Designing a variable HT PSU / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Designing a variable HT PSU

1 2 3 4 next last

Move Thread

LAN_403

Adam Munich

Wed Dec 01 2010, 02:24AM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

So for my x-ray machine I'll need a battery powered high tension PSU. Since the lead box weighs 25lbs, the wooden box will likely weigh another ten, weight is starting to become an issue. Since Li-ions are way out of my price range I'm going to need to use an AGM battery. I have a 12V 6.5Ah (about 8 lbs) one that needs a home, so I'd figure I'd use that. It will probably last for about 10 minutes of continuous x-raying, which is plenty of time.

Essentially what my PSU needs to do is;

A) Supply 35 to 70kV (variable)
B) Supply at least 3mA of current.
C) Not sag under the load of the x-ray tube (important for consistency)
D) Be unipolar (one +HV output)
E) Operate on 12V DC.

So, any ideas? I really don't want to spend more than 70 dollars on this.

Ash Small

Wed Dec 01 2010, 06:29AM

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

I'm really a beginner here, but why can't you use a ZVS driven flyback with a multiplier?

Will it not run off 12V?

(Sorry if this is a stupid question)

Adam Munich

Wed Dec 01 2010, 06:37AM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

It would, but there are a few problems.

A) It's going to be expensive
B) In order to adjust the kVp, I would have to adjust the voltage that goes into the ZVS. The only way I can really do that is with a resistive voltage divider, but large rotary rheostats are really expensive. I don't even know if a ZVS will work at say... 4V. Since it won't be continuous duty I could probably get away with a 100 watt rheostat, but even that is expensive.
C) Another problem with a CW is that they tend to sag under load, and an x-ray tube is quite a hefty load.
D) Finding a suitable AC flyback is a big job itself.

Ash Small

Wed Dec 01 2010, 08:07AM

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

OK, I'd imagine that the ZVS circuits that use two voltage inputs would work as long as you have 12V to drive it.

Can the output voltage be stepped, or does it have to be continuously variable? Could you not switch resistors in and out?

(This wouldn't overcome the problem of sag, Which I don't fully understand anyway. can you explain a bit more?)

I'm interested in this as I'm concidering building a similar circuit myself.

radhoo

Wed Dec 01 2010, 09:58AM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

I built a variable 50KV supply (can be extended):
- one LM317 + high power transistors to adjust the low voltage 0..30V (schematics here:

- a ZVS driver and a flyback chosen that the discharge starts at minimum 1 cm (some of my flybacks are giving very low voltages only, you should start with at least 10kv) . Funny this was a low power flyback.
- the multiplier, to reduce the ripple effect I've split it in two parts, a negative multiplier and a positive multiplier.

Multiplier details here:

Freitsu

Wed Dec 01 2010, 10:50AM

Registered Member #3147 Joined: Sun Aug 29 2010, 10:53AM
Location: Finland
Posts: 56

A little hijack, sorry Grenadier.

I didn't think this update that I'm going to make deserves it's own thread.

I decided to finally try to construct an X-ray PSU that will put out 60-70kV, this time I'm going to try a CW multiplier driven by a quite new ac flyback that I found at the recycling station (it's good to have connections

)

This is the HV diodes that I'm going to use:

And the caps:

And 2 of these resistors in series to limit the current:

Adam Munich

Wed Dec 01 2010, 11:26AM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

@ash

When you put a load on a CW, the voltage sags.

@radhoo
Nice, but it doesn't look like there's much current there.

@Freitsu
Have fun with that!

What I need is one of those high frequency x-ray transformers, one with an open core. That way I could just build a ZVS or an SG3525 based driver and be done with it.

Wolfram

Wed Dec 01 2010, 12:32PM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

Grenadier wrote ...

@ash

When you put a load on a CW, the voltage sags.

When you put a load on any supply, the voltage sags, unless the supply has feedback to compensate for this. As I see it you have three options.

1. Make a relatively stiff supply that doesn't sag very much under load, so that the sag is negligible. This means an over-designed voltage multiplier with much more expensive components than needed, as voltage multipliers tend to have inherently high output impedance. High frequency transformers for such high voltages are not practical here, as the enormous inductance and self-capacitance for a very-high-voltage secondary would give it a very low self resonance frequency. And while you might be able to make a flyback output 70kV, having it deliver 70kV at 3mA while being reliable is not realistic. Summary: transformer and multiplier is the best alternative, but multiplier will need to be over-designed to get low enough sag.

2. Same as 1, but use a voltmeter on the anode voltage and adjust transformer voltage to get right tube voltage under load. Should work fine as long as the tube draws a constant current, which x-ray tubes tend to do.

3. Use a voltage divider to measure the HV, and have your circuit adjust transformer power to stabilize the voltage. This is how it's done in all high frequency x-ray supplies I've seen, and this is by far the best way to do it.

What I need is one of those high frequency x-ray transformers, one with an open core. That way I could just build a ZVS or an SG3525 based driver and be done with it.

What do you mean by open core? All high frequency x-ray supplies I've seen use an intermediate-voltage transformer and a CW voltage multiplier.

I think the simplest approach would be to use a 60Hz x-ray transformer and a cheap 12V-110V AC inverter modified for variable output voltage in your case.

radhoo

Wed Dec 01 2010, 01:06PM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

Grenadier wrote ...

@radhoo
Nice, but it doesn't look like there's much current there.

Probably you're talking about the film. If you look carefully there's the TV HV Resistor to limit the current.
limited:

unlimited:

Without that, the sparks are loud and bright, but I'm not planning to ruin the supply by drawing arcs from it.

On the other hand, Anders M. is right. If you take a look at the Glassman modules I've posted here:

You will see they use the same CW multiplier but also have a divider to control the output. So the driver running the flyback can always compensate (to some level) the voltage drop on load.

You didn't say, why do you need 3mA anyway? Do you have a X-ray tube requiring this current?

Adam Munich

Wed Dec 01 2010, 11:35PM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

@anders
I remember seeing a pic of the insides of one early machine that had no CW, just a "giant flyback." They do exist, but they are most likely a very rare item.

I could go for the 60hz + inverter method, but I'm trying to make this as light as possible. The lead box already weighs 25 lbs, I'll have an 8lb battery in there, and the wooden case would probably weigh another 15 lbs if I make it out of half inch ply. A 60hz transformer and a variac would be heavier than I'd like.

@radhoo
I just bought a Coolidge tube from xray that has a maximum of 7mA @ 70kv, but that is too much for my uses. I figured 3mA would be good.

I think I'm going to try an ignition coil based circuit before I try a CW, mostly because I can get a coil at a junkyard a lot cheaper than I can get HV caps and diodes. I was thinking of using a ZVS to step up the voltage to 120 like I did before, then feeding that into the ignition coil. I could also just go for the direct approach with a mosfet chopping up the current from the battery, that shouldn't be too hard to get working.

I'll figure this out somehow.

1 2 3 4 next last

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