Please help me finish/check my HV Circut.

[lanewaddell]

Hello all,

I could really use some feed back on my HV Circut design for my fusor using a 600 Watt 60kv precipitator power supply (see attached). I've been spinning my wheels a bit and could use the hand as this is NOT my area of expertise. : )

I believe my voltage divider is good. It should read 1000:1 directly to a digital voltmeter capable of negative polarity. 60 volts at maximum.

Where I am confused is R1 and R4 in the diagram. I know enough to know I need some kind of protection against a peak discharge from the power supply capacitance as well as protection from a long-duration direct discharge to ground but I am getting advice off-forum ranging from only needing five OY103KE resistors here amounting to 50 kOhm, 10 watt/2000v sustained/20kv pulse all the way up to 150 watt/50 kOhm. Me personally, I thought I needed to protect against the whole 600 watts or 60kv to be safe, which would require about $200 bucks worth of specialty resistors from China.

I could really use some practical advice about what is needed for R1 as well as verification that the ground current measuring resistor R4 is adequate.

Thanks!

[Dennis P Brown]

The ballast resister is talked about in the FAQ, which one should read; if memory serves, it is 60 K-ohms to 100 K-ohms and needs to handle no less than ten watts and would then need to be placed in oil for better heat removal. Twenty watts or higher would be better.

See: viewtopic.php?t=9383

Chris Bradley has an excellent post in that section relative to ballast resisters.

As for running the ground for the fusor through a 5 watt resister, huh? How will that handle 600 watts continuous?

Again, see: viewtopic.php?t=4266

These FAQ's are necessary for safe operation of a fusor. Read them.

That supply is 100% lethal and not being familiar with HV systems is certainly a major safety issue for you - this is not a safe system to use if one has no experience with HV. You really need to understand the dangers of HV, proper wiring and how to properly handle such a supply to prevent a dangerous situation. Also, that voltage level can create very dangerous x-rays, as well. You really need to read the FAQ's sections. Note plural.

[JoeBallantyne]

If you are using a precipitator supply, you don't actually need R1 at all. I used a 40kV precipitator supply without a ballast resistor, and it worked just fine.

If you are using some other supply that does not have current limiting built in, then a ballast resistor is necessary.

You need to make sure that you have earth ground DIRECTLY connected to your fusor vacuum chamber shell.

For measuring your current, you are better off just using an ammeter directly between the positive lead of the precipitator (negative lead goes to the cathode HV feedthrough) and your grounded fusor shell.

Get an ammeter that measures 15-25mA max. IMO analog meters are easier to use for controlling the pressure based on the current. Since the needle moves continuously vs periodic updates to a digital screen.

Using an ammeter is much simpler than trying to measure the voltage across another resistor that you add to the setup. Note that I do have kits for both the HV and the mA meter. My HV kit however is NOT rated for 60kV. 40kV is its present maximum HV limit - due to the HV wire being 40kV rated only. So you should use your planned setup for HV measurement if you are going to push the voltage up to 60kV. You WILL need lead shielding once you go above about 35kV or so, as your stainless fusor shell will be transparent to ~35keV x-rays and above. I would plan on lead shielding if you are going above about 30kV, just to be safe. And make sure that any viewport is shielded with leaded x-ray blocking glass no matter what. Even at 10-15kV, you will have a LOT of x-rays pouring out your viewport.

Make sure that any HV wire you buy and use in your setup is rated for 60kV. (Might be a bit hard to find, as most inexpensive HV wire maxes out about 50kV.)

Make sure you put that precipitator under oil.

My 2 cents.

Joe.

[Rich Gorski]

Lane,

Your circuit diagram is sound from an electronics point of view. My comments:

R4: A 5 watt resistor (R4) at 10 ohms is fine for measuring the fusor current. It will not dissipate 600 Watts. At maximum the current going through R4 will be 0.025mA (probably considerably less). So the power dissipated by R4 is P = I^2*R = 0.025^2*10 = 0.0063 Watts. No problem. Under normal operation almost all of the power will be dissipated in the fusor chamber and cathode grid. Even if the fusor shorted out the power supply would also shut down as it couldn't handle the load of the ballast resistor to ground. You can think of the plasma itself as having an apparent resistance of a few M-ohms and in series between R1 and R4. That's why most of the power is dissipated there.

