Vandusen Bessler wheel replication
Moderator: scott
-
- Enthusiast
- Posts: 72
- Joined: Mon Nov 17, 2003 11:38 pm
- Location: Berlin
- Contact:
re: Vandusen Bessler wheel replication
Hi Jim,
what did you do to the parameters, when you wrote bad program ?
I have the feeling, that you need to get bigger weights and shift
them more far in and out.
And as in Darrel´s design the lock and unlock
mechanism is special ( NDA secrets..) I guess
your simple ramps do not apply to his wheel.
As I said, one shift of the weight is free as it can be
supported via springs, so at 12 o clock you just have a jump
outside powered via springs, so no need for any energy over there.
Just the compression at 6 to 7 o´clock.
If you would post the WM2D file I could have a closer
look at it.
Thanks.
what did you do to the parameters, when you wrote bad program ?
I have the feeling, that you need to get bigger weights and shift
them more far in and out.
And as in Darrel´s design the lock and unlock
mechanism is special ( NDA secrets..) I guess
your simple ramps do not apply to his wheel.
As I said, one shift of the weight is free as it can be
supported via springs, so at 12 o clock you just have a jump
outside powered via springs, so no need for any energy over there.
Just the compression at 6 to 7 o´clock.
If you would post the WM2D file I could have a closer
look at it.
Thanks.
re: Vandusen Bessler wheel replication
What you see in my last post is an exact replication of Darrell's 8 foot wheel that he cannibalized from the 6 foot wheel, as close as I could determine from his videos, his many still pictures and from conversations with him. I've seen the lock and unlock mechanism. Adding bigger weights would give more power only if the original design worked. Darrell tried that and he tried making the wheel a bigger diameter all to no avail.
Please don't insult me by suggesting that I don't know what his wheel is like.
I have not seen or talked in detail to him about his latest Smith Falls wheel. If he has changed the design to add springs I'm unaware. There are springs attached to the latches to hold them latched until they are released by the cams inside the ramps.
The poor video that you have is from his 3rd from last wheel. I have 13 videos from Darrell. Most are duplicates. He would forget which ones he had sent me and would resend them.
Has he sent you pictures of his last build?
I forget what the Bad Parmeter was. It stretched the separator or the rope because clearances weren't right.
Edit: I have 12 videos with ony one duplicate. My brand new computer won't play .mov files without some sort of 35 megabyte upgrade from Quicktime which means a 3 hour download. Bummer! (but my old computer is still working)
Please don't insult me by suggesting that I don't know what his wheel is like.
I have not seen or talked in detail to him about his latest Smith Falls wheel. If he has changed the design to add springs I'm unaware. There are springs attached to the latches to hold them latched until they are released by the cams inside the ramps.
The poor video that you have is from his 3rd from last wheel. I have 13 videos from Darrell. Most are duplicates. He would forget which ones he had sent me and would resend them.
Has he sent you pictures of his last build?
See above download labeled "GEGSAssyA666.wm2d"Stefan wrote:If you would post the WM2D file I could have a closer look at it.
I forget what the Bad Parmeter was. It stretched the separator or the rope because clearances weren't right.
Edit: I have 12 videos with ony one duplicate. My brand new computer won't play .mov files without some sort of 35 megabyte upgrade from Quicktime which means a 3 hour download. Bummer! (but my old computer is still working)
Last edited by jim_mich on Wed Jul 19, 2006 12:59 pm, edited 1 time in total.
-
- Enthusiast
- Posts: 72
- Joined: Mon Nov 17, 2003 11:38 pm
- Location: Berlin
- Contact:
re: Vandusen Bessler wheel replication
Hi Jim,
many thanks for posting the WM2D file.
I will have a look at it later this week.
Am currently pretty busy and just had put already too much time
over here to "defend" Darrell´s ideas.
I am no WM2D expert, but you are Jim, so I trust your words
on it. I will see, if I can get something to work
via WM2D, but if then the program is working right
or plays fool on me, I will not know...
Give me some time for it to test and play with it.
Many thanks.
many thanks for posting the WM2D file.
I will have a look at it later this week.
Am currently pretty busy and just had put already too much time
over here to "defend" Darrell´s ideas.
I am no WM2D expert, but you are Jim, so I trust your words
on it. I will see, if I can get something to work
via WM2D, but if then the program is working right
or plays fool on me, I will not know...
Give me some time for it to test and play with it.
