Author Topic: Doorknob vs Ceramic Disk  (Read 811 times)

Offline Andrew321

  • High Voltage Enthusiast
  • *
  • Posts: 17
  • Karma: +0/-0
    • View Profile
Doorknob vs Ceramic Disk
« on: January 23, 2020, 03:57:09 AM »
I'm getting ready to build a X10 multiplier and I see online that all the more serious highvoltage multipliers use doorknob capacitors instead of ceramic disk capacitors. Doorknob capacitors are so much more expensive so I assume there is a good reason people keep using them, but I just don't know why. Can anyone shed some light on this for me?

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 192
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #1 on: January 23, 2020, 06:06:20 AM »
Andrew,

There are also other options, polystyrene or other plastic-film dielectrics.

Door-knob capacitors are expensive, but ceramic disk capacitors often have several issues.  I use them, but with care.  Personally, I go to film capacitors for better performance rather than door-knobs.

One issue is the capacitance vs. voltage curve.  Cheap ceramic capacitors can loose 90% of their capacitance at rated DC voltage, especially larger capacitance values.  A second issue is loss when charging and discharging.  My initial voltage multiplier (static-electricity generator) uses 30kV ceramic disk capacitors.  However, when initially driving the input at 20kHz, some of the capacitors overheated and failed due to their internal dielectric losses.  (Even w/o DC load on the multiplier, stray capacitance can cause quite a current at 20kHz.)  Since I didn't need much DC current, dropped the input frequency to 700Hz, and it works fine.  Finally, the closer lead proximity makes potting or oil immersion more critical.
David Knierim

Offline kilovolt

  • High Voltage Enthusiast
  • *
  • Posts: 13
  • Karma: +1/-0
    • View Profile
    • Hochspannung als Hobby
Re: Doorknob vs Ceramic Disk
« Reply #2 on: January 23, 2020, 05:01:27 PM »
I have worked successfully with both capacitor types. But the fact is that doorknobs can bear much higher currents than any ceramic filter capacitor. Especially in the lower stages there are sometimes massive current peaks, which can cause cheap capacitors to burst.

But doorknobs have another advantage: With the screw connections the construction is more stable and better, especially if you want to add toroids or balls for field shaping.

This is my latest bipolar Setup with doorknobs, it puts out more than 70cm arcs:


If you need some ideas, have a look at my website. I presented three cascades, one with 100kV, one with 180kV and a bipolar with up to 700kV open circuit voltage:
http://kilovolt.ch/kaskade.htm
http://kilovolt.ch/kaskade_180kv.htm
http://kilovolt.ch/kaskade_bipolar.htm

Best regards
kilovolt
« Last Edit: January 23, 2020, 05:12:30 PM by kilovolt »
All information on my part without guarantee! I reject any liability for personal and/or material damage. Everyone is responsible for his own safety.

Offline MRMILSTAR

  • High Voltage Technician
  • ***
  • Posts: 154
  • Karma: +4/-0
  • Cedar Rapids, Iowa
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #3 on: January 23, 2020, 08:39:10 PM »
That is a very impressive looking setup kilovolt. One question I have is about structural integrity. It doesn't appear that anything really holding it up. It also appears that you aren't using resistors on the output terminal. Does this concern you?
Steve White
Cedar Rapids, Iowa
Retired electrical engineer

Offline kilovolt

  • High Voltage Enthusiast
  • *
  • Posts: 13
  • Karma: +1/-0
    • View Profile
    • Hochspannung als Hobby
Re: Doorknob vs Ceramic Disk
« Reply #4 on: January 23, 2020, 09:55:05 PM »
Hello Steve

I'm definitely using many resistors on the output of both multiplier towers, but these resistors can't be seen in the picture above. On the following picture you can see the whole setup, the resistors can be seen in the foreground:



"One question I have is about structural integrity. It doesn't appear that anything really holding it up."
What do you mean? Every cap is holding each other by a grub screw, which is going through the toroid between two caps. Same at the bottom on the acrylglass.

Best regards
kilovolt


« Last Edit: January 23, 2020, 10:09:34 PM by kilovolt »
All information on my part without guarantee! I reject any liability for personal and/or material damage. Everyone is responsible for his own safety.

