Giant Wimshurst Machine - could use some guidance / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Giant Wimshurst Machine - could use some guidance

first previous 15 16 17 18 next

Move Thread

LAN_403

AwesomeMatt

Sat Jun 28 2014, 01:59AM

Registered Member #4454 Joined: Sun Feb 26 2012, 12:47AM
Location: Western Canada
Posts: 74

wrote ...
You could probably salvage some very thin and springy flat metal from an old wind-up watch or an analog movement meter.

Hrm... Lemme check my bucket...

2002

Not a single damn spring.

When I was a kid, MacDonalds had those yank-back-release spring powered toys. A trip to a thrift store should satisfy that.

Also, if resistance is too hight, I'll just add more and longer brushes. Right now there's only 2 brushes per... brush. If I slap on a run of 8 instead... will be 1/4 the restistance.

Went to a carbon fiber place today, they sell in amounts of 500lbs only... yet... have no samples on hand. so, I can't sneak a sample and use that, nor could I order a small amount. Foey.

wrote ...
If the machine works only at high speed the most probable problem is leakage.

Interesting. Okay.

wrote ...
Verify if the machine works without the charge collectors. With only the neutralizers in place it shall charge the disks very evidently.

That is the first thing I tried yesterday. I used to get static and hiss before I'd even given the handle a full crank, before I even built the combs. Same deal, absolutely nothing until I get it going really fast.

wrote ...
I like to clean these machines with furniture cleaning products containing silicone and wax, instead of with alcohol.

Why? Would any residue of the cleaning product be considered "dirt" or would it be considered beneficial insulator?

...

So, the biggest sparks I could get this time were maybe 5", with the leyden jars hooked up. I was sweating up a storm trying to keep it powered at that speed.

But, woohdamn were they loud. And the farther apart I had the electrodes, the more loud they were. Disproportionately. Naturally... voltage required to jump an airgap is linear with the gap. But, energy stored in a capacitor is proportional to voltage squared... so... doubling the gap is 4x the energy in the snap.

My camera didn't pick them up very well this time. Overwhelmed the sensor I think. It's showing 24 long arcs across the middle of the conductor poles. Maybe just lens artifacts. Kind of a crappy deal, since I hardly spent any time watching it with my own eyes.

I was told the prototype has served every purpose it could, and I'm done working on it. Pack it up, head home. So, other than the crappy video I got today, maybe I won't get any more, perhaps for weeks or months.

They're suggesting that if the fabricator has questions, that they hire me as a consultant, and, if necessary for me to tweak or experiment with the prototype more, to come back and do that only if they need me to.

It's Friday so people were ready to get out of there. We had to stop anyway when the mis-sized crank finally bent its bolt and snapped the arm.

Well, the demo stopped. I was, umm, just cleaning up and went back for more. No more crank in the way, I hooked up the cordless drill. I umm, maybe let some of the smoke out of the first one, but the second one made it spin up nicely. I think I got somewhere around 10-14" sparks. I could see the flashes through the brown paper on the discs :p. I had to spin the wheel increasingly fast to accomplish anything at higher voltages. Not an issue of "Just store more energy" but, seemed "Must overcome increasing leakage to have any left over to store." Interestingly, if I could get to a certain voltage at all... I could get regular sparks there. Like, easily every 3-5 seconds, but if I didn't get one for maybe 10-12 seconds, I never would.

I had the earmuffs on and it was loud. At 5", they sounded like gunshots and made me jump the first time I heard it when someone else was cranking. At 10+"... I swear I could feel the sonic shockwave.

At some point before I could see if the Leyden jars were going to puncture, the crossed belt ripped itself apart and today spun to a close.

Gimme a few hours or a day to compose the video and I'll post it.

Antonio

Sat Jun 28 2014, 03:39PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

Carbon fiber brushes can be problematic. If they lose filaments they short-circuit the machine. There are always sparks at the neutralizers, and the brushes may burn and leave carbon dust, again short-circuiting the machine. And they always cause startup difficulties.
The best brushes are metallic, built in a way that touches the disks lightly, but with good contact.
Alcohol may dissolve the glue of the sectors, and the glue residue left may be not a good insulator, leaving the disks contaminated. It may also cause cracks in acrylic. I have seen a complicated acrylic part with several holes and threads that I tried to clean with alcohol after finishing it crack in several pieces.
I see that you know now the problems with belts and cranks...