Many thanks.
re: Vandusen Bessler wheel replication
Darrell and I have exchanged posts occasionally since the first of this year. His last post was on June 6. He was soliciting me to fabricate some spoke guide rollers for the weights to ride on.
I have not replied to this last post.
Ralph
I have not replied to this last post.
Ralph
re: Vandusen Bessler wheel replication
this guy has lost it, i picked this up from his forum.....!
hartiberlin
Administrator
Hero Member
*****
Posts: 1354
Stefan Hartmann, Moderator of overunity.com forum
View Profile Email Personal Message (Online)
Re: Overunity, Green Energy, Powered by Gravity!
« Reply #75 on: Today at 03:03:51 AM »
Reply with quoteQuote
Posted: Wed Jul 19, 2006 1:32 am Post subject: Re: re: Vandusen Bessler wheel replication
--------------------------------------------------------------------------------
Quote
ovyyus wrote:
Yes, I have the self-starting video where he walks up to the wheel, positions it, lets it go and it turns for less than 1/4 rotation. The full video is 000_0235.mov at 12,525 KB. You probably have a section of that.
Why does the wheel only turn about 1/4 rotation from that specific position? Because it is made top heavy in that specific position. Why doesn't Vandusen show at least one full rotation? Because the top heavy wheel will only accelerate for about 1/4 rotation from that specific position before it begins to slow and stop. Vandusen points to the fact that it makes 2 x 30 deg cycles knowing full well that that is only the case because he has made it top heavy in that position. Why? In order to create the impression that his wheel actually works when in fact it does not - so he can scam people for money.
Okay, I also got now this video.
Well, it is now totally clear what I am seeing.
Watch the weights and spoke´s ends !
The unlock system at 12 o ´clock is not working properly.
Thus in this video he first has to fix it at 2 o´clock, so that the wheel
can self starts. He must push it back to the right position.
That is at about 30 seconds in the videoclip and then the wheel
actually accelerates !
You see then, that the whole
torquearm of the right side is heavier than the whole left side !
Now then the wheels turns and accelerates and
just about before the end
at second 53 to 54 you see a weight again at around 1 to 3 pm
again not having resetted and unlocked and is not in the outer position,
so at this position it is loosing the heavier torquearm,
cause the upper pushout mechanism has failed for this weight.
This is why it slows a bit down in this special video 54 second long video...
Now as I have seen this video this even tells me even more, that this
thing works !
Quote
Michael wrote:
Stefan, you say the principle is valid. Please explain what the principle is so I and others can check the physics.
Thanks,
Michael
You always have 6 weights at the right side of the wheel 3 inches
more far away from the axis than on the left side.
That means enough torque to reset ( lock and unlock)
only 2 weights every 30 degrees by 1.5 inches.
And if you do it with springs you store the energy in one
place and get it for free in the other place so you have to spend
the energy only to unlock and shift one weight at a time ,
so you just need only to push in the weight at 6 to 7 o ´clock and
get the unlock at 12 o ´clock for free there if you use springs.
Then you only have to overcome the lift of one weight and the spring
compression at 6-7 o´clock via the heavier torquearm of the right side of the wheel.
This is why Darrell says there is at least a 12 spoke machine needed to do it,
otherwise the torque is not enough to lift the weight !
Hope this helps.
hartiberlin
Administrator
Hero Member
*****
Posts: 1354
Stefan Hartmann, Moderator of overunity.com forum
View Profile Email Personal Message (Online)
Re: Overunity, Green Energy, Powered by Gravity!
« Reply #75 on: Today at 03:03:51 AM »
Reply with quoteQuote
Posted: Wed Jul 19, 2006 1:32 am Post subject: Re: re: Vandusen Bessler wheel replication
--------------------------------------------------------------------------------
Quote
ovyyus wrote:
Yes, I have the self-starting video where he walks up to the wheel, positions it, lets it go and it turns for less than 1/4 rotation. The full video is 000_0235.mov at 12,525 KB. You probably have a section of that.
Why does the wheel only turn about 1/4 rotation from that specific position? Because it is made top heavy in that specific position. Why doesn't Vandusen show at least one full rotation? Because the top heavy wheel will only accelerate for about 1/4 rotation from that specific position before it begins to slow and stop. Vandusen points to the fact that it makes 2 x 30 deg cycles knowing full well that that is only the case because he has made it top heavy in that position. Why? In order to create the impression that his wheel actually works when in fact it does not - so he can scam people for money.
Okay, I also got now this video.
Well, it is now totally clear what I am seeing.
Watch the weights and spoke´s ends !
The unlock system at 12 o ´clock is not working properly.
Thus in this video he first has to fix it at 2 o´clock, so that the wheel
can self starts. He must push it back to the right position.
That is at about 30 seconds in the videoclip and then the wheel
actually accelerates !
You see then, that the whole
torquearm of the right side is heavier than the whole left side !
Now then the wheels turns and accelerates and
just about before the end
at second 53 to 54 you see a weight again at around 1 to 3 pm
again not having resetted and unlocked and is not in the outer position,
so at this position it is loosing the heavier torquearm,
cause the upper pushout mechanism has failed for this weight.
This is why it slows a bit down in this special video 54 second long video...
Now as I have seen this video this even tells me even more, that this
thing works !
Quote
Michael wrote:
Stefan, you say the principle is valid. Please explain what the principle is so I and others can check the physics.
Thanks,
Michael
You always have 6 weights at the right side of the wheel 3 inches
more far away from the axis than on the left side.
That means enough torque to reset ( lock and unlock)
only 2 weights every 30 degrees by 1.5 inches.
And if you do it with springs you store the energy in one
place and get it for free in the other place so you have to spend
the energy only to unlock and shift one weight at a time ,
so you just need only to push in the weight at 6 to 7 o ´clock and
get the unlock at 12 o ´clock for free there if you use springs.
Then you only have to overcome the lift of one weight and the spring
compression at 6-7 o´clock via the heavier torquearm of the right side of the wheel.
This is why Darrell says there is at least a 12 spoke machine needed to do it,
otherwise the torque is not enough to lift the weight !
Hope this helps.
re: Vandusen Bessler wheel replication
Stefan .. I'd just like to add to Bill's top heavy description of why it appears to accelerate thru a couple of sectors. Gordy's photo should help.
Anyone who has built a wheel where the spokes are not supported by a rim knows that it is notoriously difficult to get the dang wheel exactly balanced (symmetrical) before you even add weights & shifting mechs to it. A normal wheel with a rim almost guarantees that each spoke is a certain number of degrees apart & all the spokes are in the same plane.
When there is no rim to provide structural integrity, complete symmetry & vertical alignment, almost invariably if you were to measure the distances between each spoke point to point you would find some differences in the lengths measured. They require a lot of tweaking to get 'just right' in terms of distances apart between each spoke & to ensure that they are all aligned in the same vertical plane.
If they are not exact & symmetrical then effectively the wheel considers itself geometrically as two halves, where one half has more mass than the other. The wheel then wants to accelerate to move its GoG to below the axle to what we call the keel position.
Especially if you preposition it for demonstration purposes with the SAME spoke each time for maximum effect. If it could do that for each spoke tested sequentially, from the top position, from a standing start, it might lead to more meaningful conclusions about this ramped wheel design ?????
Anyone who has built a wheel where the spokes are not supported by a rim knows that it is notoriously difficult to get the dang wheel exactly balanced (symmetrical) before you even add weights & shifting mechs to it. A normal wheel with a rim almost guarantees that each spoke is a certain number of degrees apart & all the spokes are in the same plane.
When there is no rim to provide structural integrity, complete symmetry & vertical alignment, almost invariably if you were to measure the distances between each spoke point to point you would find some differences in the lengths measured. They require a lot of tweaking to get 'just right' in terms of distances apart between each spoke & to ensure that they are all aligned in the same vertical plane.
If they are not exact & symmetrical then effectively the wheel considers itself geometrically as two halves, where one half has more mass than the other. The wheel then wants to accelerate to move its GoG to below the axle to what we call the keel position.
Especially if you preposition it for demonstration purposes with the SAME spoke each time for maximum effect. If it could do that for each spoke tested sequentially, from the top position, from a standing start, it might lead to more meaningful conclusions about this ramped wheel design ?????
re: Vandusen Bessler wheel replication
Spot on Fletcher.
- ken_behrendt
- Addict
- Posts: 3487
- Joined: Wed Mar 02, 2005 7:45 am
- Location: new jersey, usa
- Contact:
re: Vandusen Bessler wheel replication
Buenas dias, muchachos...
I'm having a little insomnia, so I decided to take a shot at making a WM2D model of Vandusen's Wheel...or, at least, what I think it is supposed to be.
My model wheel is, like his earlier one that was claimed to work, 6 ft in diameter. After that, I made some variations to make it easier to excecute on WM2D, but I think that what I have is essentially the same as the designs he is using.
My wheel has an unloaded mass of 5 lbs and carries twelve 1 lb weights. Rather than use ramps and latching mechanisms, I decided to just attach the 12 weights to two sets of slots. To replace the ramp, I used a curved slot shaped so that the weights would be forced to orbit closer to the wheel's axle on the left or ascending side of the wheel during CW rotation. This curved slot is not part of the wheel, but, rather, fixed to the background behind the wheel. The weights are further forced to undergo radial motion toward and away from the wheel's axle by straight line slots which are attached to the rotating wheel. I find that the use of slots greatly speeds up the rate at which WM2D can compute simulation frames as compared to using actual parts to make ramps and latch mechanisms.
The 12 compression springs have an UNcompressed length of 1.5 ft, a minimum compressed length of 0.5 ft, and a K value of 2 lb/ft. These parameters assure that, neglecting CF, there will be enough energy stored in a spring to completely lift a weight near the 6:00 position up the ramp section there.
So, how did the model perform?
Well, it was not self-starting, so I had to attach a motor to its axle to rev it up in a CW direction. The motor was set to have a speed of -30 rpm's for only 5 seconds after the sim started. That is, after 5 seconds, the motor was disconnected from the model wheel's axle.
As can be seen from the rotational velocity graph for the wheel, as soon as the motor disconnected, the wheel began to lose speed. By about 40 seconds it had slowed enough to begin oscillating back and forth about an equilibrium position as the offset CG of the 12 weights was not providing enough torque to lift the 12:00 weight against the pull of gravity and the resistance of its compression spring. By about 200 seconds the wheel was completely motionless.
Needless to say, this quick model has reconfirmed my believe in the unworkabilty of various ramp designs...regardless of whether they use springs of not.
ken
I'm having a little insomnia, so I decided to take a shot at making a WM2D model of Vandusen's Wheel...or, at least, what I think it is supposed to be.
My model wheel is, like his earlier one that was claimed to work, 6 ft in diameter. After that, I made some variations to make it easier to excecute on WM2D, but I think that what I have is essentially the same as the designs he is using.
My wheel has an unloaded mass of 5 lbs and carries twelve 1 lb weights. Rather than use ramps and latching mechanisms, I decided to just attach the 12 weights to two sets of slots. To replace the ramp, I used a curved slot shaped so that the weights would be forced to orbit closer to the wheel's axle on the left or ascending side of the wheel during CW rotation. This curved slot is not part of the wheel, but, rather, fixed to the background behind the wheel. The weights are further forced to undergo radial motion toward and away from the wheel's axle by straight line slots which are attached to the rotating wheel. I find that the use of slots greatly speeds up the rate at which WM2D can compute simulation frames as compared to using actual parts to make ramps and latch mechanisms.
The 12 compression springs have an UNcompressed length of 1.5 ft, a minimum compressed length of 0.5 ft, and a K value of 2 lb/ft. These parameters assure that, neglecting CF, there will be enough energy stored in a spring to completely lift a weight near the 6:00 position up the ramp section there.
So, how did the model perform?
Well, it was not self-starting, so I had to attach a motor to its axle to rev it up in a CW direction. The motor was set to have a speed of -30 rpm's for only 5 seconds after the sim started. That is, after 5 seconds, the motor was disconnected from the model wheel's axle.
As can be seen from the rotational velocity graph for the wheel, as soon as the motor disconnected, the wheel began to lose speed. By about 40 seconds it had slowed enough to begin oscillating back and forth about an equilibrium position as the offset CG of the 12 weights was not providing enough torque to lift the 12:00 weight against the pull of gravity and the resistance of its compression spring. By about 200 seconds the wheel was completely motionless.
Needless to say, this quick model has reconfirmed my believe in the unworkabilty of various ramp designs...regardless of whether they use springs of not.
ken
On 7/6/06, I found, in any overbalanced gravity wheel with rotation rate, ω, axle to CG distance d, and CG dip angle φ, the average vertical velocity of its drive weights is downward and given by:
Vaver = -2(√2)πdωcosφ
Vaver = -2(√2)πdωcosφ
re: Vandusen Bessler wheel replication
Ken, you could screw up a wet dream!
Pins in slots will not act like rollers on a ramp. Rollers touch the ramps at different locations and different angles than the spoke. Rollers have inertial rotating energy to deal with. Rollers are free to move (bounce) off the ramp when conditions dictate whereas slots constrain such movement.
If a cake recipe called for sugar you would substitute salt because they both look the same to you.
Darrell's wheel does not have springs pushing weights inward. The spring concept variation was conceived and posted by Stefan at overunity.com.