Offline MRMILSTAR

  • High Voltage Technician
  • ***
  • Posts: 154
  • Karma: +4/-0
  • Cedar Rapids, Iowa
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #5 on: January 23, 2020, 10:42:22 PM »
Good to see the resistors. How much resistance are you using?

I understand that the caps are all bolted together. It just looks like the towers might be a little wobbly. Maybe the picture is misleading and its very sturdy. My observation wasn't intended as a criticism. Its a very impressive project.
« Last Edit: January 23, 2020, 10:47:06 PM by MRMILSTAR »
Steve White
Cedar Rapids, Iowa
Retired electrical engineer

Offline kilovolt

  • High Voltage Enthusiast
  • *
  • Posts: 13
  • Karma: +1/-0
    • View Profile
    • Hochspannung als Hobby
Re: Doorknob vs Ceramic Disk
« Reply #6 on: January 24, 2020, 09:03:52 AM »
Hello Steve

No problem, I didn't take that as a criticism. Apart from that, criticism is also very welcome, because it is helpful for improvements.

The upper capacitors are partly mounted a bit crooked. It is due to inaccuracies of the balls.

The resistor chain consists mainly of 10kOhm / 11W resistors and contains about 800kOhm. In principle you only have to make sure that the peak currents of the capacitors cannot destroy the diodes, so it is sufficient to dimension the resistors in a way that the surge currents of the diodes are never exceeded.

Best regards
kilovolt
All information on my part without guarantee! I reject any liability for personal and/or material damage. Everyone is responsible for his own safety.

Offline Andrew321

  • High Voltage Enthusiast
  • *
  • Posts: 17
  • Karma: +0/-0
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #7 on: February 08, 2020, 03:27:06 PM »
Good to know they don't hold up too well. Perfect for testing and learning because they are cheap, but I will be using something better in the end.

A few questions about voltage multiplies in general that I'm a bit confused about. When discharging the Cockcroft–Walton multiplier I hear that there are very large surge currents that can destroy the diodes and so resistors are needed. I didn't think that the discharge takes place through the diodes though, but maybe I'm wrong? Or is it that once discharged the inrush current that charges them is too high?

When it comes to full wave multipliers, you have the outer capacitors feeding bridge rectifiers which charge the central capacitors in parallel. In this, the capacitors discharge in series from one end to the other and no discharge current flows though the diodes, correct?

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 192
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #8 on: February 08, 2020, 06:55:44 PM »
Both (or all three in the "full wave" case) of the capacitor columns are charged to the same high voltage.  One is static, in the sense that the bottom end is tied to ground and the top end is the multiplier output.  The other is fed with AC at the bottom.  The DC voltage across each capacitor in each column is roughly the same.  (Except the bottom capacitor of the AC-fed column has half-voltage).  When the output arcs to ground, all the capacitors get discharged.  The output column discharges directly.  The AC-fed column(s) discharge through the diodes.

Another way to look at it:  Consider a 10kVpp AC input to a 10-stage multiplier.  19 of the capacitors have 10kV DC across them.  (The bottom AC-fed capacitor has only 5kV, as it's half-wave rectified version of the input AC voltage.)  So, the output is 100kV DC.  The top of the AC-fed capacitor column is oscillating between 90kV and 100kV.  If the output is discharged from 100kV to 0V, the top of the other column can't remain at 90-100kV.  The diodes conduct to pull it down too.

CW multipliers are ideal for providing DC voltage.  For making arcs, the necessary resistors lower efficiency.  Marx generators are another option for high-voltage DC arcs, but they have their own set of efficiency losses.
David Knierim

Offline Andrew321

  • High Voltage Enthusiast
  • *
  • Posts: 17
  • Karma: +0/-0
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #9 on: February 09, 2020, 05:20:17 PM »
Ahhhh I think I get it. So then what kills the diodes is the surge current from the AC feed capacitors through the diodes, not the main discharge from the static discharge capacitors. So another way to limit the surge current thought the diodes would be to use smaller feed capacitors and then larger discharge capacitors. It would take a bit longer for the static capacitors to charge up but would reduce the need for big resistors.

