2

I need a DC 4.2Kv 300ma power supply, and am looking for advice.

I'm currently using a the guts of a microwave oven (transformer/diode/capacitor) to power a magnetron for a physics experiment. The transformer outputs 2.1kv into a diode/capacitor voltage doubler to produces a 50% duty cycle 4.2KV potential at 60hz. The resulting signal from the magnetron is noisier than an LA freeway at rush hour.

According to this paper changing the unit to a regulated HV power supply, controlling the filament heater separately, and reducing the filament current after it reaches operating conditions should allow it to phase lock onto an external signal and operate as a low noise amplifier.

For the HV side, my first instinct was to buy a commercial supply. I priced them, and after regaining consciousness thought I should ask here.

Question: Is there a way to smooth out the power from my existing supply or another inexpensive alternative I should consider?

Thanks in advance.

(Amateur: yes. Kid/teenager: no. Experience in electronics: Kit-building. Follows safety procedures/common sense: yes...Safety advice is always appreciated.)

Added additional details on safety and end use to comments.

ElizabethGreene
  • 161
  • 3
  • 2
    Safety question: Are you experimenting under some sort of professional supervision? Otherwise it sounds risky. At least let professional to examine your setup. – Eugene Sh. Sep 04 '15 at 18:23
  • 2
    To reiterate @Eugene: Microwave magnetrons are incredibly, mind-bogglingly dangerous, insofar as I've heard. In fact, there's a saying I've heard: "Don't attempt any project that involves a magnetron". If anything goes wrong, you could easily end up with clouded and boiled corneas and all manner of disturbing injuries. Oh, and the magnetron itself contains easily exposed highly toxic beyrillium. Please do ensure that everything is checked by a professional before you attempt this, as Eugene says. – 0xDBFB7 Sep 04 '15 at 19:09
  • 1
    I am always fearful when I see a question like this. People can cause so much damage with a magnetron if they don't know what they're doing. I was relieved to see the answer that you posted [here](http://electronics.stackexchange.com/a/189007/38335). Your knowledge and attitude are reassuring :) One example I remember was a guy who had cut a 4" hole in the front and back of his microwave oven so he and his dad could dry lumber by passing it through... After doing this for a while, he realized it might be dangerous and asked about it online!! – bitsmack Sep 04 '15 at 20:23
  • Hmm, I had a look at the paper you posted on DropBox and after reading the introduction my bull-shit-o-meter starts to go crazy. Perhaps it is all legit but then again to transmit power the noise figure is unlikely to matter that much, the filament being 'back heated' if even possible is unlikely to have much effect. A whole genre of these 'wonderful' inventions relies on them being difficult to realise while people gaze at the designs with starry eyes. Even if you could bean ALL the power to just one 'collector' You would get a max of 750W of RF energy that will cook anything in the path... – KalleMP Sep 04 '15 at 20:51
  • 1
    ... also note the list of references used, just one by some other author. I would be interested if the next issue of this periodical included a retraction or at least some interesting comments from other readers. The Beryllium danger is probably over hyped by those that don't understand it. The ceramic insulating parts may be made of beryllium oxide and fine dust liberated by grinding these can be dangerous, the HV stored in one of those caps may kill you a lot faster. Take care. – KalleMP Sep 04 '15 at 20:56
  • Thanks for the concern. My magnetron does not have the beryllium insulator, so no risk with that. The magnetron is firing into a fully enclosed resonator inside a faraday cage. I have LED/shottky microwave indicators inside the cage, and a leakage indicator with an audible alarm outside the cage. I enclosed the power supply in a lexan box because cats, and I put a key switch on it because kids. – ElizabethGreene Sep 05 '15 at 01:10
  • The most dangerous part I see is the HV terminals on my the magnetron. I intend to cover those with "something", but right now they are the recessed rubber covered oem style spade terminals. I have to pull the magnetron soon to add water cooling, so it's a good time to make changes if you have recommendations. – ElizabethGreene Sep 05 '15 at 01:24
  • KalleMP it's much worse than 750 watts. My resonator has a Q of over 1000 at the peak frequency. If the resonator blew apart it would release many many times more power than that. As I do not want to die, I take safety seriously. The high Q correlates to narrow bandwidth. That's why I'm trying to clean up the spectrum. I want to pack more energy into the center of that band and less lost as heat. – ElizabethGreene Sep 05 '15 at 01:35
  • You question asks for a 1.2kW power supply, you can expect 50% max efficiency from the typical set-up. Sadly you cannot expect more than 100%. Methinks the lady doth troll too much (I may be wrong in which case I apologise for dismissing your otherwise earnest learning attempts). I am also interested to know how you tested the ceramic insulator composition on your magnetron. – KalleMP Sep 05 '15 at 17:30
  • @KalleMP I explained that poorly, my apologies. A microwave resonator stores and amplifies microwaves by trapping them inside a closed cavity. With no way to escape, the microwaves form standing waves. A resonator cannot store energy forever. Energy is lost as the electrical resistance of the cavity walls converts the energy from electromagnetic waves to heat. How efficiently the cavity can store the energy is referred to as the Q factor. Q varies by frequency. My cavity has a measured Q peak of 1100. Q is defined as 2pi * energy stored/energy dissipated. – ElizabethGreene Sep 05 '15 at 18:10
  • When the cavity is "full", the energy dissipated equals the input power. That's equilibrium. Then the stored energy equals (input power * coupling loss coefficient) * Q / 2pi. Assuming 40% coupling efficiency, my cavity is holds ~70 times the input power when fully charged. I currently have a magnetron rated for 600 Watts of input power, but it only receives that power 50% of the time because of the power supply. 16 ms on, 16ms off. 16ms on, 16 ms off. With this question, I am trying to build, buy, or beg a continuous output power supply to see if that changes my experiment results. – ElizabethGreene Sep 05 '15 at 18:22
  • Let us [continue this discussion in chat](http://chat.stackexchange.com/rooms/28820/discussion-between-kallemp-and-elizabethgreene). – KalleMP Sep 05 '15 at 18:23
  • If it doesn't, then my next step is to try and narrow the input frequency spectrum. That is the subject of Mr. Clark's paper. In hindsight, mentioning that confused the issue significantly and and that was a mistake. I thought that by mentioning the desired use for the kit I might get this question more attention than the normal "don't play with microwaves, kids" response. Sadly, I have only gotten the boilerplate answer so far. – ElizabethGreene Sep 05 '15 at 18:27

1 Answers1

1

The simplest solution to this problem is to replace the existing "Villard Circuit" pulse-type voltage doubler with a voltage doubler circuit that outputs solid DC. This modification is significantly less expensive than a new power supply. It requires only two appropriately rated diodes and two capacitors.

Two safety related modifications are also suggested. First, a small bleeder resistor should be connected across each capacitor to drain away charge when the device is not in use. This will reduce (but not eliminate) the possibility of an accidental discharge.

Second, a fuse or circuit breaker should be installed if not already in place.

Finally, to round out the updated power supply, consider adding a cooling fan and voltage/current meters.

Special thanks to @kalleMp for inspiring this answer. I am in their debt.

ElizabethGreene
  • 161
  • 3
  • 1
    Could you diagram the difference between a "Villard" doubler and whatever you found that you consider a "solid-DC-output" doubler? – feetwet Jun 11 '16 at 21:11