R1: If you use a ballast resistor it will dissipate 10s of Watts. P = I^2*R = 0.025^2*100,000 = 62.5 Watts.

Joe's thought of using an analog meter in place of R4 is sound. I also like the idea of not having to wait for the digital display to update especially when things are not stable. Don't forget to include a 100V gas discharge device (or a neon lamp) across the terminals of the analog meter or R4. This provides protection in case R4 or the analog meter decides to open up. If it opens and you loose ground connection, without a protection device you could see high voltage down at the analog meter or on R4.

R3: You might want to make a correction to R3 (300K) since your volt meter (I assume digital) will have an impedance of around 10Mohms. So the 10M is in parallel with 300K so your 1000:1 divider is a bit off like 3%. Not much to worry about however so I'll let you go ahead and do the calculation for resistors in parallel if you want to be a bit more accurate.

Rich G.

[lanewaddell]

Dennis,

Thank for the information admittedly even the countless FAQs I've read I've forgotten about. I read them both twice.

Joe,

Thanks again for the input. So you didn't ballast your China Special precipitator? You weren't worried about a direct discharge? I believe the extensive study done on these from several years back indicates they do have a current limiting feature (an induction wire by the transformer) but that it has to be removed because it cuts it out on ignition. Mine doesn't appear to have that red wire.

I like your suggestion of just using an anmeter, but in the second FAQ linked above the suggestion was to always measure across a resistor. The anmeter sounds much simpler to me though.

A dumb question though, why do the anmeter wires have to be rated for the maximum 60kv voltage of they are on the ground leg and protected by a gas tube in parallel? When protected shouldn't the worse case voltage just be 100 v if the anmeter opens?

I should state that although this is rated for 60kv, I'm only ever going up to 45.

[lanewaddell]

Rich G.,

Thank you so much for the analysis, that was SUPER helpful!

I noticed you used 0.025 amps - for 60 kv and 600watts isn't this 0.010 amps? Is that a 2.5x peaking factor applied?

[lanewaddell]

Also, it would be good to know if I'm going to kill myself using mineral oil to pot the supply.

It has a very high breakdown rating, is cheap, easy to get and I've seen other HV DIY projects use it.

[Dennis P Brown]

Mineral oil is ok and people here use it. Auto synthetic engine oil is also very good (in tests superior to many transformer oils from a research paper I read.)

I've run my fusor x-former through a milli-amp gauge mounted atop the HV feed (full 30 kV and 100 ma) with no problem. It was air isolated. However, I generally run a milliamp meter off one leg of my diode bridge (no resisters) and that works well but can be misleading at times (get's fooled by the plasma when resistance gets very high.)

I measured my variac voltage output and convert that into KV after calibrating the HV output using a HV probe (Giga-ohm resister in series with a 50 ua meter.). That variac system too can lie (again, the plasma resistance) but is an ok guide.

RMS matters but yes, 600 watts at 60 kV (includes the RMS value) will yield a max. of 0.01 amps.

[Rich Gorski]

Lane,
I just used the maximum current rating Joe gave for the ammeter as a worst case, but you’re correct to use 0.010 as that is the max the supply can deliver. My analysis numbers just get smaller by 2.5 times.

As for the mineral oil, that is great to use to insulate Hi voltage and you are safer using it than not using it. I use it to both insulate and cool my 50K ballast resistor. It’s far more important that your hi voltage cables are rated for your voltage (60kV ??) and be careful around any exposed metal connected at the Hi voltage feedthrough. To get to voltages above 30kV the feedthrough connector on the air side will have to be on a corona globe, immersed in mineral oil or maybe silicone to prevent corona. Corona will start in air around the sharp corners of unprotected metal terminals somewhere around 20-30kV. The large size and smoothness of the globe greatly reduces electric field strength (no sharp points).

Rich G.

[lanewaddell]

Hey all,

So I sat down and worked through what you all said. I think I've incorporated everything correctly. For the ballast resistor how does this look?

https://www.mouser.com/ProductDetail/71 ... 06E50K00JE

[lanewaddell]

For insulation the precipitator has horrible thin insulation that I'll assume basically isn't there.

