you'll learn with experimentation that will benefit you or others in the future. No one ever achieved greatness from a sofa, other than perhaps Mae West.

On the other hand, if you are trying to improve the Model T's performance, I'd say that 15 million of them could be started with the crank and driven to their owner's satisfaction during the course of its lifetime, so then what's the point?

Whether or not you continue is entirely up to you, but know that Model T flywheels have been modified very successfully in the past, using rare earth magnets, so decide for yourself why you're doing this and proceed appropriately.

Above all, have fun and learn stuff.

[browning]

Scott, there has been anxiety recently about the loss of Fun Projects voltage regulators and just a few posts ago information on how to rewire a starter for 12 volts. Apparently there is some interest in this type of thing. It may well be that a good stock mag might work with the solid state regulator to provide 12 volts. If nobody tries nobody will know.

[A Whiteman]

Am I as crazy as my wife thinks I am?

Do you mean that there actually are wives out there who DON'T think their husbands are (at least a bit) crazy

[browning]

One of the guys who works for me has one - he won’t sell her though!

[A Whiteman]

so then what's the point?

Well, there are not 15 million of them on the road any more and no original dealers stores that I have found at least, so it is not unwise to safeguard the future of the hobby by exploring alternatives when 'original' bits get a little hard to find....

[browning]

330 million people in the country and most don’t need a stent. The ones who do are glad that someone looked into them.

[Scott_Conger]

Browning: you asked

Am I as crazy as my wife thinks I am?

So, Browning and Whiteman...did you not read my first sentence responding to your question?

You obviously have the interest, intellect, curiosity and apparently the equipment to do this, so why not?

How on earth can anyone read this and think I am dissuading anyone from experimenting?

I often suspect that people simply scan responses looking for either an argument or to support their position against all odds...you have no quibble from me, I support your experimenting if it makes you happy, and I said so... GOOD GRIEF!

[MKossor]

The obvious question was whether a Model T mag could supply the current necessary to charge a 12 volt battery through the cheap regulator.

The answer is Yes and No. Yes the Model T magneto can supply the current necessary to charge a 12 volt battery (modestly at ~ 1A steady DC) but the Model T magneto output is a rather hostile power source with severe EMI from the firing coils with voltage spikes to around 300V, widely varying AC output of around 60-100V peak with widely varying frequency so No, that voltage cannot simply be rectified (half wave only being referenced to engine ground) and used to drive a cheap regulator to charge the battery.

I did design a high efficiency isolated switching regulator with sophisticated transient voltage and EMI protection that could withstand and full wave rectify the Model T magneto output then regulate to produce a clean 13.8VDC output that could supply a steady 1A battery charge over the full range of normal engine operation (600-2000 RPM). Given a projected retail price of about $150, I decided to shelf that project rather than try producing them to compete with the ultra simple 5A diode and 1156 light bulb $5 pulse charger solution that's been used successfully for over a decade. Bottom line; I had fun and learned stuff. By all means, have at it and experiment.

[Gen3AntiqueAuto]

Spin that at 2000 rpm and post the results.

[browning]

The flywheel is mounted in the lathe and the mag mounting is being discussed. Maximum speed of the lathe I have is 1550 rpm but that should make an interesting test. I will be able to change speeds easily and am anxious to see if the voltage increases in a linear fashion. Same with varying air gaps. I suspect I’ll have to use smaller magnets to find the sweet spot. I’ll keep you posted - pics when appropriate.

