Thanks for your advices,
K.
Flyback, either directly or through a step-up transformer.
If you have lots of money and not much time check out Apex amplifiers
(they got bought a couple of years ago, I can't remember by whom).
That'll give you a basic, ordinary, op-amp that just happens to work at
your voltage range. It saves you having to roll your own.
If you want to do it without much effort get a tube amplifier line
transformer. Hammond or similar. Or an isolation transformer that
presents 230V at the secondary. Then use a voltage doubler and two
electrolytic caps, like this:
http://en.wikipedia.org/wiki/File:Bridge_voltage_doubler.svg
Depending on the transformer you'll end up with 600-800V. There will be
ripple on it so if your amp has a poor power supply rejection ratio
(PSRR) you may need a regulator behind it. If you use a 230V iso
transformer you'll need larger electrolytics so your voltage doesn't dip
below 500V.
If you need some offset just use a 2nd setup like this and add the
voltage at the bottom. But (important!): Mind the limited isolation
breakdown voltage on the secondaries, there comes a point where it'll go
bzzzt ... *PHUT*
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Apex (bought by Cirrus) was my first option, but when I saw their crazy
prices I have decided to evaluate other cheaper solutions...
Thanks to you and Joerg for your advices.
Bye
K.
>I am working on an amplifier for driving piezo stacks and I
>need a sinusoidal voltage of 500 V peak with mean value of 500
>V. To keep things "quite simple" I'll work at slow frequencies,
>so that the current in the stack (capacitance of 1 uF) is
>"reasonable", say it max 0f 50 mA. As a first attempt, I am
>evaluating the circuit proposed in AoE p. 169, which has been
>suggested in this ng in the past for similar applications.
>This amplifier (or other equivalent designs) employs a 1kV
>power supply for the power mos, so my question is: how can I
>realize 1 kV power supply for this kind af amplifier (actually
>I need 3 amplifiers)?Rising transformers?Multiplying
>circuits(diodes+caps)?
I have not designed any high voltage devices but I have _used_ piezo
actuators and would like to make a point here, from a user's
perspective. This assumes you are making a product which is going to be
sold to other people.
We designed in some piezo actuators which used about 100V. Then we tried
to get the equipment certified to the generic industrial safety
standard, IEC61010. The biggest problem we had was with these *******
piezo actuators and their drivers, because the voltages were about the
magic 42V dc (30V ac) figure. The company who designed them and their
drivers seemed totally unaware that there are safety regulations to obey
as soon as you exceed these voltages. They used bare enamelled copper
wire to the piezos (because they were designed for a vacuum). IEC61010
counts enamel as "no insulation at all". The high voltage wiring needs
to be adequately insulated. If it is in a vacuum use PTFE insulation.
And make it something easy to solder, and colour coded. Inside your
amplifier, make sure there is clear segregation / optoisolation /
protection. The IEC61010 tests will assume your 500V can appear across
the safety components, so they will need to be adequately rated.
You might also like to look at Linear Tech application note 118:
http://www.linear.com/pc/downloadDocument.do?id=26421
--
Nemo
Me? I'd go the old-fashioned way, step-up line transformer, then
rectify.
The I'd use... are you ready for this... vacuum tubes ("valves" on the
other side of the pond) for the amplifiers.
Everyone always wants exotic (and failure prone).
...Jim Thompson
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
You can never be too prepared for the REPRESSION!
Ye olde 6KD6 would be a candidate here:
http://www.nj7p.org/Tube4.php?tube=6KD6
> ... ("valves" on the other side of the pond) for the amplifiers.
>
Or "lampes" where Fred lives :-)
> Everyone always wants exotic (and failure prone).
>
Nah, done it transistorized, many times. What are you afraid of could go
wrong?
Under five bucks, meaning even less than a Russian tube:
http://ixdev.ixys.com/DataSheet/99036.pdf
Short a tube, get a flash... no damage. Now do that to a
semiconductor ;-)
Thou shalt not build a contraption this is not protected against the
evil forces of a pipe wrench falling onto the output terminals :-)
As long as the wrench falls onto the terminals but not onto the amp
itself the tubes are fine I suppose...
robert
Real designers don't use pipe wrenches around exposed electronics.
--
Offworld checks no longer accepted!
Designers don't but installers and maintenance folks do :-)
One episode: "Wow, how did that happen?" ... "Well, we have this 5-ton
truck and ..."