Getting the ball rolling... / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Getting the ball rolling...

first previous 8 9 10 11 next

Move Thread

LAN_403

StevenCaton

Mon Jun 08 2009, 04:12PM

Registered Member #1845 Joined: Fri Dec 05 2008, 05:38AM
Location: California
Posts: 211

I matched them, assuming i had retained the initial state, and got an arc (but only from power from one transformer - the inactive transformer did not add to it, but neither did it cancel it out).

So are you saying that you cannot draw arcs from both transformers when they are plugged in. Plug the two in, (disconnect all other wiring that leads to the main gap, etc. ) and verify that you can draw an arc off each one. When they are both phased up, then the arc that you could draw from the bank should be much thicker and fatter. You did do this right?

Many of my connections were made by twisting copper wire together and then crimping it - i know it's not the safest thing ever, but it's been working pretty well so far - but it's just occurred to me that all that twisting might be generating a huge amount of inductance. Is this another thing to be tackled to unleash the actual potential of this thing?

EDIT: With only 30mA (one transformer) the recommended LTR cap is something between 17 and 18nF. I conservatively estimate my entire bank at 44nF. Could this disparity alone cause dramatic underperformance? Would an excessive cap size decrease performance even with both running in tandem (60mA demands ~34nF, so i would still be over) and should i try disconnecting a number of my bottles to boost performance?

This would not cause a decrease in performance. The only thing that could cause a decrease in performance would be that your cap bank was too small. Then you would not be utilizing the full energy that your transformers could charge those things to on each fire. (bang energy) (Someone correct me if this is bad advice)

Well, at this point several things come to mind that could be messing up your coil right now.
1. The NST bank is messed up, not phased properly, one transformer is dead, etc.
2. The adjustment of your electrode spacing on your spark gap could be limiting things. (experiment with this one) I have a suspicion that you might have the electrodes spaced way to far away.
3. Your toroid might be too big. (or small even, this one takes some experimenting usually)
A very large toroid will sometimes limit output. (same with small)

Oh, and by the way, congratulations on your first spark

, even though the length is suffering right now.

Herr Zapp

Mon Jun 08 2009, 07:16PM

Registered Member #480 Joined: Thu Jul 06 2006, 07:08PM
Location: North America
Posts: 644

Blitzorn -

If you are only getting 3" arcs, obviously something is VERY wrong. There's a good probability that your coil is either (a) just grossly out of tune, or (b) has a problem with the NST power source. However, working through a methodical, step-by-step analysis WILL determine what is wrong, and WILL get your coil working properly.

Here are my observations, generally following the sequence of your June 8 post:

1. First, ALWAYS start the tuning process with the coil configured as close to the calculated "tune point" as possible. Starting at one end of the available tuning range and slowing incrementing towards the other end is not good practice. Every time the spark gap fires, a large pulse of energy from the primary cap is dumped into the system. In a tuned coil, this energy is transferred into the secondary,and emerges as arcs & streamers. If the energy is not being transferred into the secondary, it has to go somewhere else, which may ultimately overstress the NSTs or the primary cap, or cause damage somewhere else.

2. Your arc target was a "flexible brass floor lamp". If this was a conventional floor lamp as sold in the US, they typically do not have a three-pin AC plug, and the lamp frame is NOT grounded to anything. Even 3" arcs contacting the lamp frame were causing insulation breakdown in the lamp wiring and finding their way directly into the AC hot or neutral lines: VERY bad practice!

ANY ARC TARGET MUST BE CONNECTED TO YOUR RF GROUND!!

3. Your observation about the "cold" NST is significant. Instructions were previously provided on how to individually check for HV output from EACH leg of EACH NST. Did you perform these tests prior to "phasing" the transformers? Each secondary output from each transformer should be capable of generating approximately the same length arc to the NST case.

4. There clearly is a question about correct phasing of the transformers. After verifying proper HV output from ALL FOUR transformer outputs, go back and CAREFULLY repeat the phasing process, making sure that both transformare are turned on, etc.

5. Are you using your variac for this initial testing? Are you sure that you have it connected up properly? Have you verified (with your voltmeter) that the variac delivers 0-120 volts to the NSTs UNDER LOAD?

6. You mention that you "estimate" your primary tank cap value to be .44nF.
How was this "estimate" obtained? Did you use a salt water bottle capacitor calculator, like the one at classictesla.com?

