1. DC-DC Buck Converter in Inner Detector Environment
J Kierstead BNL
Environment
Radiation 75 Mrad, 2 x MeV neutron equivalent
Magnetic field 2 Tesla
Limited volume
Implications
No magnetic materials
Simple Design (limited number of parts)
Buck Converter?

2. Buck Converter Strengths and Weaknesses
Efficiency
Failure modes are well understood and easy to isolate failures
Requires higher voltage rated components making more susceptible to radiation damage and other failures
Noise

3. Buck Converter Simple Buck Converter
Magnetic field requires that we use air core inductors
Vin = 20 V, Vout = 2V, Ipeak= 1A
Lmin = (1.8/fswitch) henries
If fswitch = 1MHz
Lmin = 1.8 x 10-6 henries
Lmin = (Vin – Vout)ton / Ipeak
ton = Vout/ (fswitch*Vin)
fswitch = switching frequency

4. Practical Air Core Inductor
Within a factor of 2 or so
L = 10-6 N2r henries, r is radius in cm
To get L = 1.8 x 10-6 henries
for 1 MHz switching
If cm < r < 0.4cm
Then 30 > N > 22
Length of Cu wire between
35 – 55 cm with diameter of
0.3 – 0.4 mm.
Length of coil = 1 – 2 cm.
COIL IS TOO BIG
DC resistance about
0.1 ohm giving voltage
drop of 0.1 volt

5. Practical Air Core Inductor cont.
Solution: Increase switching frequency to > 10 MHz
Lmin = 1.8 x 10-7 henries
Then If cm < r < 0.4cm > N >7
Length of Cu wire = cm, diameter 0.18 – 0.2 mm
wire, Length of coil = 1.5 – 2 mm
MUCH BETTER
But now we are moving into frequencies where we will have more trouble with the switch(es).

6. Mosfet Switch Issues
Mosfets suitable for switching 20 volts are not available in rad-hard technologies (at least not to 75 Mrad) ?
Switching losses are proportional to switching frequency.
Mosfets and pulse width modulators for switching at 10 MHz and greater are not readily available

7. Gamma Irradiation Test 1.5 Volt VTM
Conditions:
Dose rate 1.4 x 104 rad/hr
1.5 Volt, 90 Amp
VTM powered with commercial supply
Fan cooled
Voltage and current monitored on input and output
Device failed between 50 and 60 krad
Switching MOSFET(30 volt) tested independently on this device and failed between 3-4 Mrad.
Control ASIC is probably what failed in the VTM at 50 krad
Failure of the device was abrupt and not observed.

8. Conclusions
Switching regulators have significant technological challenges when used in this environment
No magnetic materials allowed implying air coil inductors/transformers
At low frequencies (<1 MHz) coil is too large
At high frequencies(> 10 MHz) efficiency decreases due to switching losses and parts not readily available
No known commercial switches for radiation environment