Bedini Ferris Wheel Regauging Motor

[John_Bedini]

Bedini GT3

Itsu,
That is right thanks for posting that as I have no more room to put diagrams up.
Thanks.
John B

[Ron Chase]

Scott and RS,

This what we know for sure:

1. All four wires are wound together on center coil, wires not twisted.
2. Definetely only two wires coming out of each coil. I was at the conference and can verify this.
3. All #15 wires of the coils are hooked in parallel and terminate at two posts on the back of the switching control box.

The question remains as to how a magamp function can be achieved if the control wires are also connected in parallel with the power wire?

Thoughts anyone?

Ron

[John_K]

Updated Schematic

Hi All,

Thanks Itsu and John B for the corrections...

Updated schematic attached.


John K.

[nvisser]

I just inverted the polarities on my 2 slave coil so they atract. I also adjust the coils accordingly. It runs much better. Much better charging than before.
Now after the latest info from John it looks like the slave coils should be double the resistance than all combined parallel wires of the main coil. For some reason I read it the other way round and have my slave coils half the resistance of the main coil. I will change it by winding an extra winding on my main coil to lower the resistance and see what the improvments are.

[John_Bedini]

Bdini GT3

Nvisser,
This is one way to do this, Good thinking Nvisser keep going.
John B

[txaggie00]

Nvisser,
This is one way to do this, Good thinking Nvisser keep going.
John B

@John B,

We are not quite there yet, but you said that if you started your wheel CW, it would hit the magnetic field and jump back in reverse. That's how you knew the timing was right. My question is... Is it the right slave coil that is hooked in parallel in reverse polarity? Or both as Nvisser pointed out? Also, you had mentioned one of the control wires being reversed. Am I understanding this correctly? Is this how you bias the main coil?

Thanks as always,
Scott

[BrentA929]

Center coil

John B,

There has been lots of speculation recently in regards to the center coil and how the windings are connected...

Can you clear up what you have said about the slave coils being half the impedance of the center coil, yet your measurements that you gave us a couple of days ago are very different.


Thanks, Brent

[genessc]

I'm not sure its about it working like a mag-amp so much as its about being able to separate the functions of the basic SG into two unidirectional flows.

The slave coils are affected by the rotor magnets BEFORE the main center coil is affected. This means that the incoming magnet induces a current flow in the slave coils, which is also somehow routed thru the main coil in the opposite direction to the flow that the power pulse will cause on the main coil.

In the SG this flow is on the trigger wire. It is there and is in motion turning on the base of the transistor for the collector pulse to fire which inverts the flow on the trigger winding causing the transistor to instantly turn off. However the trigger strand is actually rigged to freewheel with the 1n4001 diode from emitter to base and so it does till the rotor brings in the next flow which returns the flow on the base winding to what it needs to be to start this cycle again.

So perhaps the slave coils are there to induce this flow on the main coil to act as the Tensioner flow which is then Hit hard and inverted by the repelling pulse that pushes the magnet over TDC of the center coil and moves the rotor along. (this might be why the rotor is noted to get slightly pushed backwards before taking off forwards?)

In all my experiments aimed at trying to reproduce the SG's "hyper" charging effect on a capacitor, only the inductive trigger could induce this effect. I tried optos, reed, relays, etc and none of them let the same effect manifest, and the main difference that I observed was the flow on the trigger strand that was removed due to the new means of triggering the switch.

I've yet to still just put a small directional flow on a 1:1:1 matched coil and then hit it with a larger pulse in opposite direction on a separate strand to then see if the same "hyper" charge effect was noted.

Anyways thats how I am sorta looking at it. Maybe it will be helpful for others too.

Take care,
Gene

Scott and RS,

This what we know for sure:

1. All four wires are wound together on center coil, wires not twisted.
2. Definetely only two wires coming out of each coil. I was at the conference and can verify this.
3. All #15 wires of the coils are hooked in parallel and terminate at two posts on the back of the switching control box.

The question remains as to how a magamp function can be achieved if the control wires are also connected in parallel with the power wire?

Thoughts anyone?

