Model generator build from scratch

Help Support HMEM:

Steamchick

Well-Known Member
HMEM Supporter
Joined
Apr 2, 2020
Messages
1,013
Reaction score
338
Location
Sunderland , UK
Working in a cable manufacturing company back in the mid-1970s, I wound coils for a magnetic flux switch for a production device. The coils were wound on a simple wood former, with a paper inserted at every 4th or 5th layer, and the whole painted liberally with shellac before the next few layers were wound. This was based on the method used by an "old-timer" who did the similar job of making coils a few decades earlier. He said either paper (we used good newspaper) or cotton or linen (his wives old sheets!) were used as the fibre reinforcement. The finished coil was wrapped with strips of cotton (I used an old pillow-case for the fabric) and finally painted with more shellac. Shellac is basically a solid varnish that is dissolved in Meths to make the lacquer. As it has been used since Victorian times it has a good pedigree, and coils can be unwound if soaked in meths for a few days. I guess modern Epoxies (I guess 2pack?) are stronger, but the coil when mounted on a stator is going nowhere. Shellac seems to be good for 50 years or more - I don't know if epoxy will last as long? However you impregnate the coils I guess it will be a messy job.
I have a Monkey Puzzle tree that "bleeds" large quantities of resin. I know that insects trapped in the resin become "Amber after about 50 million years or so, But I won't be around to see that. Also I know the Japanese used the Pine resin to make their Lacquer-wear - pre-plastic era moulded and decorated boxes, cups, bowls (usually black). When I was there in the late 1980s a guy told me they were replacing the lacquer with modern plastics, but on transformer and motor windings etc. it was still still the preferred option - cheap and proven for decades. Is it a simple case of dissolving the resin in hot water to make the lacquer? - Anyone know?
Enjoy!
K2
 

Steamchick

Well-Known Member
HMEM Supporter
Joined
Apr 2, 2020
Messages
1,013
Reaction score
338
Location
Sunderland , UK
An observation on your test rig. The Orange stator looks like some Nylon or other non-magnetic material. "Of course" (an electrical engineer would say) you need an iron core to get to "practical" magnetic circuits so the magnetic field strength is much higher and the voltage and current will consequently rise.
Maxwell explained it all in the 1870s and wrote the text book still in use today. I'm not an Electrical engineer so won't try and explain. I'd have to read the book - but I suggest you do instead? Or the others on this thread?
Something to do with B=nI, dB/dT proportional to V? etc.
Cheers!
K2
 

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
An observation on your test rig. The Orange stator looks like some Nylon or other non-magnetic material. "Of course" (an electrical engineer would say) you need an iron core to get to "practical" magnetic circuits so the magnetic field strength is much higher and the voltage and current will consequently rise.
Maxwell explained it all in the 1870s and wrote the text book still in use today. I'm not an Electrical engineer so won't try and explain. I'd have to read the book - but I suggest you do instead? Or the others on this thread?
Something to do with B=nI, dB/dT proportional to V? etc.
Cheers!
K2
Il go with the setup i have at the minute, the epoxy works great its just the effort of mixing a new batch for every coil isnt practical. Like you seid they only need to hold there shape when instaled. Im making a set of 12 formers and then it wont matter if it takes a day to dry the lacher/varnish.
But they do need to be very acurate to creat as little gap between the magnets as possible.
The test rig is printed abs plastic.
Ive seid in my first post and many time now im not using iron if at all posible! Please see the link i posted to a previouse build by manfred
 

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
I
Good luck with it, as you say there do not seem to be many sources for kit dynamos/generators.
I am in the UK and came across this little unit on ebay:

I make these lol, ive made 100's just finishing another batch of em. Im looking to make somthing a little more interesting and maybe produce as a kit.
 