I built a Marx generator once and it worked ok, but I found it a bit loud with all the sparks going off all the time. I also found that the capacitors seemed to discharge between themselves with out the main arc actually going off. I didn't have a lot of experience back then and I'm sure I could build a better one now. I plan to revisit them after I conquer this ZVS and multiplier I'm working on. I'm actually going to use the voltage multiplier and ZVS driver to power the marx generator and I'll build my own capacitors for it to keep the coast down.

I've been looking at diodes to buy and I almost settled on these https://www.mouser.ca/ProductDetail/583-R5000F-T
Seems like they should be good but looking at the data sheet has me worried that they might not be designed for high frequency.

Also, when building a voltage multiplier, I should only be concerned with the reverse voltage DC value, not the AC, right?

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 192
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #10 on: February 10, 2020, 05:26:26 AM »
Andrew,

Glad it's making sense!  Yes, smaller capacitors on the AC-fed column will allow lower output resistance.  The output current splits by the capacitor ratio.  If the DC output column capacitors are 10 times larger, then they'll supply 10/11ths of the discharge current and the AC-fed column will supply 1/11th (9%).  As you mentioned, it will take more cycles to charge that way.

Those diodes are relative slow.   For CW multiplier use, the caps need to feed the reverse-recovery charge of all the diodes.  So, I'd use these for at most a few kHz.  They do have a nice large surge-current capability - 30A for one 60Hz cycle.