Ron

[John_Bedini]

Bedini GT3

Txaggie00,
My machine runs counter clockwise, it does not make any difference which way the machine runs. But what ever way the machine goes when the timing is right the machine will go backwards stop and then go the direction you have chosen.
You can use repulsion and attraction if you want to do that and that was my answer to Nvisser and nothing more. I'm not recommending that you do that but you can if you want too.
My machine is in repulsion so everything is wired the same.
John B

[genessc]

Just to Add. A mag-amp is in effect a "current switch". If a DC current is applied to a bias winding, the AC flows current will then be allowed to drive the load. (note I said current, not volts... The sinewave will be allowed thru even if the DC bias is off, however it has no Current to back the volts so the smallest load will drag that sine to a zero base line.)

I have since verified with my Mag-amp that using a cap to collect charge for the DC bias doesn't make any difference to how long the on-time duty needs to be for the home-made mag amp, using 1000va E-I core trafos, needs to be to run the load. It still needs to be ON 98%+ to get the full AC current to the load. (which at this point is pointing to no switching on the DC bias for this test setup.)

I will also note that this is not an I core mag-amp. All the flux is held inside these trafos via their E-I core setup. Thus it can't be directly compared with a I core mag-amp... I don't know that much literature exists on using open cored laminates for mag-amps. I do see the slight gap in flyback trafo ferrites tho, What is the gap for? Is that what lets the magnetic flux drop at a much faster rate than if the core were wholly enclosed?

Interesting thread... Keep up the good work all.

Gene

I'm not sure its about it working like a mag-amp so much as its about being able to separate the functions of the basic SG into two unidirectional flows.

The slave coils are affected by the rotor magnets BEFORE the main center coil is affected. This means that the incoming magnet induces a current flow in the slave coils, which is also somehow routed thru the main coil in the opposite direction to the flow that the power pulse will cause on the main coil.

In the SG this flow is on the trigger wire. It is there and is in motion turning on the base of the transistor for the collector pulse to fire which inverts the flow on the trigger winding causing the transistor to instantly turn off. However the trigger strand is actually rigged to freewheel with the 1n4001 diode from emitter to base and so it does till the rotor brings in the next flow which returns the flow on the base winding to what it needs to be to start this cycle again.

So perhaps the slave coils are there to induce this flow on the main coil to act as the Tensioner flow which is then Hit hard and inverted by the repelling pulse that pushes the magnet over TDC of the center coil and moves the rotor along. (this might be why the rotor is noted to get slightly pushed backwards before taking off forwards?)

In all my experiments aimed at trying to reproduce the SG's "hyper" charging effect on a capacitor, only the inductive trigger could induce this effect. I tried optos, reed, relays, etc and none of them let the same effect manifest, and the main difference that I observed was the flow on the trigger strand that was removed due to the new means of triggering the switch.

I've yet to still just put a small directional flow on a 1:1:1 matched coil and then hit it with a larger pulse in opposite direction on a separate strand to then see if the same "hyper" charge effect was noted.

Anyways thats how I am sorta looking at it. Maybe it will be helpful for others too.

Take care,
Gene

[gyula]

....
I don't know that much literature exists on using open cored laminates for mag-amps. I do see the slight gap in flyback trafo ferrites tho, What is the gap for? Is that what lets the magnetic flux drop at a much faster rate than if the core were wholly enclosed?

Interesting thread... Keep up the good work all.

Gene

Hi Gene,

In case of flyback transformers (used in the horizontal output stages of CRT television receivers or CRT computer monitors) the air gap is used for preventing the ferrite core saturation. You ask if the gap is for the quicker flux collapse, I think it is possibly so but this can be only of a secondary importance, the main reason is to avoid saturation.

In a magamp the DC bias current is just for controlling / shifting the magnetic operation point of the core on its B/H curve, bringing the core towards or away from the saturating condition. This way the AC coils 'see' a changing permeability, hence they will have a changing (XL) inductive reactance.

So your question on open core magamps could be answered as there is no much sense for using open cores because a very huge DC input current would be needed for influencing the B/H curve, this cannot be a goal.

rgds, Gyula

[genessc]

Hi Gyula,

Its been a little while eh? Glad to hear from you.

