Double window motor/generator build journal

[Bradley Malone]

don't have any hard readings...everything is to unstable at full speed because of the wobble from the magnet being off center. but it seems to pull somewhere around half an amp untill it gets up to speed then it levels off slightly below that but i was using a 10 amp panel meter so not very accurate. That's running on 12 volts. it got up to about 1000 rpm and it just shook to much for me to want to push it. so back to the bench to figure out a way to balance this thing. the way i built it i had to slide the shaft through the coil and the magnet and glued the magnet in place. So now i can't just pull the shaft out to balance it i have to do it while its in the coil. Pain in the neck but its all a lesson. on one hand though the torque on this thing seems to match what i have heard about window motors. for drawing as little as it does it has got quite some force behind it.

[Bradley Malone]

Well today has been better. after fighting with the thing for about 4-5 hours this morning, I got everything balanced. spun the thing up and it hit 30,000 rpm if my math is correct. I used a small coil hooked up to my multimeter which has a Hz setting. it climbed to a peak just over 500 Hz which using a normal 2 pole magnet that is 500 revolutions a second multiplied by 60 seconds in a minute equals 30,000 rpm! Also I have the diode from the collector through a neon and if i spin it up then hold the reed in different locations sometimes I can get the purple flash I have seen in coil shorting videos.

EDIT; math was right but the measurement wasn't...my trigger magnet was adding extra pulses to the readings. i put an isolated magnet on the shaft away from the others and it read around 200 hz which adds up to about 12,000 which i am still happy with seeing as i am only using 1 pulse per revolution untill i get my bedini/cole circuit running. back to the bench!

[Bradley Malone]

Tidied up a bit...still looks like wires everywhere, but everything is in its place. I used the connector strip along with the alligator clips so that if I burn out a transistor I just unscrew it and put in a new one. Another thing I change is I separated the bi-filar coil into its separate coils and hooked one up reversed to its own transistor so that I could use 2 reeds and have a bipolar circuit. When I measure the current draw though my digital multi-meter it reads steady at 39 miliamps running from a 12 volt drill battery. If I change the timing to close to the zero point (I mean the point where the poles 90 degrees from the coil plane) the speed is much greater with less torque and draws around 28 miliamps. When I run it at the high speed point the caps stop charging at 12 volts, But when I run it at the higher torque lower speed timing the cap (880uf 250v bank of 3 caps in parallel) charges to 30 volts in a second maybe two and just keeps going. That's my next step is to time the charge rates at the different timing points until I find the best location. I also plan on making something to accurately read the speed. No cash to get a tach at the moment.

[Bradley Malone]

https://youtu.be/b_1rAs01AEs simple video of it running. i have noticed that there is only one spot where the voltage will just continue to climb on the cap. if i get it off that point it just stops at the charge its on thats why after that long i only had 29 volts in the cap because trying to film while holding the reeds i couldn't keep it in the right spot.

[Bradley Malone]

Tom, hoping you can chime in.
I have read post where you said that the window motors run best when loaded to 2/3 of its unloaded speed. By that do you mean that if you have a generator coil that slows the motor to that speed to get the best results via the generated voltage plus what you get from the CEMF into a cap. or are you saying that loading the motor somehow increases the CEMF into the cap. I can understand the first because loading the motor doesn't change the current much ,if I am correct, so you could load it with a generator without changing power in thus getting more. On the other hand though I see JB load his machines with fans...so i am just hoping to get a bit of clarification on that.

Another update, I built a small setup with a push button toggle switch, and a double pole single throw relay. The way I hooked it up allows me to use a small bank of 5 x 2.7v 5F super caps, in series to make 13.5volts, as the drive power to run the motor. While the motor is running off the super caps it is charging a battery. When the motor starts to slow down I push the toggle and it flips the relay while disconnecting the battery negative from the positive of the caps and when the relay switches it dumps a fresh charge into the super caps in about half a second and when the toggle comes back out its back to running and charging the battery. i don't have everything tuned yet but when i was doing this The battery started out at 12.99, I would tap the toggle and charge the caps about every minute. The charge would drop the battery to about 12.88 then it would quickly climb back up to about 12.95 then creep back up to 12.99. I did this for an hour and the battery ended up at 12.90. One thing I noticed was at one point in the voltage drop of the super caps while driving the motor it would slow as the voltage dropped and the battery was charging, But it would reach a point where the rate at witch it slowed down ...slowed down (sorry if that's confusing) When it would reach about 5 volts it would maintain that speed longer and it seemed the current draw dropped considerably because it will run at that voltage for while...but it won't charge the battery once it does this leveling out type of thing.

[Tom C]

Bradley,

the window motor is a torque motor, its best efficiency (this is from what I have heard John say) is you run it as fast as it will go, load it down to 2/3 of its fastest speed, that is the most efficient range of operation, current draw goes down a bit and work gets done. A window motor under load is not a self runner, its just very efficient. you can adjust current draw thru timing and load. I have never seen a window motor doing other work besides running itself go over unity. and I have only seen one self power. and here lies the problem, its not ENOUGH to have it self run, you need it to self run while powering other loads.

the fan on an SG is another example of adjusting current draw, the fan slows the machine a bit, and it does work by moving air. not a lot of work but some, it is one method of power extraction.