AwesomeMatt

Mon Jun 30 2014, 09:48AM

Registered Member #4454 Joined: Sun Feb 26 2012, 12:47AM
Location: Western Canada
Posts: 74

Here's the (private) vid:

Or for those just looking for a picture, here's the old sparks versus the new ones:

2058

BLAM BLAM BLAM.

I don't know that I"ll get a chance to work on it again, but, I'm about half way there, spark wise. Which means I'm about 25% there energy-wise. Final sparks are going to be lotsa fun.

The discharge right now is insane. The lightning bolt is already 2" thick at the one side.

The spark gap is still to narrow to be interesting, since almost every spark is just a straight line.

It's adding weird artifacts on video, little pre-arcs that are aligned left-right okay, but up-down wrong. I cut most of them, but they most often appear the frame before the blue spark (which often doesn't appear at all). I think they're some kind of lens effect, or, they're some residual image from the first feelers before the full arc.

A long odds thing will be to try to schedule some free slow-motion camera time to see what's up.

Antonio

Mon Jun 30 2014, 01:43PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

The spark images at wrong places appear only in digital cameras. Probably some interference on the circuits of the image sensors. You didn't take off the backing paper of the disks, that is probably conductive and leaves the machine partially short-circuited.

teravolt

Mon Jun 30 2014, 04:23PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

are you using the pvc Lydon's and do you have a picture

AwesomeMatt

Tue Jul 01 2014, 01:40AM

Registered Member #4454 Joined: Sun Feb 26 2012, 12:47AM
Location: Western Canada
Posts: 74

wrote ...
You didn't take off the backing paper of the disks, that is probably conductive and leaves the machine partially short-circuited.

You're the 3rd person that's suggested to that as the likely fault. I concur. If I ever get to work on it again, I'll get a hand taking it apart and stripping the discs properly.

wrote ...
are you using the pvc Lydon's and do you have a picture

Yes, am still using the 4" PVC pipes, 42" long. 18" gap at the top, 4" gap at the bottom, 20" of aluminum tape.

Pictures.. no. But you can see them right at the end of the video. They are vertical and next to the 2" PVC output supports.

Why do you ask?

.

Question barrage. Any answers or feedback would be very helpful and documented in a tutorial.

1 - Instead of foil, would there be a point to using plates? I was thinking of stainless plates. Durability forever. Any electrostatic concerns?

2 - Would there be a significant reduction in leakage by rounding the edges of the stainless plates?

3 - Would it be helpful to CNC a small trough out of the acrylic discs for each of the sector plates to sit in?

4 - For LOTS of use, suppose it's spinning almost all day every day... any idea on using a protective (possibly disposable) film like Antonio did on his, with holes cut for buttons? I'm worried about the mechanical aspects and rubbing on the edges of the much softer adhesive surface covering.

5 - I still don't understand how a Wimshurst works. I have a fair knowledge of electrical things, ohms law, power law, that kind of hand-on level stuff, but I have very poor grasp of the actual physics. Could anyone walk me through what is actually happening, where the charge comes from, circuit-wise what the circuit is, etc? I know that at some point things will "click" and then all the little separate pieces I know but can't integrate together will fall into place.

teravolt

Tue Jul 01 2014, 02:21AM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

With van de graphs the more capacitance the more voltage but that doesn't seem so in a wimshurst just more current. I was just curious how much voltage a peace of pvc tubing will hold off. Shouldn't you be getting arcs that are about 18"? when I built my van de graph my biggest enemy was charges that bled off even the type of rubber the belt was and the way I joined it is a big deal. I think that your wimshurst is very nice and well done.

Antonio

Wed Jul 02 2014, 12:06AM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

AwesomeMatt wrote ...

Question barrage. Any answers or feedback would be very helpful and documented in a tutorial.