Unless you fairly accurately know your capacitor value, you can't be sure that have enough "tuning range" in your primary to achieve resonance. If you did not carefully calculate the capacitor value based on the physical dimensions of the bottles, please do so and report the value. Also describe the bottles themselves, their dimensions, and provide a detailed description of the bottle cap array, its wiring, the salinity of the electrolyte in the bottles, etc, etc.

7. You mention that most of your connections are made by "twisting" the wires together. This won't crerate extra inductance, but the primary circuit sees very high current pulses and simply twisting wires together is not good practice. Ideally, solder everything, or use carefully-crimped-on (or soldered on) ring terminals and screws & nuts to make sure that all connections are clean and tightly clamped.

8. How are the connections made to your primary coil?

9. How are the connections made to the magnet wire on your secondary? You did remove the insulating varnish so you had clean, bare copper exposed, right?

10. IF your transformers were improperly phased (or one was not even turned on), your spark length SHOULD improve as soon as you correct your power supply problems, assuming that everything else is correct. However, the coil still could be way out of tune, or have some other design defect or construction error that is limiting output.

11. In your post of 6-7, you mention that you are now using a single-segment air-blast spark gap similar to:

What are you using to provide airflow to the gap?
Are you using suction or pressure?
What is the actual distance between the spark gap electrodes?

Please review the above comments, and provide an update.

Regards,
Herr Zapp

Blitzorn

Tue Jun 09 2009, 02:25AM

Registered Member #2128 Joined: Thu May 21 2009, 03:17AM
Location: Seattle, Washington, USA
Posts: 58

SteveC -

Thanks much! tonight things should really get up to speed properly.
As far as top load goes, i have another smaller one i could rig up for later in the tuning process. Base = covered.

Herr Zapp -

Herr Zapp wrote ...

2. Your arc target was a "flexible brass floor lamp". If this was a conventional floor lamp as sold in the US, they typically do not have a three-pin AC plug, and the lamp frame is NOT grounded to anything. Even 3" arcs contacting the lamp frame were causing insulation breakdown in the lamp wiring and finding their way directly into the AC hot or neutral lines: VERY bad practice!

ANY ARC TARGET MUST BE CONNECTED TO YOUR RF GROUND!!

I may be a first timer, but i'm not a total nincompoop - the test was outdoors, the lamp was not plugged in, and the lamp's base was RF grounded (the brass plating runs from head to base). I think i'm alright on this one...

With regards to phasing - I have tested both transformers individually and KNOW they work. After i did this, however, i suspect one of them was inadvertently switched off. When i "phased" them and tested my phasing, i got a spark, so i assumed success. However if the phasing had been inverted with one transformer simply not running, there would be no efficacy in the phasing test. Had they both been switched on and powered up, there would not have been a spark. I was able to verify this theory by switching them one at a time, alternating left NST, right NST, and the pattern of result was spark, no spark, spark, no spark. Regardless of how they started out, if one was broken the pattern would be spark, spark, no spark, no spark. If both are working and phased, the only combination that yields no spark would be off/off, or 1 out of 4 possibilities. Since the cycle would be on/off, on/on, off/on, off/off, we can infer that there is no spark at off/off and on/on - a clear indicator of improper phasing. I am fairly confident that this is what happened and i will address it in a few minutes, it should be a quick fix if i'm right.

The variac is being used for testing and does, certainly, without a doubt, tested with a voltmeter and practical experimentation, deliver 0-120VAC. Not a problem.

I entered my best measurements into the classictesla calculator, and got more than one nF per bottle - far higher than any examples i had seen on any other page for the design i am using. Essentially everywhere i found a design like mine they had somewhere between 750 and 900pF per bottle. My bottles are brown glass, ~1/10 of an inch thick, with about five inches of usable cylindrical body. Each has about 300mL of nearly (although not quite) saturated brine in it, upon which rests a tablespoon of motor oil. They are wrapped in 2 layers of grilling aluminum foil held on with a ring of electrical tape around the neck. Each has a 12 AWG copper wire inserted through the lid, sunk down into the brine, and secured with hot glue. The lids are snapped on and wrapped thoroughly in electrical tape. The bank is arranged four bottles wide and twelve bottles long in an open box made of particle board. Lining the bottom of the box is a bed of aluminum foil that comes some distance up the sides. On this rests a narrow sheet of industrial copper that fits the box like a glove. The copper sheet feeds up and over the side of the box opposite the inputs, and it joined to a wire that serves as the negative side of the circuit. The bottles are held together in the center of the box by crumpled and rolled foil padding that is snug around them and between them and contacts the copper sheet along with their bases. The input sides are connected in parallel and divided into four rows that meet at the input end.