Tom C

[Bradley Malone]

Thanks Tom your always a help. Is that the fastest it will go at any timing because like i said before there is a point where it is very fast but has little torque and won't charge the caps higher than 12 volts, but at a different point it runs slower with lots of torque and charges caps very quickly. so am I supposed to go with the high torque timing and base everything off the top speed from that point.

As far as it not being overunity...I am fine with that for now but I don't see many people mixing these high efficiency systems and that is what I am shooting for. Once I have the window motor running at its peak (knowing it needs a load). I plan to use that requirement to try out other high efficiency generators or lenzless generators and things like that to see what I can get out. Because if the window motor runs very efficiently while driving a low lenz generator I am thinking there could be possibilities. I have the first coil I made with the copper tape, because its flat copper I couldn't bend it around the shaft so the shaft had to be inside the coil witch has me stuck with a timing issue until I get hall sensors, because reeds switch on both poles. but if I take that coil and set it right on top of the other coil (without the other magnet) it becomes a normal generator coil, so if i need to load the motor to reduce current with a load...why not cause the current using the gen as a load? I have multiple taps on the coil so between that and just pulling the generator further away to find the right spot where the power generated loads the motor just right. One last thing, I have not verified this but from what I have read about coil shorting if done correctly it does not load the rotor so I am going to also try using the flat coil on top with a shorting circuit as a secondary output. Also has anyone tried running a window motor from a tesla switch?

[Tom C]

you will just have to experiment.

Tom C

[Bradley Malone]

Already doing that , I just did not know if I was missing something with the high speed no-charging spot. But im finding the happy point and seeing whats happening I put a cone shaped toroid (from a crt yoke coil) around the shaft outside the motor coil and it will output 6 volts at 25ma without changing the motor speed hardly any. It does not have much wire on it though so I may rewind it with more.

[min2oly]

Thanks Tom your always a help. Is that the fastest it will go at any timing because like i said before there is a point where it is very fast but has little torque and won't charge the caps higher than 12 volts, but at a different point it runs slower with lots of torque and charges caps very quickly. so am I supposed to go with the high torque timing and base everything off the top speed from that point.

As far as it not being overunity...I am fine with that for now but I don't see many people mixing these high efficiency systems and that is what I am shooting for. Once I have the window motor running at its peak (knowing it needs a load). I plan to use that requirement to try out other high efficiency generators or lenzless generators and things like that to see what I can get out. Because if the window motor runs very efficiently while driving a low lenz generator I am thinking there could be possibilities. I have the first coil I made with the copper tape, because its flat copper I couldn't bend it around the shaft so the shaft had to be inside the coil witch has me stuck with a timing issue until I get hall sensors, because reeds switch on both poles. but if I take that coil and set it right on top of the other coil (without the other magnet) it becomes a normal generator coil, so if i need to load the motor to reduce current with a load...why not cause the current using the gen as a load? I have multiple taps on the coil so between that and just pulling the generator further away to find the right spot where the power generated loads the motor just right. One last thing, I have not verified this but from what I have read about coil shorting if done correctly it does not load the rotor so I am going to also try using the flat coil on top with a shorting circuit as a secondary output. Also has anyone tried running a window motor from a tesla switch?

You can force the reed to fire on south or north only. Place a "biasing" magnet close to the reed, when your timing magnet approaches you'll see how it works 😃
http://www.digikey.com/Web%20Export/...ent_magnet.pdf

[Bradley Malone]

You can force the reed to fire on south or north only. Place a "biasing" magnet close to the reed, when your timing magnet approaches you'll see how it works ��
http://www.digikey.com/Web%20Export/...ent_magnet.pdf

Awesome! I did not know that. thank you for your info. I won't have to worry about the reeds anymore though I got all the parts I need for a hall triggered b/c circuit in the mail today (been waiting about a month) I'm beta matching transistors now and looking for a pcb layout that I saw a while back. But that is very handy for a few other projects im working on.

[Bradley Malone]

I have my bedinin cole circuit about 95% done and im stuck on something. In johns icehouse ciruit there is a voltage divider for the hall sensors...but there are no values. On another circuit i have seen of Johns it has 12 volts straight to the hall. So is there a reason for the resistors in the bipolar circuit other than limiting the power that the hall uses?

[Bradley Malone]

Up and running with the B/C bipolar circuit. I just put power straight to the halls and it works fine. I will look into the resistors later to see if I can save a bit more energy. The charging is way better with this than the reeds, And it seems to run smoother. I may have been adding extra pulses in with the reed, but with the halls everything is nice and precise. Now to start measuring

[Tom C]

look at your hall sensor data sheet see what voltage and current it likes.

Tom C

[Bradley Malone]

I did...they take 6ma each at up to 24 volts. i have them hooked straight to 12 volts and it draws 12ma if I don't have the rotor spinning. so it appears that the voltage regulator in them has everything where it needs to be. what throws me off is the circuit has 3020 halls witch is what I am using...and if I add resistors to give the halls what they are already doing themselves it just seems redundant. Like I said though everything runs without them so I will try with them in later test