Driver for F6T5

[burhanmz]

I need some help with this driver for driving F6T5 CFL bulbs.

See attached figure (mostly made in Fritzing software except for the T1, the first one I saw on my desktop)
These are 4 F6T5/BLB (black UV light blue).
My power source is from a step down transformer to bridge rectifier to smoothing capacitor ending up with 13v (cap = 2200uF/50v , bridge rectifier 20A , transformer 220v/15v @ 4A secondary so basically 60W) all safe so far
Astable mode 7555 (cmos version)
R1 and R2 = 5.6K ohms (couldn't find those in fritzing )
freq = 859KHz
Duty cycle 67%
IRF540 configured in saturation mode (for switching - I checked datasheet and IRF540 can tolerate up to 100ns pulses, so I quite safe)
T1 is ferrite core transformer (so it can handle the high freq) 0.6ohm/10ohm

The circuit works but with two problems
1. the MOSFET gets hot
2. only one of the F6T5 lights up

So I can individually light up all four F6T5s but together only goes ON... I feel its got something to do with standing waves (since when I connect all of them together - only the right most lits up; the one furthest from the circuit)

And I can not understand why the MOSFET gets hot. Its rated for more than 20A continuous.
with duty cycle 67% the Vgs = 0.67 x 13v = 8.7v
VDS= 13v

Can someone help me point out what I am missing here?

 

[ronsimpson]

And I can not understand why the MOSFET gets hot. Its rated for more than 20A continuous.

With a big heat sink it can handle 20A.

with duty cycle 67% the Vgs = 0.67 x 13v = 8.7v

???????

VDS= 13v

12V on schematic and 13V in real life.

1. the MOSFET gets hot 2. only one of the F6T5 lights up

1. Yes it does. A heat sink will help. ......2. You are not sending enough power to light 4 bulbs.

freq = 859KHz

Really? Can the 555 run that fast? The MOSFET will get hot in part because this is too fast.

T1 is ferrite core transformer (so it can handle the high freq) 0.6ohm/10ohm

did you make the transformer? Do you know if the transformer is in saturation?

 

[burhanmz]

And I can not understand why the MOSFET gets hot. Its rated for more than 20A continuous.

With a big heat sink it can handle 20A.

I have got the MOSFET on a heat sink, this one from digikey
https://www.digikey.com/product-detail/en/WA-T220-101E/WA-T220-101E-ND/2202817

with duty cycle 67% the Vgs = 0.67 x 13v = 8.7v

???????

the out of timer directly feeds to gate of MOSFET, and the out of timer is a function of the duty cycle percentage times power source (so 8.7v but I measure 8.2 at gate to ground)

VDS= 13v

12V on schematic and 13V in real life.

yeah sorry about that, the power source is about 13.5v, couldn't find it in Fritzing, so used 12v to display it - But VDS is about 13v and across the T1 about 0.5v

1. the MOSFET gets hot 2. only one of the F6T5 lights up

1. Yes it does. A heat sink will help. ......2. You are not sending enough power to light 4 bulbs.

yes but it shouldn't get hot at all! as long as its inside the saturation range (I mean its hardly taking 1A, and thats next to nothing)
how exactly can I increase the power? I thought using a MOSFET the sort of IRF540 would take care of it - since I have more than ample power source

freq = 859KHz

Really? Can the 555 run that fast? The MOSFET will get hot in part because this is too fast.

I might be wrong, but I have specifically read that CMOS 555 timers (aka 7555) are rated for operations upto 3.5MHz and some datasheets (the newer SG electronics) claim 10MHz at 5v to 15v

T1 is ferrite core transformer (so it can handle the high freq) 0.6ohm/10ohm​

did you make the transformer? Do you know if the transformer is in saturation?

I salvaged this transformer from an old 40W energy saver bulb - which I found was working at 150khz / plus I googled to confirm what sort of transformers would support such high frequencies which lead me to my scavenger hunt for it in right place -
transformers in saturation start to spark at such high frequency at the terminal ends ... mine does it at about 1MHz, which is why I choose a slightly lower frequency