[browning]

Some interesting and surprising data. I completed the lathe setup with the flywheel and mag and was able to make some voltage tests this afternoon at various speeds. The voltages with the stock magnets, recently recharged, yielded the following at .020 air gap:

364 rpm 1.0 volts
492 rpm 3.0 volts
648 rpm 7.0 volts
876 rpm 12.5 volts
1150 rpm 13.5 volts
1550 rpm 17.5 volts

With the new samarium cobalt magnets installed:

364 rpm 5.0 volts
492 rpm 10.0 volts
648 rpm 13.0 volts
876 rpm 16.5 volts
1150 rpm 21.5 volts
1550 rpm 27.5 volts

Ironically the voltage increase at cranking speed was almost exactly 500 percent while at running speed it was only increased by 57 percent. That is as fast as my lathe will run so from 1550 on up is anybody's guess. My assumption is that the fields are large enough that the overlap at higher speeds has a tendency to minimize the gain. Seems like a good thing as far as crank starting goes. Just for giggles I played with the air gap and found that if I increased the gap to .165 with the new magnets the readings were the same as with the stock magnets at every speed.

[TXGOAT2]

That sounds great to me. Perhaps there is some self-regulating effect in play, maybe due to different magnetic field patterns. The voltage curve also looks like it would do a better job of powering headlights or other accessories, like a battery charger. I assume current is increased along with voltage. Loading the output and observing voltage drop would probably indicate whether amperage is increased along with voltage at low speed.

[speedytinc]

Some interesting and surprising data. I completed the lathe setup with the flywheel and mag and was able to make some voltage tests this afternoon at various speeds. The voltages with the stock magnets, recently recharged, yielded the following at .020 air gap:

364 rpm 1.0 volts
492 rpm 3.0 volts
648 rpm 7.0 volts
876 rpm 12.5 volts
1150 rpm 13.5 volts
1550 rpm 17.5 volts

With the new samarium cobalt magnets installed:

364 rpm 5.0 volts
492 rpm 10.0 volts
648 rpm 13.0 volts
876 rpm 16.5 volts
1150 rpm 21.5 volts
1550 rpm 27.5 volts

Ironically the voltage increase at cranking speed was almost exactly 500 percent while at running speed it was only increased by 57 percent. That is as fast as my lathe will run so from 1550 on up is anybody's guess. My assumption is that the fields are large enough that the overlap at higher speeds has a tendency to minimize the gain. Seems like a good thing as far as crank starting goes. Just for giggles I played with the air gap and found that if I increased the gap to .165 with the new magnets the readings were the same as with the stock magnets at every speed.

Great work. Testing & producing meaningful #'s make for usable comparisons. I am following your experiment closely, as is another T buddy.

Your samarium voltage #'s have the look of a good hot mag output as is. IMO. No extra air gap would be desired to mute the output.

Your stock magnet voltage outputs are really poor. You could not have reached magnet saturation. If I had these readings from A fresh engine build, I would be real disappointed.
The last 2 I did showed 35V output @ hi speed. I test individual charged magnets to hold a 4# sledge. These motors start easily with a 1/4 pull. The extra? magnetism doesnt seem to detract from their performance. Both T's do 65+.
I am aware there is an optimal output vs drag for such operations as the Montana 500. I dont know what that closely held secret is.
I have also been present @ an in car magnet recharging session with a big dc arc welder. The result with one sparking was 25v output.

My work with magnet charging is fairly new, maybe 8 years only. I suppose I could be over doing things, but, so far I dont know of any ill effects.
Is 17.5v @ 1550 rpms in the normal range for a good mag?
Feed back appreciated.

This is not a criticism. I hope this information helps you with this project.

[Original Smith]

I had the magnets charged in 1962 on my 1913. They are still going strong today with over 55,000 miles.

[browning]

It looks like these magnets will reduce the weight of the flywheel by about 8 1/2 pounds. Do you guys think that would make a noticeable difference in performance, all else being equal?

[speedytinc]

It looks like these magnets will reduce the weight of the flywheel by about 8 1/2 pounds. Do you guys think that would make a noticeable difference in performance, all else being equal?

There is also the benefit of not worrying about a magnet breaking or getting loose.

[Scott_Conger]

The loss of kinetic energy will make engaging low gear more difficult, and require more RPM and more slipping to get engaged particularly on an incline.