Note: there are two empty cardboard boxes at one end as space filler.

I know my caps aren't total crap, but i also find it hard to believe they are as good as classictesla wants me to believe.

I know that just twisting them is not good practice, and it's really just a stand-in until i have the time and the baseline tuning to swap in parts that are more permanent, safer, cleaner, and prettier. The summer will see a lot of tinkering on this baby for optimization and cosmetic repair, but also a lot of permanent fixtures will be added to replace my hasty jury-rigging.

The inside end of the primary coil is simply soldered onto some 12AWG wire that is wrapped around it, and the tap point is a fuse-holder.

Both ends of the secondary were thoroughly sanded and bare when they were connected. The grounding end is simply locked in a copper grounding lug, the top load end is woven around a ring terminal and secured with a machine screw, this terminal attaches it to a short length of heavier gauge copper wire that connects it to the toroid.

My spark gap is being ventilated by suction from a 1.5hp shopvac. I have run it with gaps between 1 and 5mm, the five sparked but it did so erratically, slowly, and unusually loudly, so i cut the power almost immediately. I didn't realize they were as far apart as they were. I don't have a set distance right now, but i think for tonight's tests i'll stick with about 1.75mm unless there's some outcry...

I think we can safely assume that the coil IS grossly out of tune, but i sense that that isn't the ONLY problem it's got going for it. If we're lucky, just fixing the NST phasing and adjusting the primary tap will do the trick.

More later! Wish me luck!

Herr Zapp

Tue Jun 09 2009, 06:27AM

Registered Member #480 Joined: Thu Jul 06 2006, 07:08PM
Location: North America
Posts: 644

Blitzorn -

It sounds like you are getting closer to success ...... let us know how the coil performs after you "fix" the NST phasing. (Make only one change at a time!)

One more question on your bottle caps ..... how far up the sidewall of the bottle does the foil extend? You stated that the foil is secured with electrical tape at the neck of the bottle, which almost makes it sound like the entire bottle is foil covered.

Generally, you want the foil to stop at or just slightly past the end of the cylindrical portion of the bottle. You want at least 2-3" of exposed glass between the end of the foil and the bottle cap. If you have this entire area covered with electrical tape, you may be having electrical "tracking" (leakage) under the tape, especially if there is any chance of salt residue on the glass. If there is 2-3" of exposed glass, you can carefully wipe it down to remove any salt residue, salty fingerprints, etc.

Regards,
Herr Zapp

Blitzorn

Tue Jun 09 2009, 06:51AM

Registered Member #2128 Joined: Thu May 21 2009, 03:17AM
Location: Seattle, Washington, USA
Posts: 58

Tonight's testing is complete. The news is good!

My theory about the transformers was correct, it only took a little tweaking to get the proper phased output.

I'm not going to go into the gory details of the tuning, but after adjusting the tap point and the spark gap a bit, we coaxed out some much more satisfying discharges... some upwards of ten inches, and the thing actually behaves the way i expected now, although still not peak performance. What i have now will work more than great for the demo.

Herr Zapp, when i said secured at the neck i meant where the cylindrical portion meets the neck. I linked to a photo of the cap bank, if you missed that.
Here are a couple (amateur) photos of the thing running:

All seems to be going according to plan. Thanks again guys! It's Blitz!

CT2

Tue Jun 09 2009, 02:05PM

Registered Member #180 Joined: Thu Feb 16 2006, 02:12AM
Location: Ontario, Canada
Posts: 187

On a little side note, and I know you said you had you lamp not plugged in and properly grounded. But I remember one time when I was young and just starting out I had built a little coil in my shed, and i had a brass floor lamp in there probably similar to yours. I wasn't letting the coil arc to it or anything like that but I guess it picked up enough voltage from the nearby coil to arc through the insulation on the power wires running up to the bulb, because when I grabbed it (coil off) to move it out of the way, I had a funny sensation in my hand and arm, that was weird... grabbed it again, yep I'm getting electrocuted, thats the end of that lamp, and I'm just glad I didn't get a serious shock.