I've found it difficult to find fast HV diodes in normal electronics parts distribution.  I've gone to Chinese parts from EBay, and had good luck so far.  Depending on the voltage you need, here's some links for 30mA at 15, 20, 25, and 30kV:
https://www.ebay.com/itm/HVRT150-30mA-15KV-100nS-HV-Rectifier-High-Voltage-Diode-20pcs/401463844295?hash=item5d791c1dc7:g:008AAOSw0W5aO87s
https://www.ebay.com/itm/10pcs-2CL20KV-30mA-High-Voltage-Diode-Rectifier-HF-Tesla-Coil/382415572940?_trkparms=ispr%3D1&hash=item5909be8fcc:g:JhcAAOSw76JcMyne&enc=AQAEAAACQBPxNw%2BVj6nta7CKEs3N0qVIA%2Fl%2BWrTfyJdvQbYaSA2PBSlY61x5eAYWC3QVGs6DO4Rxigadgx9P7cC56%2FLhFkz%2BVNddX75pW%2B%2FBD5%2FMgw%2FtLtaNEXXe9zJS205l7oupX1ZPywMBo%2BENdoHueQA4x2whiqCI41WIdcCuAOAsZPNpCMK8FOHVpE2ioFhTbLiqyzIQmMhw1txTN9AiBimqf%2BCVNt93TRf4MMZ95I0O1vXH2YfnTHTeZDvZOeEQuSp0lvY3NAmffmGlwgN89pogxre5HV0iIOciidPsaiivq8gZ4PbD%2B%2FbJ0wSTeMcrbu%2F2NORy3y5R0J%2B%2Brc0jfvQGNXN0tUKgjKF0ZTbOJXdrp%2BQni7gzDHbZwjN2Ps7om3zJj4ra2QK%2FSMPUSdbhRpyWjRCJrJjMvvzkyL8Q8oP6sPtbqD4iOQ02OJC3PYfiTD7GQv2e2Ci%2BuPeTaeMczvrkUyjEIMC4Mt1QBHCAmJjlPjlmsECaB%2FZCrOwVmvTKmUn5aw7Zdhl08Wvmh%2F83QiSlatYwHrS%2BPcSohsfRa3YhYwyBg4dNecwMIX8eDwtdLKU2Hj%2F6mbcCystoDuz%2B9GuqBKaBZCLqhm1fjr%2FAVMDyTxt1FQqcip32YRiM9Q%2Fk%2FSJm0T%2BkfupeiM8kHWQi8RVo%2F9B8IEVIpH0MRI2ENBqkDwwdfcmOMnXhjvhAATgDUhBpg%2BPBIHLpUer5%2FAzPOCkybzbKKbDBDlxyPqFCBAczp%2BRm19RfLhVf84Uhk62qzgtLqQ%3D%3D&checksum=382415572940a42d186c494248ed8d1f8803446e0d3b
https://www.ebay.com/itm/US-Stock-12x-2CL25KV-25KV-30mA-100nS-High-Voltage-Diode-Rectifier-HF-Tesla-Coil/401295878181?_trkparms=ispr%3D1&hash=item5d6f192825:g:L6oAAOSwVrBalm5t:sc:ShippingMethodStandard!97070!US!-1&enc=AQAEAAACQBPxNw%2BVj6nta7CKEs3N0qWj8RhVGMcqidy%2FrYqjIDyalLsvkMdDJZC82HCKjYL3isishWWhioPJhD22hZVdiPy099u22UhUvMwamHIV%2BSKzy6v%2BxFxQMD%2FuNkhaKBIGOE1PRQUfA0CvOfa%2FNYaXwmoAm6kscbukmw1phGpa6It8ARS59Cl4devhi3SopK%2F4z%2FP442lUA7v1Lbg4abvyo%2FE3Pq6HTLKzH0%2FDonRaVOGyQTFb5X8MhytsP0lc%2FR8%2BQF10QzHJqRcOz7PcZf9H%2F82HiwUgedb1jKUGckkBHlXtmpjpTQNo3d1rBPudd01zD2t5pMgnJUbQ47ApIKAcBQXd5PHrOnXQ6HJ9Mslb1b6DCePS2quVjQeAvFjuhGa0W%2F60nlpygVPCBRTWjURR3heuuC05hECWVRxZ2x6d7X1OMVSGXLPmP1ZF2a01DjNSrmCJjr%2F9z6u9%2FcJxxHXZEwvfQekz%2FP21IpMfWqWxj2tYLI2sDDn5CpoJwSCVq3F5FaKqK6MP1clhuOU4c6HQIfKKhaY%2F3LxrKiHbFAfdEnu9Vz5zn3Ox6WVEt2IQXxqwE%2FfH0ZpoE8efHvHK6SolMwWkwR%2B1qZABQC0HoiMn5W0rTEztgPn70o0Fs3%2B%2Bh%2BuB0DPsPAgVPSjPHgKKSMiJid0SswzvBBGQO8YwqrcLaziucs4Otsfvag75kOXwJA7UUBt36%2BGaKT0DlzkrecN6c9IuuHrJVcfQOC3SwkPNjOhqHMLEpn%2Feu0AH5lBGx68XJQ%3D%3D&checksum=401295878181c7ef6e17669144b58fb84364e997aaf3
https://www.ebay.com/itm/6pcs-30mA-30KV-100nS-High-Voltage-RECTIFIER-Diodes/153434145388?hash=item23b963426c:g:tnIAAOSw-FpbL7mC

5mA versions are much more available, but that may not be enough for your plans.

For the diodes, reverse DC voltage is easiest to understand, but AC voltage is just sqrt(2) lower (presuming full-wave rectification).  If you are referring to the capacitors, they do experience some AC voltage (due to the AC current), which can be the cause of failure for cheap disc ceramic caps.
David Knierim

Offline Andrew321

  • High Voltage Enthusiast
  • *
  • Posts: 17
  • Karma: +0/-0
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #11 on: February 13, 2020, 01:51:14 AM »
I was thinking of going the ebay rout, just the delivery times can take so long. So when it comes to these diodes, is 100mA overkill? Or will 30mA be just fine? Im wanting to do something like how Kilovolt did with his twin towers that shoot 75cm fiery arcs. I have some HV diodes and caps to test with and I built a full wave multiplier that has 3 stages. From the second stage to gound is fiery but from the 3rd to ground is just lots of loud arcs and no fire. I figure the impedance goes up as you add more stages so if I add more stages it will just be more non-fiery arcs, but longer. I'm thinking I need to increase the capacitance at the charging columns in order to get the higher stages to be fiery, correct?