The mechanism then for preventing saturation of the ferrite is to keep the ferrite from being able to establish the same density of current all the way around the *mostly* closed loop. This prevents it from saturation which means what? What is saturation of a ferrite core and what are the effects?

In my benched mag-amp, No DC on the bias winding leaves the triple trafos then unsaturated which implies that the inductance of the asymetry is whats keeping the current from closing its loop thru the load.

Just shorting the dc bias winding is the same as leaving no DC power on the bias winding as it doesn't affect the inductance of the transformers in mag-amp config either. (which is why this winding is connected in bifilar common end connected config to "null" the inductance that it would generate if it were connected in bifilar series connected fashion.)

This implies that if the AC is meant to flow the DC bias winding must "saturate" the ferrite cores so that the asymetry of the dual trafos on one side of the load trafo becomes as though its non-existant. That would explain why it needs a good sized current of ~4-4.5amps at ~9vdc to "bias" and saturate those two trafo cores.

Its not so much that the "AC coils 'see' a changing permeability, hence they will have a changing (XL) inductive reactance." but is more that the DC bias on the bias winding causes saturation which removes the asymetric inductance on one side of the AC windings output. The AC has the changing inductive reactance because its AC. Its able to flow because the DC bias has saturated its windings on its two transformers and thus made "transparent" the asymetric inductance that was gating the current on the AC load.

One thing that isn't accounted for is a rotational magnetic field provided by perm magnets on the rotor as the method for changing the inductive reactance of the coils. As I've noted on a prior post to this thread (and another I believe found the same, Jerdee.) if you put an LCR meter on an inductance... and swipe a magnet past the inductance you see the inductance go UP, instead of down which is what shorting the inductance does, as well as applying a DC potential to the inductance (albeit about 1/4 of a shift in reduced inductance.)

Take care, Thanks for the reply,
Gene

Hi Gene,

In case of flyback transformers (used in the horizontal output stages of CRT television receivers or CRT computer monitors) the air gap is used for preventing the ferrite core saturation. You ask if the gap is for the quicker flux collapse, I think it is possibly so but this can be only of a secondary importance, the main reason is to avoid saturation.

In a magamp the DC bias current is just for controlling / shifting the magnetic operation point of the core on its B/H curve, bringing the core towards or away from the saturating condition. This way the AC coils 'see' a changing permeability, hence they will have a changing (XL) inductive reactance.

So your question on open core magamps could be answered as there is no much sense for using open cores because a very huge DC input current would be needed for influencing the B/H curve, this cannot be a goal.

rgds, Gyula

[Ron Chase]

Magamp

Thanks Gene and Gyula for your insights. I don't see how a classic magamp function is possible with the present main and slave coil configuration so it must be in the direction of unidirectional flow as you mentioned Gene. Let us know your results with the 1:1:1 coil experiment.

Best regards, Ron

[RS_]

Hi all,

I can see that the incoming magnets on the slave coils are providing a mag amp like bias on the main coil because of their offset, but why call the 2 18ga wires, control wires, when they are drive wires, unless their size difference makes it work different somehow, that's the part that does not make sense to me.....

So, I thought that maybe JB was doing something more like a traditional mag amp, and providing a major B/H bias with the 2 Control wires at different points in the duty cycle..... and if so, without using the 18ga, that was gonna change the Drive coil lengths to get the same main coil DCR, but what do I know.......I thought I would throw it out there, and see if I understood what JB is doing or not........

As it seems I am TOTALLY Wrong, here are the numbers for all 6 strands in parallel, down to the centimeter, based on a AWG Wire Ga table from Rea Magnet Wire Co, and crunched with my Mathcad magnet wire calculator.....

hope this helps......

