Instability between a 16A triac 'dimmer' and 3300uF caps.

[Doug Coulter]

This book, ISBN 0-07-052236-7, will save you more in fried parts than the money it'll cost you.

Good book on switching supplies

This wheel is already round - take it from there, you'll have more fun.

[Jerry]

You know you can't turn off a triac or SCR once it's on

I don't really know much about those. I was planning just to use FETs or IGBTs, with some sort of comparator/logic circuitry. Not planned it yet, but I've got a warm fuzzy feeling it's something I'm likely to be able to accomplish.

So, this is just an isolated variable DC power supply you are trying to build?

[chrismb]

Yes, one that is 16A capable and, ultimately, practice a topology that will allow me to control 32A (I have a transformer twice the size of this to make work, if I figure out a simple way to do this).

[Jerry]

Then your best bet is a variac on the from end of the transformer. Otherwise you are going to have to scale up something existing. 0-300v itself is no big deal. Handling 5kw and ultimately 10kw is asking for a lot. That is some serious power. Plus you will need some sort of short and overload protection or it could get very interesting. If you use a circuit breaker on the output it needs to be rated for DC. At this voltage most circuit breakers will just draw an arc instead of cutting power so you need a breaker with magnetic blow-out.

So, variac -> isolation -> rectifier -> filter cap -> dc circuit breaker is really going to be your only option.

[chrismb]

I have a 6A variac and will be using that for fine control to achieve lower voltages for currents within that capability.

But to find a fit and working >10A variac here in UK, with the money I'm prepared to put into this, is highly unlikely.

Ultimately, I do not need a particularly steady supply, just one that will work predictably with a reasonable amount of ripple, because I am going to run a 100kHz switching supply on the other side of this that will give me a more precise, high current DC I am working towards. This will regulate away these fluctuations, by PWM regulation.

[Starfire]

Use a variac - it gets round inductive and impedance issues - welding transformers are two phase

[chrismb]

welding transformers are two phase

I've given up trying to work out what people call 'two phase' John. It says '2P' on my site transformer, but I'm told that stands for 'pole'. It seems illogical to me to call anything other than 180 deg phase separation '2-phase', but I've found most folks don't let a trivial matter like 'what is logical' get in the way of naming conventions!!!

[Jerry]

Thats is not two phase, that means two poles.

[Jerry]

Use a variac - it gets round inductive and impedance issues - welding transformers are two phase

No, they are not. You cannot get two phase out of single phase. The accepted definition for two phase power is two legs (poles) 90 degrees out of phase with each other. Two phase comes in two configurations, three or four wire. Though three is pretty rare. Three wire is used on the little TMP50 turbo pumps.

Heck, wiki has a page on it.

http://en.wikipedia.org/wiki/Two-phase_electric_power

[chrismb]

The accepted definition for two phase power is two legs (poles) 90 degrees out of phase with each other.

Yup, I agree with you that this is the accepted definition. Still, it is two lines of what would be a four phase system if all the phases were provided for.

One to 'two phase' conversion [as in not what is called two phase, but what is 'two phase'] is very simple as a second, mirrored, phase can be accomplished with a second winding that provides the phase in the opposite sense, at 180 degree. There is a difference between the accepted definition of two phase, and what is what you might describe as 'theoretically two phase' (the latter being 180 degree separation, because 360/2 = 180!!].