Switch Question / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Switch Question

previous 1 2 3 4 next last

Move Thread

LAN_403

Antonio

Sat Nov 03 2012, 01:02AM

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

That schematic requires a floating driving voltage at the gate, since the source follows the input with the transistor conducting. It's simpler to drive the transistor if it is between L1 and C1.

Newton Brawn

Sat Nov 03 2012, 04:13PM

Registered Member #3343 Joined: Thu Oct 21 2010, 04:06PM
Location: Toronto
Posts: 311

GamaRay:
You wrote:

"Essentially, I am asking for advice on what device to use for Q1 (that can handle 1kV (+/-) and is preferably isolated so I don't fry my function gen)."

How about replace Q1 by four sidacs in series? Each sidac can have a disrupture voltage of 240V, totalizing 960V. A very small capacitance (hi V insulation) between the Vin terminal and ground will complete the circuit .
Sidacs replaces the gas-filled spark arrestor .

GammaRay

Sat Nov 03 2012, 04:50PM

Registered Member #5323 Joined: Fri Jun 15 2012, 02:14PM
Location:
Posts: 104

Newton:

I have also used several Sidacs in series as substitute for the Spark Arrestor, they work pretty good, but can't be controlled by my Function Gen (that I know of, anyway).

Newton Brawn

Sat Nov 03 2012, 06:23PM

Registered Member #3343 Joined: Thu Oct 21 2010, 04:06PM
Location: Toronto
Posts: 311

Yes, Antonio suggestion make sense.

In your circuit the input has to be insulated of the output.
Depending of the switcing frequency you may add a small pulse transformer between the generator and Q1,.. " The input will be free of output".

Let me know the triger voltage and triger frequency

Otherwise ...

It was common in the 50th using a vacuum tube triode as cathode follower to detect the electrostatic voltages.
The plate voltage about 60V,
a 0-1 mA miliamperimeter between cathode and ground,
plate resistor about 30-50kohms,
grid resistor to ground = 10megohms...

Or

Do you want to be free of any external power supply, only the energy supplied by the "searching rod" (from Benjamin Franklin, Philladelphia 17XX ) ?

Electra

Sun Nov 04 2012, 11:39PM

Registered Member #816 Joined: Sun Jun 03 2007, 07:29PM
Location:
Posts: 156

If your spark arrestor only fires at a very low frequency, a few times a second for instance, assuming it takes a while for the input cap to recharge. What about using a high voltage reed relay instead? Iâ€™ve come across vacuum reed switches than can do about 15kv.
Itâ€™ll probably take a lot more power to switch the coil on or off, than youâ€™d ever pick up electrostatically. But itâ€™s easy to do to test the concept, as the coil is already inherently isolated.

I thinks Steve.C has already spotted the problem with using most semiconductors.
If you conducted some measurements to see how much input current you could collect, youâ€™d be able to compare it with your mosfet.

I donâ€™t know if any ultra low leakage fets for a 1000v are about, you might be able to find some lowish current ones that have a couple of uA or so.

I was thinking if you build a driver to pull the gate negative like - 5 or -10v in the off state, if that would have any affect the leakage current or not, just a guess but probably not enough to make any practical difference.

GammaRay

Mon Nov 05 2012, 09:22PM

Registered Member #5323 Joined: Fri Jun 15 2012, 02:14PM
Location:
Posts: 104

I tried the circuit using the CLARE Mosfet between L1 and C1, but didn't work (possibly because as Newton suggests the HV input needs to be isolated from the output?) Can you suggest which pulse transformer I could try?

Yes, I'd like the energy to be supplied through the 'searching rod'. One of the goals of all this is to find out what frequency is most efficient, thus I want to control Q1 with my function gen. Normally, the spark arrestor fires between 1 Hz and 100 Hz depending on the strength/distance of the nearby ion source to the searching rod, however, I wonder if a greater frequency would produce better efficiency.