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 192
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #12 on: February 13, 2020, 05:44:24 AM »
Andrew,

As a simple approximation, open-circuit voltage determines the distance over which an spark/arc can be started.  (Electrode shape matters too.)  Average current determines the thickness of the arc.  Voltage drops as the arc forms, allowing current to increase.  The details of spark vs. arc are complex and not a subject I know much about.

To make impressive arcs, high power is the first criteria.  Kilovolt mentions his transformer being good for 6 or 7kW, but it's not clear if his entire system is running that high.  Based on his 800k-ohm output ballast resistance and 35mA short-circuit current, that's about 1kW into the resistors.  Peak power is likely higher, but I don't know how much.  Perhaps kilovolt has some input power measurements he would be willing to share.

Diode currents drop linearly up the CW multiplier tower.  Bottom diode average current can be estimated by the transformer output Vpp and power (per tower).  If running 50kVpp and 2kW per tower, that's 2kW/50kV = 40mA average current per diode.  Short-circuit conditions are likely lowering the Vpp input, so will increase diode current for a given power.  So, if you want to replicate Kilovolt's results, and if he is running around 4kW (2kW per tower), then you'll need a 4kW system.  Diode current then depends on Vpp AC feed.
David Knierim

Offline kilovolt

  • High Voltage Enthusiast
  • *
  • Posts: 13
  • Karma: +1/-0
    • View Profile
    • Hochspannung als Hobby
Re: Doorknob vs Ceramic Disk
« Reply #13 on: February 13, 2020, 08:53:43 AM »
Hello davekni

it's actually quite difficult to say how much active power the setup draws from the line. The 800kOhm resistor chain together with the short circuit current of 35mA doesn't quite hit it though. The voltage that drops over the 75cm arc cannot be neglected. Additionally there are also some losses in the bridge, the transformer and in the multiplier itself (losses during recharging and corona). The setup drew about 25 to 30A at 230V on the mains side, which gives an apparent power of almost 7kVA, but of course there is a lot of reactive power, because it was measured before the big Variac.

The ferrite x-ray transformer was, if I remember correctly, designed for 5 to 6kVA continuous power. You definitely need a certain power, otherwise you can forget arcs. But the bigger problem is that the multiplier itself must not have high internal impedance, otherwise it will never convert power in the kW range. This requires capacitors which allow fast displacement currents and have a high enough capacitance. I used capacitors with 1.7nF and 2.4nF at 50kV. Switching frequency was about 25kHz.

Maybe you first want to try a smaller setup and see how it fits. I did a lot of experimentation beforehand and built smaller multipliers to get a feel for it. Another difficulty is not to get any internal flashovers between the stages anywhere. Field shaping is definitely a must.

The diodes must be able to withstand very high pulse currents. I have used 100mA diodes throughout with 10A surge. They are mostly 2CL2FP and 2CL2FM. Please do not forget that the reverse voltage of the diodes actually applies under oil. If you operate the diodes in air, it is better to connect many diodes with lower reverse voltage in series than a few with high reverse voltage. I only used the 2CL2FP because I got it as a gift.

Hope this helps!

Best regards
kilovolt
« Last Edit: February 14, 2020, 09:05:38 AM by kilovolt »
All information on my part without guarantee! I reject any liability for personal and/or material damage. Everyone is responsible for his own safety.

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 192
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #14 on: February 14, 2020, 05:42:19 AM »
Kilovolt,

At 25-30A, the variac reactive power is probably not very significant.  Even 10A out of 30A still leaves sqrt(800) = 28.3A.  The more significant part of power factor is likely from any bulk capacitance after the bridge rectifier.  Do you have bulk capacitance there, or does your inverter power track the rectified line voltage?

Yes, I expected operating power would be higher than short-circuit power.  That's why I used 4kW for the diode-current calculation example.  Even that is probably low given your line current measurements.

With 800k total series resistance, you are way below those diode's 10A surge current rating.  700kV would allow just under 1A peak discharge current, of which the diodes take half, so under 0.5A surge.

Nice project!  My largest CW multiplier is for a Van der Graaf generator alternative, so very low current.  It's max is about 250kV with a 300meg output resistor string.
David Knierim

Offline kilovolt

  • High Voltage Enthusiast
  • *
  • Posts: 13
  • Karma: +1/-0
    • View Profile
    • Hochspannung als Hobby
Re: Doorknob vs Ceramic Disk
« Reply #15 on: February 14, 2020, 08:27:03 AM »
Hello davekni

After the bridge rectifier there is a cap with 3300µF / 450V.

