Other Parts Discussed in Thread: TPA3255,
G'day team.
I'm doing a large project using 3 x TPA3255 s which is a mains/portable public performance amplifier system. The 'portable' part is where I need a bit of assistance.
To achieve this I have to have a DC-DC converter and have decided to try the UCC38085 push pull controller chip as the foundation.
However.. due to the power needed of approx 1000w output this will need to be a rather grunty Converter and will be running off a 24v Lithium Iron battery pack.
THE PROBLEM.
In order to get peak currents of around 50amps in the primary I need a transformer to suit. (24v x 50a=1200w Primary Power)
I want to run the 38085 at 100khz using an ETD49 T/F with some paralleled mosfets (IRF3710 X 3 per side)
When I did some calculations for the L/R time constant, even with a 1uH Primary Inductance and 3 x IRF3710 in parallel (giving 7mR RDS) it seems the time constant
is too long for the current to rise to 50a
TC= L/R = .000001/.007=142uS. But operation at 100KHZ would give a 5uS half cycle time.
Am I barking up the wrong tree in this or missing something all too obvious?
I can decrease the "L" further by increasing the gap in the cores, but it just seems scary to go below 1uH for L Or..I can reduce the number of FETS (use 2 instead of 3.. giving an 80amp@100deg current capacity) which could increase 'R' and reducing the TC by about a 3rd, but it's still staggeringly long compared to my PWM window.
If I slow down the frequency, say to 50khz it would double the time window, but.. still seems tough to get a 50amp current rise in that time.
Any nudges in the right direction would be appreciated.
Julian