Main Coil
for 15 Ga wire, 1150.5 meters = 3774.6 Ft = 12ohms at 37.843Lb
2 in parallel on the main coil = 6 ohms at 75.686 LB

18 Ga on main coil 1150.5 meters = 3774.6 Ft = 24.1 ohms at 18.952 Lb
2 strands of 18ga parallel = 12.05ohms 37.904 Lb

Main Coil = 4ohms Weight = 113.59 Lb

Slave Coils
15Ga 978 meters = 3208.661 Ft = 10.2ohms at 32.165 Lb

2 Slave coils in parallel = 5.1 ohms at 64.33 LB

2 strands of 15ga and 2 strands18ga in parallel on main coil, and 2 Slave coils = 2.242 ohm total DCR, 3 Coil total weight = 177.92 Lb

Sorry for any confusion...... I just want to understand the mag amp feature, and why the control wire name......

I'll go back to my soldering now.......

RS

[John_K]

Hi RS,

Your numbers look good, thanks for working that out. JB did mention that the cct works best between 2 and 30 ohms, so you're definitely in the ball park.

The way I see it is the same as you, that the the slave coils are the control wires for the "like a mag amp" (as JB put it). I can only suggest that JB might have initially thought to use the 2 x #18 wires on the main coil as control wires but perhaps changed his mind by using the slaves instead, using their offset to bias the main coil instead of additional circuitry to use the 2 x #18 wires in DC bias mode. I'm not sure though, just speculating.

@All,

I added the 15V zener and 1.5K resistor to the hall circuit last night and WOW!, what a difference it made! Previously, when the hall was powered off a 12V rail the main coil would pull around 2A. Now, with the zener/resistor change the main coil pulls around 5A. This translates into a much larger spike on coil discharge and obviously a faster charging cap, which dumps more frequently and charges much better. BTW, I only had the 2 thicker wires on the main coil connecetd in parallel. My slave coils were not connected, nor were the 2 thinner wires on the main coil.

The only problem I have is that I need to tame down the dwell to get the duty cycle down. I BBQ'd the MJL21194 after about 15 minutes, they are good up to around 150 degrees celsius so that kind of tells you how much power was going through it. The MJL21193 was not as hot, which tells me the beta of the pair was not matched. I'm going to find a couple of matched MJL complimentary pairs to share the load more evenly, which will also mean that both MJLs should switch at the same time.

JB said he did not use any heatsinks and that his transistors ran cold, so I need to figure out why. Of course, he is using a lot more wire than me, so the current is going to take longer to catch up which is probably why his current draw is a lot lower.

But I do strongly suggest the zener/resistor suggestion for the hall circuit, I could not believe it would make that much difference.


John K.

Hi all,

I can see that the incoming magnets on the slave coils are providing a mag amp like bias on the main coil because of their offset, but why call the 2 18ga wires, control wires, when they are drive wires, unless their size difference makes it work different somehow, that's the part that does not make sense to me.....

So, I thought that maybe JB was doing something more like a traditional mag amp, and providing a major B/H bias with the 2 Control wires at different points in the duty cycle..... and if so, without using the 18ga, that was gonna change the Drive coil lengths to get the same main coil DCR, but what do I know.......I thought I would throw it out there, and see if I understood what JB is doing or not........

As it seems I am TOTALLY Wrong, here are the numbers for all 6 strands in parallel, down to the centimeter, based on a AWG Wire Ga table from Rea Magnet Wire Co, and crunched with my Mathcad magnet wire calculator.....

hope this helps......

Main Coil
for 15 Ga wire, 1150.5 meters = 3774.6 Ft = 12ohms at 37.843Lb
2 in parallel on the main coil = 6 ohms at 75.686 LB

18 Ga on main coil 1150.5 meters = 3774.6 Ft = 24.1 ohms at 18.952 Lb
2 strands of 18ga parallel = 12.05ohms 37.904 Lb

Main Coil = 4ohms Weight = 113.59 Lb

Slave Coils
15Ga 978 meters = 3208.661 Ft = 10.2ohms at 32.165 Lb

2 Slave coils in parallel = 5.1 ohms at 64.33 LB

2 strands of 15ga and 2 strands18ga in parallel on main coil, and 2 Slave coils = 2.242 ohm total DCR, 3 Coil total weight = 177.92 Lb

Sorry for any confusion...... I just want to understand the mag amp feature, and why the control wire name......

I'll go back to my soldering now.......

RS