Neat project!
A few questions if you feel up to them: How did you measure the tank's current? The multiplier's current? The multiplier's output voltage? What frequency is the transformer operating at? ZVS drivers I've seen in the past have a center tap. How is this one different? Is there a rough schematic of it somewhere? Does that teeny RF switch have an internal relay, or what? Have you checked that it can reliably handle the 15A DC?
One of the reasons I bought this ZVS unit is it uses Marko's revision of the Mazzilli ZVS circuit; two input inductors allow you to pull each end of the parallel LC with fets instead of pulling across alternating legs of a center tap. This means it can be used with two-terminal primary transformers. See:
http://i.imgur.com/Q3VqOhO.png
Tank current was measured with my scope's current transformer probe, and I blew the 10W burden resistor to pieces in about 1.5sec of connection, lmfao. Had to replace it after that so I'm not going to redo the measurement. It was in the ballpark of 150Apk. It was connected to the supplied air core primary for induction heating at the time of measurement, so that current measurement isn't really valid in flyback/transformer use like I have it set up now. Aside from when designing and assigning parts, knowing the circulating reactive current isn't really important.
Multiplier current is variable based on filament voltage (conductance/impedance of the tube). Same for multiplier voltage. However, there's a rough guesstimation we can do: the ac flyback isn't operating in flyback mode, as it is driven in ZVS with a resonant primary LC circuit. With 12V input tank voltage rises to ~30Vpk. The transformer's secondary is 3000turns, primary is 6 turns, resulting in a 500:1 step up ratio. 500 x 300 = 15kV. Empirical spark gap tests show this is within ballpark accuracy. 5 stages of CW Multiplier = 75kVpk ideal output. I apply a 5% reduction/loss per stage as a rule of thumb for a decently designed CWmult under nominal load (load the mult is designed for, in this case 2.5mA), so 75k x 0.75 = 56.25kV. Judging by the penetration ability of the resultant x-rays this looks to be ballpark accurate as well. Measured input power is 152W, 56.25kV @ 2.5mA = ~140W, so again within reason.
On this transformer it resonates at 40KHz. With the supplied air core primary it resonated around 102KHz.
RC Relay Module has an internal 30Vdc 10A / 240Vac 10A relay. I'm pulling a little under 16A through it, but my ON times are never more than 10sec, so I'm not too concerned with the thermal loading. I'm sure internal sparking from back EMF will wear out the contacts faster than normal, but I'll just whack a new relay on it when that happens. Before using it with the x-ray unit I gave it a solid testing on my bench switching 15A 15V loads into a dummy load (100W 1R resistor in water).
That's me well and truly told :crackup: , of course I used the quoted rating to differentiate between the two
Ok, Question (V2.0)
Could I use one of the cheaper ZVS's, or do you
have to use the one with the bigger tank?
it depends what you want to do really. If you're going to be driving a 150W load like I am you'll want as much power handling as you can get. I mean I could remove two rows of tank caps and it would resonate at a better (higher) frequency, but I'd be tripling the current flowing through the remaining capacitors, which would probably pop them. If you don't mind a center tapped primary ($#@%ing hate them for some things) you can use a standard Mazzilli ZVS driver which is cheaper and easier to find than Marko's revision. Still though, I think this driver with coil was only $40 shipped. Fiddy grabbed one and then alerted me about it, and I grabbed one too.
(not directed at the above two quoted gentlemen.....)
I do want to reiterate to anyone thinking of replicating this or doing anything with an x-ray tube other than using it as a desk lamp with the filament ONLY..... THIS IS INSANELY DANGEROUS. Don't try it!
I dug out my calibrated ion chamber today to do some Exposure Rate measurements... and boy fellas, let me tell you, I'm regretting sticking my hand in the beam path.
10 ROENTGEN PER HOUR EXPOSURE RATE.
Yeah, I only had a 4sec maximum exposure, but damn, if that doesn't scare your scrotal hair off I don't know what will. Granted that comes out to about 3mR/sec, so I probably only got about 12mR total dose to an extremity. It's been too many years since I studied health physics and I can't remember the conversion procedure for hand-only dose to body dose (but it isn't upscaling, so using face value is good'nuff). So I estimate I dosed myself with about 15mR total radiation to date. That's about 1000 hours worth of background radiation in five seconds.
Anyway I video recorded my exposure rate tests and I'll put the video up shortly.
By the way, the battery in my saturation-proof GM detector fell below the undervoltage limit, which causes false positive readout. This happened while I was in the room setting up after arming the x-ray unit and I shat some bricks as I ran away, lmfao. Lesson learned. I *really* want a good lead apron, too bad they're about $120. I just have to settle for ALARA practices.
.