The diodes have to handle not only the discharge currents, but of course also the charging currents, and these will be much higher. In another forum someone had simulated the circuit, and the peak currents were quite high, several amps if I remember correctly. I also tried to measure the currents with a shunt and a battery driven scope at the lowest stage. But you couldn't tell for sure what was actually measured and what was only interference.
I am pretty sure that the diodes are the weakest link in the chain and I would definitely not save your money there. In the beginning I also had problems with internal flashovers between the stages. When that happens, the load on the diodes becomes very high. I strongly recommend to use 100mA diodes, but of course everybody has to know that for himself ;-)

250kV is a nice voltage ;-) How did you build the setup? Do you have pictures?

Best regards
kilovolt

« Last Edit: February 14, 2020, 09:01:21 AM by kilovolt »
All information on my part without guarantee! I reject any liability for personal and/or material damage. Everyone is responsible for his own safety.

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 192
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #16 on: February 15, 2020, 05:41:13 AM »
Kilovolt,

With 3300uF bulk cap, power factor is likely around 60 to 65%, for ~4kW of real power.  This is now based on LTSpice simulations.  Does depend on variac inductance, both parallel and leakage (series).  With different assumptions, ranges from 50% to 70%.

If the diode peak current during charging is higher than the ~0.5A discharge, then the X-Ray transformer must have quite low leakage inductance.  For 5kW (2.5kW per tower) and 50kVpp, average current of the bottom diodes should be 50mA.  I'd expect transformer inductance to spread out the current pulses to 60 degrees or so of the sine-wave cycle, but it could be shorter with low leakage inductance.  So, perhaps peaks of >10x average are possible.  Agreed that for your power level, 100mA diodes was a great choice.

Not suggesting it is worth the work, but if you want to measure peak diode current:  I'd suggest 50-ohm coax with 50-ohm termination resistors at both ends.  Wind as many turns as possible around a large ferrite core (E-core or toroid) for a common-mode choke.  One end of the coax in series with the ground connection of the bottom capacitor of a tower.  Other end to an oscilloscope.  Current is, of course, 25 times measured voltage.

I don't have any construction pictures for my large multiplier, but here's a schematic and image of the finished product:




This project was ~7 years ago, so my memory isn't perfect.  I think the diodes were 10mA 20kV.  The bottom stages used two diodes in series, as the voltage per stage is a bit lower farther up the string.  Caps near the bottom were 30kV, upper ones 20kV.  All the caps were cheap ceramic disk caps.  A couple bottom caps failed early on driving it at 20kHz.  I think the current from just stray capacitance was enough to overheat them.  Dropped to 700Hz after that initial failure - using a spark coil feed.  The caps and diodes are potted and glued to the outside of 1.5" PVC pipe using silicone rubber caulk.  (The silicone that gives off alcohol during cure, not the acetic acid type.)  That pipe is centered inside the 6" PVC outer pipe visible in the picture.  Top is a 10" stainless steel garden-gazing ball.  The multiplier is the lower ~60% of the height.  The upper ~40% is a string of 32 10meg 10kV (peak) resistors, also attached and potted with silicone rubber.

The large output resistance is is for safety.  Primary use is to charge people, making their hair stand up and/or launching foil pie tins from their hand.  For that use, ~100kV is sufficient.  I ran it to 250kV shortly after construction just to see what the upper limit was.  It hasn't been that high since.  Gets used a few days a year at different events.

Given the high output impedance, the only way I could measure voltage was by placing progressively heavier (thicker) metal disks on top to see if they could lift off by electrostatic repulsion.  (Weight per unit area is what matters here, along with the radius of the top ball.)  Even my 1G-ohm resistor strings load down the output too much, due to the 320meg output resistance and due to small capacitors with only 700Hz drive.  (That effect was much less at my original 20kHz.)  Measuring with resistor strings was also affected by the ion wind that added a bit of extra current.  Probably could have shielded enough to fix that issue.

I found your comments about charging walls and other objects around quite interesting.  I have that problem when in smaller rooms.  It takes more voltage to get hair to stand up later in the day, as the charged walls and ceiling repel the hair.  Neighboring stations notice all their objects getting charged too.
« Last Edit: February 15, 2020, 06:44:13 PM by davekni »
David Knierim

Offline kilovolt

  • High Voltage Enthusiast
  • *
  • Posts: 13
  • Karma: +1/-0
    • View Profile
    • Hochspannung als Hobby
Re: Doorknob vs Ceramic Disk
« Reply #17 on: February 17, 2020, 08:55:30 AM »
Hello Dave

Thank you very much for your explanations and math regarding the power factor of my setup. Very interesting thoughts. I never simulated the situation on my own, because I'm not familiar with all these simulation porgrams on the market  :) LTSpice seems the standard tool for this, right?

The measurement with 50Ohm termination and 50Ohm shielded cable is a good idea. Maybe I'll try out some day.

Thanks for sharing your multiplier setup. Nice design and cool housing! Good that you have optimized your device for safety.  If you carry out experiments directly with people, this is certainly the right way.

Do you also build Tesla coils? I've been building a QCWDRSSTC for a few years now, but there have been a lot of setbacks. But now it seems to work slowly. Besides, I still have good material for a VTTC. Unfortunately, as a family man, I lack more and more time for new craft projects.

Best regards
kilovolt
All information on my part without guarantee! I reject any liability for personal and/or material damage. Everyone is responsible for his own safety.

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 192
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Doorknob vs Ceramic Disk
« Reply #18 on: February 18, 2020, 05:33:22 AM »
kilovolt,

LTSpice is the tool I'm familiar with.  Use it at work and at home.  I hear about other free options, but have no personal information on those.  When LTSpice first came out, I ran some comparisons with other simulators.  LTSpice seemed to converge better and run faster on complex circuits than some of the commercial paid simulators.  I don't think it's being improved any more, so other simulators may have caught up or surpassed LTSpice by now. When I worked for Tektronix, their internally-written simulator was much better than any commercial option.  (I was on a team that evaluated the possibility of Tektrnix selling it as a commercial simulator, but such would have required 10x larger a group for documentation and support etc.)  Freescale's internal MICA simulator is the next best one to my knowledge, but also not publicly available.  My use of MICA is a decade or so old as well, so its ranking may no longer be valid either.

Yes, I've built three Tesla coils.  The latest is a DRSSTC documented on this forum:
https://highvoltageforum.net/index.php?topic=798.msg5332#msg5332
I plan (perhaps in 2021) to try something somewhat like QCW - to increase primary capacitance as the arc length increases during each firing - discussed some in the above link.  Haven't played with vacuum tubes for many decades, and never high power ones.

Yes, projects take too much time.  Even though my "kids" are out on their own, I'm still working and volunteering (science activities for Foster-Parent-Night-Out events mostly), so project progress is slow.  Keep family as the higher priority as they should be.
David Knierim

High Voltage Forum

Re: Doorknob vs Ceramic Disk
« Reply #18 on: February 18, 2020, 05:33:22 AM »

 