With the original heavy flywheel and a tuned car with spark where it belongs, a T should be able to be put into low gear at an idle, on a paved level road. That would not likely be possible with a flywheel thus lightened.

If the plan is to shave 2 seconds off of the 0-35 MPH time, while risking breaking the low drum from excessive slipping, a lighter flywheel would be ideal. At least it WOULD be peppier! ...but not something I'd be inclined to do...though I'm dead certain that a lot of people would be keen to the idea.

[TXGOAT2]

A lightened flywheel might benefit a T engine tuned to operate at higher RPM in a lightweight car. A lightened flywheel could also offset the added mass of a counterbalanced crankshaft. If a stock weight, or a heavier than stock, flywheel is called for in a particular application, a heavier flywheel could be constructed as easily, if not more easily, than a lightened one.

[TXGOAT2]

Concerning starting on magneto:
See: "Starting Something / Easy Starting Under Adverse Conditions"
Excerpted from "The Ford Owner" January 1920
(Found on Cimmorelli Website, www.cimorelli.com)

[browning]

It seems the more I look the less I understand. At the request of a previous post I disassembled the stock flywheel and recharged the magnets since I too was surprised at the low voltage the first test produced. In addition to that I replaced the magneto coil with one that I picked up from Kevin Prus Friday which was freshly rebuilt and reinsulated. The original test was done with an old somewhat greasy coil that I had removed from a running engine. I surmised that the first coil might have had a few grounded coils early in the series contributing to the low output. When I reassembled the flywheels and coil onto the lathe I took more pain to indicate them this time. They were still not perfect but within the range that they are normally installed on an engine +- .010 here and there. The recharged magnets would all support a weight of just short of 4 pounds and I checked the polarity carefully as I reinstalled them with a compass. I was certain that everything would improve with the second test this afternoon. In actuality the low speed on the lathe provided almost no measurable voltage with the stock flywheel while the highest speed (1550) increased by a volt and a half!!! I'm still scratching my head over that result. While I was at it I checked both flywheels at four air gaps of .020 .030 .040 .050 just because I could.

Stock flywheel: Rare earth magnet flywheel:
364 0.0 0.0 0.0 0.0 364 3.75 2.5 2.5 1.75
492 1.0 0.5 0.0 0.0 492 9.0 8.0 6.75 5.5
648 5.0 4.5 2.5 2.0 648 12.5 12.0 11.5 11.0
876 11.0 10.5 9.0 8.0 876 16.5 15.5 15.0 14.5
1150 15.0 13.5 12.5 11.5 1150 21.7 20.7 19.5 19.0
1550 19.0 17.5 16.5 15.5 1550 27.5 26.5 25.5 24.5

My voltmeter is about as old as I am but I scientifically checked its accuracy against my model train transformer, which has a digital readout, and found it to be pretty close. Also my bifocals don't like the scale on the voltmeter so I could be off a half a volt or so because of that but I think these numbers hold up pretty well even though they don't make complete sense. Kevin also gave me several old dead flywheels so I intend to try a mag coil recharge on one of them and see what that results in.

[browning]

I had these spaced really well on the post but when I submitted them the page realigned and screwed up the works. Sorry about that

[Scott_Conger]

If your magnet poles are +/- .010" relative to each other, that is about as far out as I would build. That said, "0" volts at 364 RPM means something is seriously awry, and the height issue is likely only a (minimal) contributing factor and not the root cause. 190RPM adjusted as you are should give a solid 2+ Volts (which is all you need to light off a coil and start the car)

[TXGOAT2]

From a 1923 Ford publication: (Cimorelli website) "Fordex" (Lots of Ford, Lincoln, and Fordson information)
Magneto output:
300 RPM: 5 volts, 6.1 amps, 26.4 Hz
400 RPM: 9.8 volts, 7.9 amps, 52.8 Hz
600 RPM: 14.4 volts, 8.5 amps, 80 Hz
800 RPM: 18.8 volts, 9.0 amps, 106.4 Hz
1000 RPM: 26.2 volts, 9.0 amps, 160 Hz

[browning]

I think my meter is the problem. I just hooked up a coil to the new flywheel and had consistent spark at 112 rpm and intermittent spark at 84.