Above oil In was going to have it in two concentric runs of PVC and then maybe use a spark plug wire to bring it to the base of the feed through.

The way I have everything else set up I'm going to try to control this from another room with Bluetooth and cameras. I'll be near the house breaker and will have a tool to hit the pre -identified breaker if things go south. I also bought a controllable 120 volt outlet that I'll hopefully be able to kill remotely.

[Dennis P Brown]

I do not like the idea of using the amp meter from the fusor to ground. You can do as Joe said and then just ground the chamber. Trusting all the components to enable a proper ground is placing your safety on that arrangement. in worse case failure, the chamber would be at 60 kV and deadly. Could arc to over parts creating a dangerous situation. I prefer a grounded chamber. Far safer. You get no second chances with these power supplies.

[Rich Gorski]

Sorry, I missed that about the ammeter.

Joe and Dennis are correct. Ground the chamber directly and connect the ammeter (or R4) between the positive lead of the 60kV supply and ground. See photo below.

Rich G.

[lanewaddell]

Thanks all!

So it is spoken, so it shall be! I appreciate the help.

Two open questions -

1.Does the ground wiring have to be HV rated going up to the anmeter? Nominally when everything. Is connected, the voltage should only be due to the resistance of the meter right? In an upset condition it should be limited to 100 volts due to the glow tube?

2. Is that resistor I picked out adequate?

[Rich Gorski]

1.Does the ground wiring have to be HV rated going up to the anmeter? Nominally when everything. Is connected, the voltage should only be due to the resistance of the meter right? In an upset condition it should be limited to 100 volts due to the glow tube?
No, the wire going up to the ammeter is connected directly to earth ground so no need for HV cable. If the ammeter were to open the glow tube which is also connected to ground would keep the supply grounded. Actually the wire could be bare copper as earth grounds often are. It should be a good gauge wire maybe 12 AWG or thicker. That promotes good grounding.

2. Is that resistor I picked out adequate?
Yes

Rich G.

[Rich Gorski]

Something to add...

The ballast resistor you chose is fine just remember that it is at high voltage so it needs to be immersed in oil as well. If you push 10mA of current through it, it will drop V = IR = 0.01*50000 = 500 volts. so the voltage rating on the resistor needs to be 500V or greater. Based on the resistor data sheet it seems to have a voltage rating of 1870 volts. So all is fine.

Rich G.

[JoeBallantyne]

Lane -

If I were you, I would NOT use any ballast resistor. Why unnecessarily complicate your life?

I most certainly would NOT in any way try to modify the precipitator supply by cutting wires or anything else.

Just use the supply exactly as it was shipped to you.

Put it in an oil bath, so that it is completely submerged in mineral oil - or synthetic engine oil.

Connect the negative output (black) to your HV feedthrough.

Connect the positive output (red) to the positive terminal of an analog ammeter. (Preferably one with a max reading of ~15mA or so.)

Connect a 100-120V GDT across the ammeter terminals.

Connect the negative ammeter terminal to your fusor shell. (You can use low voltage wire for this connection, as it will always be at ground potential.)

Connect your fusor shell directly to ground. (I would use both the neutral and ground terminals from a wall socket = both of which are wired together to ground at your breaker panel).

If you do this as described, you don't need to buy any HV wire at all. You will simply use the HV wire that comes out of the HV transformers on the precipitator supply. Which should already be rated for 60kV.

Joe.

[lanewaddell]

Hey Joe,

Thanks for that.

I was originally planning on just using the 60 kv wire that was already on the supply. I have to cut into it for voltage monitoring anyway in order to have that junction under oil.

Are you saying that these come with adequate protective circuitry? Have you had one short to ground in use a d shut off?

Just curious.

Thanks

[JoeBallantyne]

You don't have to put the HV monitoring connection under oil. You could connect it directly to your HV feedthrough along with the HV wire from the precipitator. If you don't put it under oil, you alternatively would need to encase the HV resistor lead in HV silicone putty to prevent corona.

I never used the 60kV supply that you have, but the 40kV supply that I used arced to the vacuum chamber wall a few times and survived OK. It did NOT survive many many repeated arcs (due to internal fusor contamination from vacuum grease). But I doubt that having a ballast resistor would save any of those precipitator supplies from many many repeated arcs to ground.