So just watch out, and congrats on getting the coil going! Feels great when you finally get some sparks eh?

Herr Zapp

Wed Jun 10 2009, 12:22AM

Registered Member #480 Joined: Thu Jul 06 2006, 07:08PM
Location: North America
Posts: 644

Blitzorn -

Good to hear that the coil is running better. Seeing the streamers, hearing the popping and snapping, smelling the ozone ........ all part of the addiction.

Still, there's a LOT more performance to be squeezed out this system. The well-accepted Freau formula predicts that your coil (when fully optimized), could generate a maximum arc length of 36 inches.

However, further performance improvements will get progressively more and more difficult to obtain. You will want to look at refinements to tuning, optimizing pri-sec coupling, increasing topload size, minimizing resistance in the primary circuit, optimizing spark gap performance (best performance may NOT be with the shop vac at full power), boosting NST input voltage above 120V, assembling a lower-loss capacitor (MMC), etc, etc.

(If all the wiring to the salt-water caps is #12 AWG copper, I'd just sit down for 15 minutes with a propane torch, some plumbing solder and flux and carefully solder all the loosely twisted interconnects in the primary circuit.)

Regards,
Herr Zapp

Blitzorn

Wed Jun 10 2009, 05:41AM

Registered Member #2128 Joined: Thu May 21 2009, 03:17AM
Location: Seattle, Washington, USA
Posts: 58

CT2 - The lamp may be fried - but it's an ugly old lamp anyhow, and i think it's just going to be my spark guinea pig for the rest of its life. Concerns duly noted though. And yes, i am THRILLED.

I call it Zeus Mark I

Herr Zapp - While we were running th demos today i actually smelled the ozone for the first time, i recognized it from repairing swimming pool chlorination and filtering machinery. Good times.

In a couple of weeks, when school is off my plate, i'll be back to tinkering, and queries will be addressed in earnest then.

But is it really advisable to step up the input voltage over 120VAC on the NST's? That would not be difficult, my variac goes up to 135VAC, but it just seems cruel to the circuit (and all my careful math >.>)

What kind of effect would i get from moving the primary coil higher? (it's a bit below the secondary, actually...)

Herr Zapp

Wed Jun 10 2009, 06:16AM

Registered Member #480 Joined: Thu Jul 06 2006, 07:08PM
Location: North America
Posts: 644

Blitzhorn -

Boosting NST input voltage to 135V will only very slightly increase the "risk" to your 7.5kV NSTs. They are probably assembled with the same insulation materials & design as used on 10kV, 12kV and 15kV NST's, so they should have adequate safety margin in the secondary insulation..

The slight increase in NST output voltage won't have any significant effect on any of your "math", at least not on anything that would affect tuning.

The coefficient of magnetic coupling between the primary and secondary coils is defined as the "K" factor. As "K" increases, coil output will also increase (to a point), but spark gap quenching will become more critical. Typical spark gap Tesla coils will have "K" factors between .1 and .22.

Use JAVATC or MANDK to calculate the "K" factor of your coil as currently assembled, and to determine what changes would be required to increase "K" to .15 or .18.

Regards,
Herr Zapp

Blitzorn

Wed Jun 10 2009, 05:52PM

Registered Member #2128 Joined: Thu May 21 2009, 03:17AM
Location: Seattle, Washington, USA
Posts: 58

As far as spark length goes, would adjusting the variac up really make a significant difference? is there a way to calculate the change in output wattage if i change the input voltage?

I find that a wider spark gap delivers significantly longer, more powerful sparking, but at a point they start becoming erratic and then fail entirely as the spark gap refuses to arc. Even if it runs in a satisfying and attractive way, does having a spark gap that is close the the failure point dangerously overstress the system? I am aware that it requires the caps to handle more charge, the spark gap to endure more powerful sparks, and the cold side of the transformer to process much larger pulses of power than if the spark gap is narrow, but what is the risk of damage?

I'm also curious how less ventilation could improve performance... would it permit the sparks to persist longer? If it does help, would it be worth the increased breakdown of the electrodes?

first previous 8 9 10 11 next

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