You have your weights move a greater distance than Darrell's wheel.
You have opposing weights on the ramps simultaneously.
Your ramps are not located where Darrell has his ramps.
You don't have free fall where no weights are on any ramp.
Your wheel mass and weight masses are not the same as Darrell's
Quite plainly this is YOUR version (as you said) and not a valid test of Darrell's design. If you construct Darrell's wheel correctly (as I have done) using wm2d then it will conserve energy (except friction) and coast continually but without any excess energy output (when air and friction resistance is set to zero).
Pins in slots will not act like rollers on a ramp. Rollers touch the ramps at different locations and different angles than the spoke. Rollers have inertial rotating energy to deal with. Rollers are free to move (bounce) off the ramp when conditions dictate whereas slots constrain such movement.
If a cake recipe called for sugar you would substitute salt because they both look the same to you.
Darrell's wheel does not have springs pushing weights inward. The spring concept variation was conceived and posted by Stefan at overunity.com.
You have your weights move a greater distance than Darrell's wheel.
You have opposing weights on the ramps simultaneously.
Your ramps are not located where Darrell has his ramps.
You don't have free fall where no weights are on any ramp.
Your wheel mass and weight masses are not the same as Darrell's
Quite plainly this is YOUR version (as you said) and not a valid test of Darrell's design. If you construct Darrell's wheel correctly (as I have done) using wm2d then it will conserve energy (except friction) and coast continually but without any excess energy output (when air and friction resistance is set to zero).
-
- Enthusiast
- Posts: 72
- Joined: Mon Nov 17, 2003 11:38 pm
- Location: Berlin
- Contact:
re: Vandusen Bessler wheel replication
Jim Mich wrote:
this is right, Darrell´s machine seems to have only some springs
in his lock-unlock mechanism to help the fixing and unfixing
of the weights on the spokes, but the weight´s themself are not
spring supported to move more easily...
The spring idea is my own and I guess it would make the
machine more efficient or at least at all selfrunning,
as a weight, which has gone down in the gravity field
has LOST its potential energy and could not be pulled back without additional energy. But if you store the movement in a spring you
can then make the torquearm longer on one side ( heavier),
get the rotational energy from this heavier torquearm and accelerate the wheel with it
and then pull the weight back via the spring
and thus save additional energy to move the weight.
Maybe Darrell´s spring based lock-unlock mechanism helps this a bit,
although he just uses some ramps to shift the weights,
although the ramps seems to be special designed to do it efficiently.
Regards, Stefan.
Hi Jim,Darrell's wheel does not have springs pushing weights inward. The spring concept variation was conceived and posted by Stefan at overunity.com
this is right, Darrell´s machine seems to have only some springs
in his lock-unlock mechanism to help the fixing and unfixing
of the weights on the spokes, but the weight´s themself are not
spring supported to move more easily...
The spring idea is my own and I guess it would make the
machine more efficient or at least at all selfrunning,
as a weight, which has gone down in the gravity field
has LOST its potential energy and could not be pulled back without additional energy. But if you store the movement in a spring you
can then make the torquearm longer on one side ( heavier),
get the rotational energy from this heavier torquearm and accelerate the wheel with it
and then pull the weight back via the spring
and thus save additional energy to move the weight.
Maybe Darrell´s spring based lock-unlock mechanism helps this a bit,
although he just uses some ramps to shift the weights,
although the ramps seems to be special designed to do it efficiently.
Regards, Stefan.
-
- Enthusiast
- Posts: 72
- Joined: Mon Nov 17, 2003 11:38 pm
- Location: Berlin
- Contact:
re: Vandusen Bessler wheel replication
@Ken,
could you also please post your WM2D file, so I could have a look at it
next week, when I have more time.
Many thanks in advance.
Regards, Stefan.
could you also please post your WM2D file, so I could have a look at it
next week, when I have more time.
Many thanks in advance.
Regards, Stefan.
-
- Enthusiast
- Posts: 72
- Joined: Mon Nov 17, 2003 11:38 pm
- Location: Berlin
- Contact:
re: Vandusen Bessler wheel replication
Sterling D. Allan wrote:
I agree, it is time for Darrell to come forward and present a running
machine to you or any other visitor, as there are too many
people telling negative things about Darrell.
If he really has a machine running, it is time to come forward
and present it, otherwise he will have MISSED HIS CHANCE and most of the people will only remember him of having faked the videos ??
Hi Sterling,No working prototype to see
Today at 04:55:53 PM
--------------------------------------------------------------------------------
Darrell,
Several people have been in touch with me, who have had extensive dealings with you in the past.