The supply will drop its voltage if the current rises too much. That is by design, and does make it easy to use a precipitator to drive a plasma in a fusor for long periods of time at widely varying pressures. However, when in that mode, the voltage is basically zero and the current is at ~6mA or so if the voltage was at the supply's minimum setting before firing up the plasma. At least with the 40kV supplies that I was using, if you want the voltage to stay anything close to what you set it to initially with the voltage potentiometer when there is no plasma lit, you HAVE to keep the current at 3mA or less. Go much above 3mA and the voltage just drops like a stone.

Joe.

[Bob Reite]

You do NOT want a low range millammeter in the ground return from the power supply without some sort of protection! Should it fail open, the chamber will no longer be grounded and go close to the negative high voltage output of the power supply. Measuring the voltage drop across the 10 ohm power resistor as shown in the schematic is safer.

[Rich Gorski]

The chamber is directly tied to ground as shown in the marked up schematic and will stay grounded regardless of what happens to the milliammeter. Also Lane (the poster) was advised to have a gas discharge device across the milliammeter terminals to keep the power supply grounded if it opened. But I do agree that I would have more trust in a resistor staying intact (or several in parallel) compared to a milliammeter.

Rich G.

[Richard Hull]

100 ohm power resistor might be better with 10ma in a running fusor. this would develop 1 volt across it. 10, 1000ohm 2 watt 2% resistors in parallel with a trim in parallel or series would be a 20 watt 100 ohm resistor. A momentary 500 ma surge would develop 50 volts across this resistor and 25 watt dissipation. This allows plenty of time to respond to shut down with no harm to the 100 ohm composite resistor. Besides, few power supplies would fold up trying to send 500ma to a fusor in trouble.

This assumes that If you are using a strong brute of a supply, (X-ray Xfrmr), that you already have a 50,000 ohm ballast back at the HV negative terminal. Noted as above postings): You do not need a ballast resistor if you are using a rather whimpy Chicom HV switcher supply as these either self-current limit or fail disastrously, automatically trying to deliver even 20ma to a fusor.

If you have a ballsy brute supply, a 50,000 ohm ballast, (current limiting) resistor would only steal 500volts from your HV at a running fusor drawing 10ma. (5 watts dissipated in the 50k resistor at 10ma fusor run) or 1000 volts chopped off the top of your HV to the fusor and 20watts dissipation in the ballast at 20ma fusor run current.

Richard Hull

[lanewaddell]

Mr. Reite,

I have the equipment to do either - do you still hold true to that with a gas discharge "fuse" across the anmeter?

[Bob Reite]

OK you mention that the chamber is always at ground so the metering is between the positive terminal of the supply and the grounded chamber. That would be safe. For extra protection put a 2 watt 10 volt or so Zener in parallel with the meter, so that if the meter goes open circuit, the maximum voltage would be 10 volts.

As an aside I remember working on a cheap tube type FM transmiitter that was metering the cathode current. The meteing faled open and the resulting 7500 volts at lots of amps totally fried the control and metering board before an overload tripped.

[Richard Hull]

The ammeter should always be in a series power circuit!!! Three ways to hook in an ammeter.....

1. If you hard ground the fusor shell then the HV hot negative terminal of the supply goes to the ammeter then to the high voltage grid insulator terminal ( Bad..Every thing in the ammeter circuit is forever hot and deadly!)

2. Fusor shell to ammeter to ground (best hookup as all is at or near ground.

3. Hard ground the fusor shell and hook the positive of the power supply to the ammeter and then on to a hard ground.

My God! How tough is it to understand?!!! This is hyper easy to understand as it is a simple serial power connection. #2 is the easiest and most safe as long as the ammeter circuit is robust as described above.

Formal HV, Hot negative power supplies, as sold commercially, will have a voltmeter and an ammeter built into them using #3, with the metal case at hard ground!! These are totally safe and the fusor shell goes to a hard ground

If you can't afford this kind of supply $$$$..... You must build your own supply and all 3 choices are yours to make.

For the monkey see monkey crowd, I supply a diagram. Click on it to enlarge

Richard Hull