* Recent Topics and Posts

post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
Today at 02:44:44 PM
post Re: 230 volts spark gap
[Spark Gap Tesla Coils (SGTC)]
klugesmith
Today at 07:21:13 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ritaismyconscience
Today at 06:11:32 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
Today at 04:42:34 AM
post Re: 230 volts spark gap
[Spark Gap Tesla Coils (SGTC)]
davekni
Today at 04:37:53 AM
post Re: UV-cured cyanoacrylate glues?
[Light, Lasers and Optics]
Twospoons
Today at 04:28:33 AM
post Re: UV-cured cyanoacrylate glues?
[Light, Lasers and Optics]
klugesmith
Today at 04:21:31 AM
post GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ritaismyconscience
Today at 02:41:03 AM
post Re: 230 volts spark gap
[Spark Gap Tesla Coils (SGTC)]
Quentief
Today at 12:28:20 AM
post Re: UV-cured cyanoacrylate glues?
[Light, Lasers and Optics]
Twospoons
February 20, 2020, 09:45:21 PM
post Re: UV-cured cyanoacrylate glues?
[Light, Lasers and Optics]
klugesmith
February 20, 2020, 07:28:01 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Uspring
February 20, 2020, 02:06:26 PM
post Re: UV-cured cyanoacrylate glues?
[Light, Lasers and Optics]
shrad
February 20, 2020, 10:38:01 AM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
davekni
February 20, 2020, 05:14:43 AM
post Re: Is this flyback OK for my Plasma ball ?
[Transformer (Ferrite Core)]
davekni
February 20, 2020, 05:04:50 AM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
klugesmith
February 20, 2020, 04:47:41 AM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 19, 2020, 07:29:23 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 19, 2020, 06:47:55 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 19, 2020, 06:45:23 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Teravolt
February 19, 2020, 06:41:42 PM
post Re: Is this flyback OK for my Plasma ball ?
[Transformer (Ferrite Core)]
John123
February 19, 2020, 06:10:34 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
MRMILSTAR
February 19, 2020, 05:52:20 PM
post Re: Is this flyback OK for my Plasma ball ?
[Transformer (Ferrite Core)]
ElectroXa
February 19, 2020, 05:05:50 PM
post Re: Is this flyback OK for my Plasma ball ?
[Transformer (Ferrite Core)]
John123
February 19, 2020, 04:11:22 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Uspring
February 19, 2020, 02:03:12 PM
post Is this flyback OK for my Plasma ball ?
[Transformer (Ferrite Core)]
warmania
February 19, 2020, 11:24:58 AM
post Re: HP T620 Thin Client - PSU Hack And Windows PC Modification
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
latethannever
February 19, 2020, 11:00:15 AM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
rikkitikkitavi
February 19, 2020, 10:43:13 AM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 19, 2020, 06:54:14 AM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
klugesmith
February 19, 2020, 06:51:52 AM
post Re: 230 volts spark gap
[Spark Gap Tesla Coils (SGTC)]
davekni
February 19, 2020, 03:53:13 AM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
ElectroXa
February 18, 2020, 09:44:25 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
MRMILSTAR
February 18, 2020, 09:34:06 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 18, 2020, 08:27:31 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Weston
February 18, 2020, 07:55:14 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 18, 2020, 07:32:04 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
klugesmith
February 18, 2020, 05:58:39 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
MRMILSTAR
February 18, 2020, 05:16:55 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 18, 2020, 04:19:37 PM
post Re: Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Mads Barnkob
February 18, 2020, 04:12:11 PM
post Re: 230 volts spark gap
[Spark Gap Tesla Coils (SGTC)]
Quentief
February 18, 2020, 01:25:06 PM
post Help!Capacitor that burns out after a while!
[Voltage Multipliers]
Peregrine
February 18, 2020, 10:40:58 AM
post Re: How can I make my driver better
[Solid State Tesla Coils (SSTC)]
Mads Barnkob
February 18, 2020, 09:02:23 AM
post Re: DIY 250W LED Flood Light for High Speed Filming
[Light, Lasers and Optics]
Mads Barnkob
February 18, 2020, 08:58:38 AM
post Re: Doorknob vs Ceramic Disk
[Voltage Multipliers]
davekni
February 18, 2020, 05:33:22 AM
post Re: DIY 250W LED Flood Light for High Speed Filming
[Light, Lasers and Optics]
davekni
February 18, 2020, 05:08:28 AM
post How can I make my driver better
[Solid State Tesla Coils (SSTC)]
Irido
February 18, 2020, 04:45:06 AM
post Re: DIY 250W LED Flood Light for High Speed Filming
[Light, Lasers and Optics]
klugesmith
February 18, 2020, 02:44:42 AM
post Re: HP T620 Thin Client - PSU Hack And Windows PC Modification
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
latethannever
February 18, 2020, 01:58:48 AM
post Re: UV-cured cyanoacrylate glues?
[Light, Lasers and Optics]
Twospoons
February 17, 2020, 11:30:03 PM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
SimplePortal 2.3.6 © 2008-2014, SimplePortal