High Voltage low current power supply

I would like to use a 9 volt battery to create a high voltage source. It's been awhile since I've messed with electronic componets. But it seems like I need something like a 555 timer to create a series of d.c. voltage pulses so I can then use a transformer to step up the voltage from say 9 volts to 900 volts.

Is the 555 timer the way to go? or is there a better device to turn the dc into a pulsating dc to step up with a transformer?

Thanks in advance. Robert

Indeed. The venerable Motorola (now ON Semi) mc34063A can be called the 555 of the switcher world. Their app note an920 teaches how to use it and has a high-voltage circuit. The mc34063AP1 (miniDIP package) is $1.11 at DigiKey in singles, soic reels are as low as 30 cents.

If you are going to 900V at low currents in one jump, a resonant circuit may be better. The stray capacitance of the secondary of the transformer is going to reflect back into the primary and cause losses in a normal square wave switcher.

You can go up to a lower voltage with the transformers and then voltage multiply in the output rectifier to avoid the issue.

I once make a 1000V 1mA power supply that ran at 1MHz. I just used a transistor oscillator to do the power switching and an op-amp to modulate the oscillators duty cycle. It worked quite well but I had to place the op-amp on a different bench because no op-amp would work near the transformer. The efficiency was about 50%. I figured that the losses were %10 as heat generation and %40 as radiated RF.

Unless you built it as a HUGE circuit, it's extremely unlikely that

40% was radiated as RF. A square loop about a foot on a side of 18AWG (1mm) copper wire has about 100,000 times more dissipative resistance than radiation resistance. ;-) Similarly short electrical dipoles wouldn't fare much differently.

The electric or magnetic fields may have been locally relatively strong, but the likelihood that you actually put out much 1MHz electromagnetic radiation is very small.

Cheers, Tom

You assumed a loop. Try it as a a pair of plates with 1000VAC on them.

OK. Even less EM radiation.

1MHz is a silly speed to operate a power supply at. A 60kHz switcher is all this application needs.

A little more to the point than the other replies I've seen here, you likely will want to make a flyback-type supply, using stored magnetic energy to generate a primary-side pulse that's a lot larger than 9 volts. Then the transformer turns ratio won't have to be astronomical to get your 900 volts. You can make a blocking oscillator with one transistor and a transformer, and not worry about using an IC at all, though the regulation may not be what you want.

Note that Harbor Freight sells an "electronic fly swatter," sometimes on sale for next to nothing, that generates a high voltage from a couple of D cells. In fact, I see it listed for $1.99 right now... It could be a way to get started cheap, with a circuit that works. Can't swear to the voltage though.

Cheers, Tom

OK, I measured the voltage: 4.0kV, with two reasonably fresh alkaline D cells. The circuit is apparently a blocking oscillator using a transformer and a single transistor, into a voltage multiplier. There's an 0.033uF cap at the output. "Don't touch the net while the red light is on" is an apt warning. The output cap is stressed rather far beyond its rated voltage...

You'd probably come pretty close to the desired 900V just directly off the transformer with a single half-wave rectifier. The transformer is roughly a cube half an inch on a side.

Cheers, Tom

Are you sure about that. It seems to me that 1KV AC on a pair of plates would radiate more than 1mA in a loop.

I attempted the opration at 1MHz to make the switcher be above the bandwidth of another part of the system. I ended up lower and well shielded.

What's the current through the capacitance of the plates? In other words, what's the capacitance between the plates? How large are the plates, and what's their separation? Anything else in the configuration I should consider? As Win suggested in his posting, it's still extremely unlikely that at 1MHz with an apparatus contained within, say, a cube a foot on a side (or two feet, or three feet) that you'll get anything like four times the radiation as you do I^2*R dissipation, short of using superconductors.

Cheers, Tom

A voltage tripler is often a good solution to high voltage low currents with extreme stepup ratios.

--
Many thanks,

Don Lancaster                          voice phone: (928)428-4073
Synergetics   3860 West First Street   Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml   email: don@tinaja.com

Please visit my GURU\'s LAIR web site at http://www.tinaja.com

volts.

losses

What started out as a not completely serious comment has gone way past where it should have, but to continue down this silly path:

The transformer was a toroid about 1 inch in diameter. The core was plastic. ie: This really was an air core transformer, but it was formed to limit the magnetic field it created. The turns ratio was IIRC 11:1 the winding was done with a wire rope and then connected in series. So the whole thing bulked up to maybe 1.25 inches in diameter.

The secondary side rectifier was a fast high voltage diode on about 1 inch leads. The primary switch was a highish voltage NPN in a TO-202 package. 1W went in, 0.5W came out and nothing seemed to get hot.