The conclusion I glean from them, which you are free to refute with the validation proposal I submitted earlier today, is as follows:
1) You do not presently have a functioning prototype.
2) You have never had a prototype that made a full rotation, with acceleration between rotations.
When you do achieve a functioning prototype, let us know, and we would be glad to come validate your claim.
I've updated the feature page to reflect this:
http://peswiki.com/index.php/Directory: ... vity_Motor
| Sterling D. Allan
|
| New Energy Congress
| http://NewEnergyCongress.org
|
| PES Network, Inc, Executive Director
| http://PureEnergySystems.com
| http://FreeEnergyNews.com
| http://PESN.com
| http://PESWiki.com
|
| "The best news and directory service on the net
| regarding cutting edge energy technologies."
|
| newsletter: fe_updates-subscribe@yahoogroups.com
|
| home office: 1-801-407-1292
| Eagle Mountain, Utah, USA
I agree, it is time for Darrell to come forward and present a running
machine to you or any other visitor, as there are too many
people telling negative things about Darrell.
If he really has a machine running, it is time to come forward
and present it, otherwise he will have MISSED HIS CHANCE and most of the people will only remember him of having faked the videos ??
- LustInBlack
- Devotee
- Posts: 1964
- Joined: Thu Apr 06, 2006 10:30 am
re: Vandusen Bessler wheel replication
I asked him to sign the NDA, I am currently waiting for his answer..
I am still neutral about him, however I have to take into consideration what others say here.. But I still give him a chance..
I am very near ontario, I am ready to go there at anytime ...
I am still neutral about him, however I have to take into consideration what others say here.. But I still give him a chance..
I am very near ontario, I am ready to go there at anytime ...
re: Vandusen Bessler wheel replication
Ramp design may affect performance. I put a lot of research into different ramp profiles. Should they be a simple radii? If so should they be large or small radii? Or maybe some complex curve? One thing I found is the linear velocity of the weights must speed up during some part of the movement while on the ramps when the weights move from one radius to the other. Inertia will make the weights tend to continue a same linear speed.
Moving inward at the bottom ramp, the weight needs to start moving inward. To do so it must accalarate its linear speed else flex the spoke backward. Once moving inward it will need to eventualy slow down to match the inner radius speed.
Moving outward at the top ramp, the weight must again accelerate its linear speed else flex the spoke backward. Usually it must continue to accelerate until just before matching the outer radius where it will be traveling slightly faster. Then it slows down a bit to match the outer radius speed.
A while back I wrote a program that produced the attached graphs. In Calc01.gif I used 36 inch radius ramps both top and bottom. The three black horizontal lines in the middle represent the linear distance traveled by the weight during each one degree of wheel rotation. The red lines are as the weights move on the lower ramp and the blue the upper ramp. The wheel rotates CW while the graph moves from left to right. Calc02.gif is 16 inch upper radius and 22 inch lower radius while Calc03.gif is 22 inch upper radius and 16 inch lower.
One other comment. Once the wheel is turning centrifugal force will assist the upper ramp and hinder the lower one with the two balancing.
Stefan, adding springs doesn't help. If you add an inward pulling spring then you lift less weight at the bottom while lifting MORE at the top because you need to re-stretch the spring. No net gain!
Moving inward at the bottom ramp, the weight needs to start moving inward. To do so it must accalarate its linear speed else flex the spoke backward. Once moving inward it will need to eventualy slow down to match the inner radius speed.
Moving outward at the top ramp, the weight must again accelerate its linear speed else flex the spoke backward. Usually it must continue to accelerate until just before matching the outer radius where it will be traveling slightly faster. Then it slows down a bit to match the outer radius speed.
A while back I wrote a program that produced the attached graphs. In Calc01.gif I used 36 inch radius ramps both top and bottom. The three black horizontal lines in the middle represent the linear distance traveled by the weight during each one degree of wheel rotation. The red lines are as the weights move on the lower ramp and the blue the upper ramp. The wheel rotates CW while the graph moves from left to right. Calc02.gif is 16 inch upper radius and 22 inch lower radius while Calc03.gif is 22 inch upper radius and 16 inch lower.
One other comment. Once the wheel is turning centrifugal force will assist the upper ramp and hinder the lower one with the two balancing.
Stefan, adding springs doesn't help. If you add an inward pulling spring then you lift less weight at the bottom while lifting MORE at the top because you need to re-stretch the spring. No net gain!