I didn't get to try the stock flywheel because I tried the setup at 1550 rpm and had fire flying everywhere for about 15 seconds until everything went dead. I had made about a half inch spark gap and there were about a half a dozen strong sparks together for as long as it lasted. When it quit the primary wouldn't operate so I either shorted the cap or something internal came loose I guess. Try another coil tomorrow.

[Allan]

Others have mentioned driveability issues with a lightened flywheel.The heavy flywheel in a standard model T contributes largely to its ability to cope with the two speed transmission. Making that flywheel lighter will make the engine more responsive, but the drive is compromised. A KC Warford would fix that, by giving the driver a handful of closer gear ratios to play with.

Allan from down under.

[RajoRacer]

Or a BB overhead with a lightened flywheel - if I remember (30 + years ago), my machinist removed around 11 lbs. of my flywheel - the drawing was in a Vintage Ford magazine from the 60's - the fella was a Canadian machinist and drew a dandy drawing ! My apologies for thread drift !

[browning]

I had a chance to play some more today and found some more interesting things. I speculated that the aluminum "magnets" would lighten the flywheel by about 8 1/2 pounds and someone said that might accommodate for the added weight of a Scat crank. I had one in the shop and weighed it and a stock T crank just for the heck of it and found that the Scat weighs 16.8 pounds more than the stock one. An assembly with the modified flywheel and Scat crank would still weigh about 8 pounds more than stock. Another surprise was that I had the guys dig out the oscilloscope and we set it up on the freshly recharged flywheel/mag setup in the lathe to check the output voltage. At 1550 rpm and an air gap of .030 the stock setup produced 16.1 volts consistently and 5.4 volts at 492 rpm. I would take those numbers to the bank even though I have been telling folks for years that the T wold produce over thirty volts at speed. The most interesting thing I found was that the scope showed 28 hz frequency at 492 rpm and 88 hz at 1550 rpm. I'm not the sharpest knife in the drawer but that seems low by a considerable factor. Maybe you guys can explain why the low frequency to me. I was also surprised by the fact that the trace on the scope was consistent in shape and amplitude until a coil was introduced into the circuit at which time the scope went crazy. Finally when I mounted the modified flywheel and increased the lathe speed to where the output voltage exceeded about 19 volts, the ignition coil in the circuit began to throw fire balls from the points and the output provided quite a light show as it attempted to create a "Jacobs ladder" out of my 3/8" spark gap. I feel certain that if I had allowed it to continue it would have destroyed the points or the internal windings.

[Scott_Conger]

Your voltmeter is still undoubtedly lying to you. When you get the right one into your set up, you will see the (nearly) 30V you are expecting.

What type of voltmeter are you using?

Also, the extra weight of the crank is of no use to torque, relative to the torque loss of reducing the flywheel by 8 pounds. It provides benefits, but increasing torque to offset the loss of flywheel inertia, no.

[TXGOAT2]

The added mass of the Scat crank is located closer to the center of rotation than is much of the mass of the stock flywheel/magnets assembly. Mass taken off or added out toward the rim of the flywheel/magnet assembly would have a greater effect than mass added to or removed from the crankshaft, pound for pound. 8 lbs off the outer portions of the flywheelassembly might be nicely offset by adding 16 lbs of mass to the crankshaft, most of which will be in further toward the center of rotation compared with much of what is taken off the flywheel. 60 Hz = 60 cycles per second (for instance) and 60 Hz/cps would contain 120 voltage/current peaks, 60 positive and 60 negative. (If I understand it correctly) 1 Hz = 0v to positive voltage peak, then back to 0v, then 0v to negative voltage peak, and back to 0v. Current and voltage can shift out of phase in inductive circuits, especially at higher frequencies.