[Prev][Next][Index][Thread]

Re: Solid State - air core driver??



On Thu, 25 Apr 1996, Tesla List wrote:
> 3. Tricky: build a half bridge or a full bridge to drive the primary.

Not tricky. Just drive the fets with transformers and different
source voltage is no harm. Bridge type setups are much more forgiving
becouse they clamp the voltage to the rails.

Push-pull will nead current balancing. If not it will give you trouble,
especially with saturable transformers. Air-transformers are fine but
huge.

> 4. more promising: abandon the primary for bottom feeding. I was
> going to use a ferrite transformer but Malcolm said:
> >I hope this helps. I really think an air-cored transformer for the
> >driver is a better option as power handling is only limited by the
> >wire size used (you can use it as is for operation at any power). 
> >Also, you won't have to worry about core saturation caused by 
> >imbalances in driving.

I agree with Malcom on the good points. However, you must consider
the primary inductance in parallel with the load reflected at the
turns ratio. That means you will need lots of inductance in the
primary. Something like 1-2mH is about the mininum for 200kHz operation.
(Calculate what impedance it gives at the drive frequency! Then calculate 
how much such impedance is drawing power!)

Saturation of the core is no problem. You just must wind a proper
transformer for a proper core. Then you need to remove any imbalance
probably. The cores will not be huge as some seem to think. I used a
core about 1"x2"x2" in size and did put around 1kW through it. A few of
those will make a nice transformer at low cost for low power.

> is it as simple as getting a toilet roll centre,
> winding on some thick wires to get the primary up to say, 3 or
> 4 times the minimum induction given by V=L di/dt.

Nope. Make the inductance much larger than load inductance. Then it
is correct. You are not wanting to circulate power in the primary,
you want it to the secundary!

> Using the solenoid equation to estimate the induction, and then
> wind on 10 or 20 times as many turns for the secondary.
> What kind of efficency can I expect.

Pretty poor. As a rule of thumb take around 2mH primary for 200kHz,
2mH for 100kHz and so on. Then use the turns ratio as the required
voltage step-up needs. (Yes 2mH with air core will be huge. Thats
why eg. switch mode PSU's are using ferrites!)
--
We have phone numbers already, why would we need IP-numbers! -unknown person

Harri.Suomalainen-at-hut.fi - PGP key available by fingering haba-at-alpha.hut.fi