Attachments

Last edited:

Steamchick

Well-Known Member
HMEM Supporter
Joined
Apr 2, 2020
Messages
1,013
Reaction score
338
Location
Sunderland , UK
Hi Manfred. Sorry, I had just picked-up later in this thread. If you don't want to do metalwork, is there an 'iron-filled'polymer you can use to print your parts for the magnetic circuit? I'm an Engineer. Please forgive my prejudice, but to make an electromagnetic machine but not using simple materials to hugely improve the machine is like making the proverbial chocolate fire-guard. In my humble opinion. You are doing such a good job, why spoil it by using what is effectively a "magnetic non-conductor" where you need a "magnetic conductor"?
K2
:(
 

Steamchick

Well-Known Member
HMEM Supporter
Joined
Apr 2, 2020
Messages
1,013
Reaction score
338
Location
Sunderland , UK
Sorry Darwen guy, I got my names muxed-ip.
I thought I had opened a reply from someone called Manfred. Sorry to both of you.
K2
 

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
Sorry Darwen guy, I got my names muxed-ip.
I thought I had opened a reply from someone called Manfred. Sorry to both of you.
K2
Please see this thread by 'manfred'. It will explain the sort of thing im aiming for
 

TSutrina

Well-Known Member
Joined
Mar 5, 2017
Messages
96
Reaction score
16
One time years ago when I worked for an aircraft generator manufacture, I am a mechanical engineer. He said the relationship of peak voltage is a cubic relationship to speed. Peak voltage occurs at zero output current. Thus trip the generator breaker the wires see peak voltage. Power relationship is a square of the speed term. Thus double the speed doubles results in four times the power. The magnetic flux for a permanent magnet generator is fixed so I would keep the regulator if you want to do something with the power besides dump it into a heater. I would add resistors in series with the inductor of the schematic above to represent the resistance of the wire coils. Note generators driven by the jet engine directly are called wild frequency. The peak voltage swing is quite large as it is in a vehicle because the generator is a axillary load and thus not determining the speed. That is why a wound field is used for the rotor, to control the magnetic flux and thus the output voltage. It is a lot simpler approach since the power needed is far less then that handled by the electronics of a PM wild frequency generator.
 

TSutrina

Well-Known Member
Joined
Mar 5, 2017
Messages
96
Reaction score
16
I should have said phenolic. You can get rods of bakelite, from artsy types if nothing else, but phenolic-linen would have better mechanical properties. (It's actually a fun web search -- there's a lot of interesting phenolic plastics out there. If I had a project that needed artsy knobs & levers, I'd have a collection of Bakelite rods to make them from).

Magnet iron is steel that's alloyed with silicon instead of carbon.

Steel that's hardened with carbon is both mechanically and magnetically hard. Magnetic hardness means that the steel will become a permanent magnet. It takes energy to coerce a hunk of steel into a magnet, so a magnetically hard steel causes loss each time the field reverses. This causes drag (there's actually a kind of motor that uses this to good effect -- do a search on "hysteresis motor").

Steel that's hardened with silicon gets mechanically hardened but it's hardened magnetically to a much lower degree. It's on the market as "magnet steel" or "transformer steel". It's also hard to get in less than industrial quantities; you'd basically have to luck into it. Unlike carbon steel, silicon actually lowers the ultimate strength of the steel, so it's not nearly as good a structural material as carbon steel -- but it's a worthwhile trade for a transformer or a motor.
The choice of the iron is greatly effected by the frequency of the changing magnetic flux and the the electrical frequency. For 50 and 60 Hz carbon steel doesn't present a big increase in losses. Aircraft magnetics at 400 Hz us silicon steels. And for higher frequencies the choices of materials decrease until you get to powdered metal oxides. The last choice is air. I have not studied the design. You always need a path for the magnetic flux. At reasonable frequencies magnetic engineers always choose a magnetic conducting material. The insulating materials are used where they do not want flux. I have use the simple 2D magnetic FEA program FEMM which has been around for decades for a windows program I think it is not available for other platforms. Finite Element Method Magnetics / Wiki / Magnetics Tutorial
 
Last edited:

