Rectification for homemade supply

[Ryan Catalano]

Is a filter capacitor required for the rectification system in a fusor power supply? Will the pulsing dc output in any way affect the fusion process?

[David D Speck MD]

Most fusion experimenters seem to agree that a filter cap is not required, though it seems more elegant to have one.

Big problem is that a filter cap will have a lot of stored charge on it which can bite you hard if you are not extremely careful. If you put a bleeder resistor on it, it can consume as much power as the fusion chamber itself, and you still cannot depend on a bleeder resistor for personal safety.

Also, caps rated at 30 - 40 kV are not particularly inexpensive or easy to find.

Dave

[Chris Bradley]

Un-smoothed DC looks like it might even be an improvement on capacitively filtered, because the peaks should hit higher voltages, and the fusion cross-section is better-than-linear at higher voltages, so what you lose out on in the low voltage phases of the cycle you should gain multi-fold in the high voltages... in theory ...

I think there is a thread here somewhere that looked at this precise issue. I seem to recall thinking there was nothing definitive in it either way, given the measurement uncertainties.

I wonder if there is even a question mark over whether you even need to rectify it, because if you run it with a glowing grid then the chamber itself will tend towards becoming a thermionic diode. It would be interesting to know what the relative currents to/from a fusor with a glowing grid and running AC is, between the +ve and -ve periods of the AC phase. Has anyone measured this?

But don't worry about DC ripple in the early stages of a project because ripple-free DC is not a pre-requisite, and especially if you are working with 50/60 Hz (needs a big capacitance to make any difference to ripple) then you skip the risk from heart-stopping levels of stored electrical energy.

[Dennis P Brown]

As someone not knowledgeable on electronics (but knows enough to be dangerous), I'd still like to add a few 'low tech' thoughts on this thread.

I do know that adding a capacitor to a rectified power supply is not required for the average fusor; however, adding one is often done to 'smooth out' the ripple when a diode is added to a AC power supply (half the sine wave is flipped to match the other half.)

Adding a very small (read too few farad's) capacitor achieves almost no effect in removing the ripple. So, when adding a capacitor, calculating the required capacitance (farad storage capacity) is important unless buying far more storage than called for - read very expensive method.

I will add, and my latest experience shows, adding caps to any safe power supply can create added risks - a HV power supply that is relatively 'safe' (read very low current) can easily be converted into a very dangerous supply – one that will instantly kill. Large value high voltage capacitors are not toys or should ever be added to a ‘safe’ system without extreme caution (add safety systems and relearn procedures.)

It is a good idea to consider a bleed resistor to quickly remove the charge on the capacitor when it is not being charged. As always, large, HV caps should be stored shorted so they do not self-charge.

[Richard Hull]

If possible full wave rectify!!!

After this, no cap is needed or required to do fusion. This is especially true in 60hz supplies as any truly suitable cap will store a good deal of energy and will motor boat pulses into the system of many kilowatts as you approach the delicate threshold of glow vs. no glow. This is rough on grids and diodes in the system. Leave the cap out of the circuit.

So, for 60hz line supplies, a full wave rectification will warrant 120 hz operation without too much stored energy except magnetic energy in the transformer core when to get to glow mode potentials. This is usually softened by the use of a 20 to 50 kilohm 50 watt wire wound resistor coupled with the fall to zero with each 120 hz alternation. (Such a resisitor is hard to locate today, except at hamfests.)

If you are using a modern switcher supply, (best of all worlds), there is little stored energy in its caps and the power supply usually has electronic, settable limiters built in.

I have used a full wave, non-capacitor filtered supply since 2004 and can readily hit the mega-neutron mark with 55kv 120hz peaks with about 40kv rms indicated on my metering.

Richard Hull