1 - Durability is improved, more difficult to cut, can get loose and fly away, brushes break more easily. I used aluminum plates in this machine and saw all this:

2 - Yes.

3 - The brushes will like this, but this goes agains the idea of letting the brushes touch only the sectors, and not the spaces between them.

4 - Insulated sectors increase the performance significantly, because the sectors can be larger (more current) and the insulation is better (more voltage). I didn't think about the added protection, because it's difficult to apply the covers, and I don't have a good solution for buttons that can be easily removed. The material that I used appears to be quite difficult to scratch.

5 - The machine works due to concentrated electric field when a sector touches a neutralizer brush. Several sectors at the same time attract opposite charge to the touched sector, resulting in more charge on it that was present in the other sectors around it. I have a more lengthy discussion of this and simulations here:

Dr. Slack

Wed Jul 02 2014, 08:21AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

It 'clicked' for me after I asked the question about the angle of the neutraliser bars.

1) Look up electrophorus

This was one of the earliest electrical machines to multiply charge, and is the simplest place to start. The Wimshurst works on exactly the same principle.

2) What happens in the electrophorus is that you have a source of charge (a). This doesn't get used up during operation, so it performs over and over. You bring a conductive body (b) close to the source. The field from (a) polarises the conductor so there is charge separation between the closest and furthest bits from the source. Another way to look at it is that as the conductive body is conductive, it distorts the field of (a). This means that it is internally charged, and the field from that charge is neutralising the field from (a). Now you ground the point on (b) furthest from (a). Charge flows to/from the excess on that side to ground. Another way to look at that is the conductive body (b) has just got bigger by adding the conductive ground contact to it, which extends further away from (a). This allows the charge to separate further, and in doing this, some charge flows between ground and (b). Remove the ground contact, to stop charge moving onto/off (b), and pull it away from (a). The remaining net charge on (b) from the polarise-ground-unpolarise process redistributes itself all over the surface. As you further pull (b) away from (a) and other grounds, its capacitance to ground falls, and so with constant charge, the voltage rises.

3) A Whimshurt automates this process.

4) Let's assume that the bottom sectors on the *other* disc are charged (just go with me on this). These are (a). Now consider a sector near the bottom of *this* disc. It will be polarised by the charges on the other disc. As the disc turns, (b) passes under and contacts the grounded neutraliser brush, which allows charge to flow. Shortly afterwards, contact with the neutraliser is broken, and the disc carries the now charged sector across the midline of the machine, where it now acts as (a) to the other disc's (b)s when they get to their neutraliser bars. Finally, it is carried out to the edge of the machine, reducing capacitance from the various grounds so increasing the voltage. The high voltage sectors then dump their charge into the collecting combs.

5) So how does it start? No two bits of metal insulated from each other will have the same voltage, even if you don't walk across a carpet and then touch a segment. So a Wimshurst machine will *always* have a charge imbalance somewhere on it to start it. Because the brushes and the segments are conductive, the slightest voltage difference will cause charges to flow, that will then self-start the machine. This is in contrast to a Bonnetti machine, which is basically a Wimshurst with no segments. No current can flow here until the voltage is high enough to cause corona, so the neutraliser combs can spray charge onto the disc. A Bonnetti is typically started by charging a glass rod with silk or fur or something, then holding it opposite a neutraliser comb, to get charge spraying.

Other electrostatic machines.

The Kelvin Water-drop machine uses electrophorus charging. Subsequent separation uses gravity.

If you made a Van de Graff belt with alternate metal/plastic chain links, then it's called a pelletron, and you could charge it electrophorus style, but even those are more usually charged by direct spraying

Now this is the sort of post that ought to go into the, as you correctly point out, moribund Wiki.

kilovolt

Mon Jul 28 2014, 02:41PM

Registered Member #2018 Joined: Tue Mar 10 2009, 09:56AM
Location: Switzerland
Posts: 74

AwesomeMatt, how is the actual status of this project now? Did you perform some improvements futher or is it discontinued?

Very best regards
hvpower

first previous 15 16 17 18 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob