Test: Integrated Power Switches in HV Applications (fwd)

[Steven Roys <sroys@xxxxxxxx>]

---------- Forwarded message ----------
Date: Tue, 16 Mar 2004 20:10:59 +1030
From: Matthew Smith <matt@xxxxxxxxxx>
To: High Voltage list <hvlist@xxxxxxxxxx>
Subject: Test: Integrated Power Switches in HV Applications

Hi All

Some time ago, I acquired half a surplus rail (Ebay) of the Linear Technologies
LT1171 integrated power switches.  These devices include an SMPS controller and
power FET all in a single TO220 package.  The LT1171 switches up to 2.5A at
100kHz.  (1A and 5A devices are also available, as is a 40(?)kHz range.)

These devices appeared to me to offer a simple build, low component count way of
building switching power supplies, both for microcontrollers (5V from 12V) and
also low power (battery) high voltage power supplies.

In my "treasure" box, there was a ferrite transformer and voltage multiplier
module retrieved from a laser printer power supply.  This was supposed to derive
7kV from a 24V supply.  Would this work and give me a proportionally lower
voltage from battery power?

I built a PCB for this - only 7 components including the HV module, as no
feedback components were included.

Testing with 4 x AA NiMh cells (4.8V), my 5kV electrostatic voltmeter did not
budge although oscilloscope showed that the module was getting drive.

Adding another pack of four cells (9.6V), the meter registered a healthy 2.4kV.
    Foolishly, I forgot to record the current draw, but it was very low and I
calculated that the 2Ah cells would last a good few hours with the circuit
unloaded.  (Draw a spark and up the current goes!)

I have since installed the circuit in an old Post Office high voltage tester
which originally used a bulky flyback circuit, all built on perfboard.  This
device runs of a pair of the old-style 6V lantern batteries.  With this 12V
input, output is up to 2.8kV.

I was concerned that the ferrite used might not be suitable for 100kHz, but it
only gets vaguely warm.

I'm inclined to try the same circuit with a small TV flyback.  LT1171 is
internally current limited so almost impossible to "fry" unless one forgets to
connect the snubber...

Conclusion:  for low power HV applications, these are worth a look.  If starting
again, I would probably look to use the LT1070, which switches 5A at 40kHz.

If anyone else has used these or similar devices from other manufacturers in HV
applications, I'd be interested to hear results.

Cheers

M

-- 
Matthew Smith
Kadina Business Consultancy
South Australia
http://www.kbc.net.au