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
Hi, ive not had much time for this project recently but i have settled on the final design now.
I have a satisfactory method of producing the coils and will add details of that soon.
i have done most of the pattern work now and hopefully soon will have somthing to show. But so far just patterns and hours and hours of sanding and painting.
The patterns are all 3d printed and will be made up as matchplate patterns on plywood boards.
20210414_143143.jpg
 

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
Hi all, managed to get some more done on this project.
I finished my little coil winding machine now and that turned into a little project of its own. But its a solid little machine that i hope does a lot of work.
20210502_155215.jpg

20210502_155236.jpg

The coil spools each have a scew that simply acts as a cam on the leaver to press a button on a cheep stock counter.
20210502_155257.jpg

The spool winders have a layer of teflon tape on the inner faces.
The coil bobbins are clamped between the two parts and then mounted onto the winding machine.
20210502_172505.jpg

I use two part liquid epoxy make the coils solid. The epoxy is added with a small brush as the coil is winding. The former can then be removed from the machine and left to cure.
20210502_172856.jpg

20210502_174939.jpg

The coils take about 1 minute each plus a little time setting up and gives me plenty of time to do a set of 12 before the epoxy starts setting after 30min.
Il see how they all turned out tomorrow and do some more tests. Ive now got some bigger magnets to try also these are 25x10x3 and much much more powerfull and also cheeper!
Not much more to show, but nearly finished all the paterns and have finished all the drawings now.
I should have this project wrapped up soon, im going to have a go at casting some bases tomorrow so will update on that.
Hers a pic of how the finished generator will actually look Like..
20210501_222020.jpg

Best regards.
Luke.
 

Richard Hed

Well-Known Member
Joined
Nov 23, 2018
Messages
1,031
Reaction score
228
Location
Seattle
Hi all, managed to get some more done on this project.
I finished my little coil winding machine now and that turned into a little project of its own. But its a solid little machine that i hope does a lot of work.
View attachment 125026
View attachment 125027
The coil spools each have a scew that simply acts as a cam on the leaver to press a button on a cheep stock counter.
View attachment 125028
The spool winders have a layer of teflon tape on the inner faces.
The coil bobbins are clamped between the two parts and then mounted onto the winding machine.
View attachment 125029
I use two part liquid epoxy make the coils solid. The epoxy is added with a small brush as the coil is winding. The former can then be removed from the machine and left to cure.
View attachment 125030
View attachment 125031
The coils take about 1 minute each plus a little time setting up and gives me plenty of time to do a set of 12 before the epoxy starts setting after 30min.
Il see how they all turned out tomorrow and do some more tests. Ive now got some bigger magnets to try also these are 25x10x3 and much much more powerfull and also cheeper!
Not much more to show, but nearly finished all the paterns and have finished all the drawings now.
I should have this project wrapped up soon, im going to have a go at casting some bases tomorrow so will update on that.
Hers a pic of how the finished generator will actually look Like..
View attachment 125032
Best regards.
Luke.
Too cool. Too cool!
 

ShopShoe

Senior Member
Joined
Oct 1, 2010
Messages
1,065
Reaction score
223
FYI - Shellac

You can still buy shellac. The traditional way is to buy shellac flakes and dissolve them in alcohol to the strength you need.

Google "Shellac Flakes" for sources and more information.

--ShopShoe
 

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
Thanks shopshoe, i did look into shellac but when searching ebay it mostly showed stuff for furniture all different with added beeswax and the like. So wasnt in the mood for experimenting and wasting money. I also looked into baking the coils with ethonal but seams like more work. The epoxy was less then a fiver delivered and is enougth to make a hundred coils.
The coils turned out nice and solid and uniform..
20210504_072214.jpg
 

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
Hi all, a bit more done on this today.
I have plenty of coils to do more experiments with now, i made another test stator housing and did a nice neat job of wiring it up. But it was late in the day and ive gone and wired it all wrong so its usless and will start again tomorrow.
20210504_072221.jpg

20210504_170112.jpg

Also i need to make a 12 pole test rotor now as ive decided on making the unit single phase ac. I got these neat little single phase bridge rectifier units to convert to DC. they come as kits that just require soldering and fit nicely in the base cavety.
20210504_114520.jpg

20210504_163627.jpg

20210504_163642.jpg

More work done on the patterns. All the patterns have holes for using pins to alighn the two sides of the matchplates. I nake them with the runners and gates included and just have to add the sprue holes when building the mold. Pics will explain. Still need a little filler on the holes then one more round of paint and sanding.
20210504_141233.jpg

20210504_141344.jpg

20210504_141646.jpg

20210504_144958.jpg

I should have some more to show tomorrow.
Regards.
Luke.
 

