Using PWM to drive a ferrite transformer

Thread Starter

Peter McLeod

Joined Jul 3, 2020
I am in the process of designing a specialised welding machine that acts like a TIG in a vacuum. As in a lot of welders I have to create a short pulse of high voltage at an RF frequency. Because it is such a short pulse I can use a charged capacitor as the source and switch an IGBT with PWM. This is connected to the primary of a ferrite transformer. The primary is a single turn and the secondary is 12 turns. The secondary is taken to two electrode in a low vacuum with low pressure Argon gas.
My capacitor is 560uF charged to 500v and there is a 500mOhm resistor to limit the initial primary current. With 12 turns this should give me around 6Kv from the secondary, which is enough to create ionisation in the vacuum chamber. The capacitor will discharge in 0.84mS if no PWM.

My PWM will be 25 to 30Kz with around 5% duty. Which lengthens the discharge time to around 17mS.
I have tried to find out how plausible this is and mostly found that any problems will be due to the fact that the primary voltage is a square wave and could lead to saturation. However I am hoping that the short duration of the pulse might mean that the saturation will be minimal. (if it pulls the output down to 5Kv that would still cause ionisation)
The transformer is an ETD44/22/15 N87 SMPS Transformer with 0.5mm gapped core. (Farnell 2673494)

Can anyone help here please.