I'd like to try the suggested IGBT, anyone recommend a specific one to try?? I would also like to try the Vacuume tube if I knew which one to try (never used tubes before).

Thanks to all.

Newton Brawn

Wed Nov 07 2012, 02:24AM

Registered Member #3343 Joined: Thu Oct 21 2010, 04:06PM
Location: Toronto
Posts: 311

Gamaray :
You wrote :
"a circuit that, when placed very close to an ion source such as a cathode tube style television set or an Ionic Breeze air cleaner, etc, scavenges enough airborne ions to power small loads such as an LED or small motor."

This induce me to believe that the purpose of your circuit is get the hv energy of air and convert to 2 ~ 3V lighting a led,

Then, you do not need a device (as the CLARE) to do it . The switching will be done by the arrestor.

The best efficience will be the achieved by direct discharge of the input capacitance to the coil, the dischage can be done by a air spark gap, gas-filled spark arrestor, or 4 sidacs.
The dischage also may be performed through a reed reed relay without any coil to avoid any stray capacitance, In this situation the reed relay works as a sparg gap in a high vacuun chamber, ( the disrupting voltage shall be determined)

what is the input capacitance ??
what is the lenght of your lightning rod ?
Can you post one good picture of your 12mH coil ?

depending the above parameters think your device will be self powered...

Ash Small

Wed Nov 07 2012, 11:36AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

GammaRay wrote ...

I would also like to try the Vacuume tube if I knew which one to try (never used tubes before).

Thanks to all.

I think most old TV's had a tube in them rated for around 5kV, and sufficient wattage for your needs, but I don't have much in the way of part numbers. Maybe someone more knowledgable in this area could suggest something?

Electra

Wed Nov 07 2012, 10:34PM

Registered Member #816 Joined: Sun Jun 03 2007, 07:29PM
Location:
Posts: 156

If you ion source behaves as a constant current source up to a certain voltage, then you may want your circuit to operate at the highest input voltage x current point, for maximum power transfer.
When your spark gap or discharge tube fires it will discharge the input capacitance to the voltage where the spark extinguishes, then it will have to wait till it climbs all the way back up again.

If you could maintain a higher average Vin.

Example

Say you found (I) collected is 10uA, you chose a input capacitor of 0.01uf (a guess)

Voltage of spark gap or whatever 1000v

Time to reach 1000v
T= CxV
------- = 1 Second
A

E=1/2 CxVsquared

0.01uf charged to 1000v discharged to 0v (to make it simple) = 0.005mJ
0.005mJ transferred at 1Hz = 5mw.

Or

0.01uf charged to 1000v discharged to 800v. Energy start 0.005mJ
0.01uf charged to 800v. Energy left 0.0032mJ

Energy transferred = 0.0018mJ at 5 Hz = 9mw

Ok did make a lot of simplifications ignored the time to discharge the capacitor assumed it very fast compared with what it takes to charge, and neglected the difference between the out put voltage as itâ€™s low 5v, and the current was constant all the time.
In a real circuit there will be losses as well.

Hope thereâ€™s no silly mistakes I made in that.

Newton Brawn

Thu Nov 08 2012, 12:13AM

Registered Member #3343 Joined: Thu Oct 21 2010, 04:06PM
Location: Toronto
Posts: 311

Hi Electra,

I agree with your calculations:

Time of a 10uA to charge a 0.01uF cap with 1000V:
I=c*dv/dt >> dt=c*dv/I = 0.01*0.000001*1000/0.000010 = 1second

Energy in a 0.01uF cap charged to 1000V:
E= 0.5*c*V*V = 0.5*0.01*0.000001*1000*1000 = 0.005joules or 5 milijoules.
This 5milijoules are tranfered to the coil during the discharge.

What I did not understand is why the capacitor shall dischage only 1.2miijoules in the coil if the capacitor has delivred 5milijoules to the coil. Do you mind to explain ?
Thanks

previous 1 2 3 4 next last

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