Richard Hed

Well-Known Member
Joined
Nov 23, 2018
Messages
1,031
Reaction score
228
Location
Seattle
Hi all, a bit more done on this today.
I have plenty of coils to do more experiments with now, i made another test stator housing and did a nice neat job of wiring it up. But it was late in the day and ive gone and wired it all wrong so its usless and will start again tomorrow.
View attachment 125152
View attachment 125153
Also i need to make a 12 pole test rotor now as ive decided on making the unit single phase ac. I got these neat little single phase bridge rectifier units to convert to DC. they come as kits that just require soldering and fit nicely in the base cavety.
View attachment 125154
View attachment 125155
View attachment 125156
More work done on the patterns. All the patterns have holes for using pins to alighn the two sides of the matchplates. I nake them with the runners and gates included and just have to add the sprue holes when building the mold. Pics will explain. Still need a little filler on the holes then one more round of paint and sanding.
View attachment 125157
View attachment 125159
View attachment 125160
View attachment 125162
I should have some more to show tomorrow.
Regards.
Luke.
Luv it!
 

darwenguy

Well-Known Member
Joined
Oct 26, 2013
Messages
115
Reaction score
88
Hi all, today i made a full test rig with the full 12 coils and a new rotor with 12 poles.
The new magnets are much more powerfull and care must be taken as they can bite! When they snap together They are hard to seperate.
For this test it is setup as single phase AC, the 12 coils have 100 turns of .25mm wire each and are wired in series with each coil alternating in the winding direction. The two ends are wired into the bridge recfifier and a digital voltage display is wired to the DC output of the rectifier.
The generator is spun via the lathe with equal sized pullys and a drive band so that the lathe rpm display is rougthly the same as the generator rmp.
I had good results and this is the readings at the dc output...
At 500rmp =12v dc and 1.8amp
At 1000rpm=24vdc and 3.6amp
It will even produce 5v at 250 rpm.
As low as 12v a good spark can be seen across the terminals.
The voltage display isnt very acurate and draws some voltage but good for display for the photo.
20210505_141108.jpg

20210505_141116.jpg


20210505_141807.jpg

Im quite pleased with these results and the final model should be a little better as the air gap between coils and magnets will be half the size.
Regards.
Luke.
 

Steamchick

Well-Known Member
HMEM Supporter
Joined
Apr 2, 2020
Messages
1,013
Reaction score
338
Location
Sunderland , UK
Hi Luke, looks good! It generates volts. But the real test is a plot of volts against amps - say at set speeds. 200rpm, 400rpm 600rpm, 800 rpm 1000rpm. For generator analysis you can not rely upon just volts. Really that is just a speed relationship as it is dB/dt. (B = Magnetic flux intensity) but generators need to produce amps as well. You can but a combined voltmeter and a meter digital unit for just a couple of £. Then you have the tool to determine how powerful the generator really is. But be careful. As all the windings are effectively in series, the same current flows in them all, and too much current will simply burn-out coils. So check the current rating of the wire and DO NOT EXCEED THAT CURRENT when testing across a load. For varying loads, 12v car bulbs are pretty good resistors. The should also mean you can easily study current by connecting more bulbs in parallel.
This is the sort of size for a 12V. Generator for a motorcycle of 100 to 250 rating. But without iron cores you won't expect to get anywhere near that. My guess is only 100dth.of that power. Not over 1A. But prove me wrong?
 

Latest posts

Top