Sources for high voltage power supplies?

[Michael Noone]

"Chris" <cfoley1064@yahoo.com> wrote in news:1111420964.293831.31780
@g14g2000cwa.googlegroups.com:

Hi. You've got yourself a passel of problems here. Let's take 'em
one
at a time.

1) SAFETY. When you're dealing with high voltage DC, you're talking
about potentially killing someone, either yourself, one of the
researchers, or an operator or someone else who wanders in and touches
the wrong thing. Before anything else, when you're developing a high
voltage fixture for R&D, you have to build a reverse mousetrap to
prevent someone who's smart but distracted from accidentally
electrocuting themselves. For something small, this could be as
simple
as a small Lexan box with a lexan door with magnetic latch and
microswitch safety. The idea is to immediately neuter the power
supply
in the event someone could conceivably get buzzed off. Not all that
difficult if you use your head.

I have suggested that everything be fully insulated - but it remains to
be seen whether or not they take my advice. Certainly the circuits I'm
building for this beast will all be fully protected like you have
advised.

The thing is, you should have done this first, and now you're under
pressure to produce something. Well, you have to DO THIS FIRST. If
you're out of time, do it first. If they've promised to let you go if
you don't give them something immediately, do this first and let them
fire you. There are more important things than a job, and the next
guy
will have a head start. And besides, it shows you're not just
slacking
off, and are willing to work toward a solution. That may be in
question at this point, which is just about the worst situation you
can
be in.

Oh - don't worry about my job security I'd take a lot to get fired -
and even if I did it wouldn't be *that* big of a deal... Mostly I'll
just get chewed out if they think I'm being too slow.

2) HV KNOWLEDGE. This part is a little tricky. Apparently you don't
have much experience here. That can be a literally fatal problem. It
might be better to ask around, and talk to somebody who's familiar
with
HV. If no one's available, find an old EE or engineering tech who's
familiar with the art, and have him teach you. It won't take long.
If
you're in a city of any size, you'll probably be able to find someone
without too much problem. If you need it right away and can't find
anyone, call one of the temp services and ask for the services of
someone who's familiar with HV fixturing. He can teach you a lot in a
couple of hours. If you happen to be in the Chicago area, use my
email
-- I'll be happy to kibbitz (for a reasonable fee ;-)

I'll try to do some substantial research about HV. I'm an EE undergrad
student and we haven't even mentioned HV design yet. Everything we've
covered so far is about small signals.

3) PEOPLE PROBLEM. I think Don Lancaster once said something like,
"For any job there's the technical problem and the people problem."
Believe it or not, the people problem is every bit as important as the
technical issues most of the time, and in this case, it's become the
predominant factor. You're dealing with researchers here. They've
admitted that the electronics end of the project is outside of their
sphere of competence -- that's what you're there for. Also, their job
is to research. If they knew exactly what to do, they wouldn't need
you. An ever-changing power supply requirement isn't a fatal problem
here. Being in a "chicken-and-egg" hangup is, and you're the one
that's going to look bad.

OK, some practical advice. You've got to walk in with a plan, and
take
charge here. Your "extemporizing" time has pretty much evaporated, I
guess. Here's what you might do:

* Stop at the local hardware store. Buy some thick plexiglas,
acryllic cement, and aluminum angle stock, along with a selection of
6-32 and 8-32 screws, nuts and lock washers. Start building your
safety enclosure immediately. Farm out some of the machining if you
can -- it will save time.

* Once that's done, your first job is to give these guys something to
work with so they can find some answers to give you. Start out by
renting a HV DC power supply for a week -- that should cost less than
$200 USD plus shipping. Spec a good one with an emergency shutdown
circuit, and be sure to wire your reverse mousetrap safety microswitch
into the circuit before you let 'em at it. Let them play, and be
creative and curious. Watch what they're doing. Help to keep them
moving. Ask good questions:

- Do you have an idea of voltage requirements yet?

I just e-mailed them again to see if they had decided on anything - I'll
get back to you on that.

- How about maximum current? (If this is an electrostatic
application, you probably won't need a tenth of the current you're
talking about. But you'll never know until you plug something in and
check it out. It's your job to measure things.)

They want a max output of 20ma on each of the 10 outputs - so .2A total
max output, though in reality it should be quite a good deal less than
that.

- How much ripple voltage(many electrostatic applications go nuts
with
a superimposed AC ripple)?

I don't think there will be any ripple - but the load will be changing,
if that will be a problem

- Do they need soft start (voltage ramps up to value)?

Nope.

- How about arcing on shutdown?

I'm not quite sure what you mean here - what would be arcing?

- Do they need operational control over the power supply? If so,
what
bandwidth? Or can you "make do" with a couple of relays if you need a
bidirectional supply?

What exactly do you mean by operational control? As far as I know -
initially they'll need to set the voltage - and after that they'll just
need to turn it on and off.

Once you get good answers, you can start to look at whether a "home
brew" job will be good enough, or if not, how much you're going to
need
to spend on a store-bought power supply.
By the way, working on the assumption that this is an electrostatic
application, I'd like to second Mr. Hill's suggestion to look at the
Spellman HV catalog, if for no other reason than to learn about what
questions to ask. They're good folk to know if you have HV problems.
One of their technical resource articles in particular may be of some
interest:

High Voltage Power Supplies for Electrostatic Applications
http://www.spellmanhv.com/tech/article_detail.asp?id=8

This will help you in the process of specifying the power supply, and
lists some of the "gotchas" you'll have to watch for. Don't overlook
the possibility of just giving them a call.They do have apps engineers
to help.

I'll have a look through it. Thanks for the link.

Most importantly, work on the relationships here, and give them
reasons
to have confidence in you. You're a force multiplier, and have been
put on the job to make them more effective. If you look at it that
way, you will have more success, and won't be the "people problem".
You have no idea what a good job you have here. Take advantage of it.
I've been where you are now, before I learned how not to do it. It
isn't fun, you don't look good, and there is a way out.

Good luck
Chris

Thanks,

-Michael

 

[Chris]

Michael Noone wrote:

I have suggested that everything be fully insulated - but it remains
to
be seen whether or not they take my advice. Certainly the circuits
I'm
building for this beast will all be fully protected like you have
advised.

I'm afraid I don't like what I'm hearing here. As a member of a
project team with assigned responsibility for the electronics, you are
in charge of making sure the electrons and holes stay where they they
belong. That's true whether you build or buy the supply. The supply
itself can be perfectly safe, but if it's improperly applied in a
project where you're assigned responsibility for that portion, you're
responsible, at least from an ethical standpoint. Also from an
employer's standpoint. If you were doing research in an industrial
ceramics lab, as the project EE, and someone was hurt because of an
omission of yours, you'd get fired. And rightly so. It shouldn't cost
very much to make it safe.

The other thing is, your reverse mousetrap shows you're taking
initiative, are willing to work hard to make the project happen, and as
an extra added bonus, may give you a little more time to learn what you
need to know. You don't think every EE who's assigned a project has
all the knowledge they need to complete it before they start? Ho ho,
nyuk nyuk. And they want you to start working productively on day 1,
whether you know everything you need or not. In the real world, it's
common to use the "swiss cheese method", and work your projects by
doing the parts you know how to do while scrambling to get up to speed
on the rest.

<snip>

I'll try to do some substantial research about HV. I'm an EE
undergrad
student and we haven't even mentioned HV design yet. Everything we've

covered so far is about small signals.

But, as I mentioned above, there's enough practical advice that can be
given about dealing with HV to take an experienced EE or engineering
tech a few hours to teach. Not substantial, but practical and
necessary. You can't get it in books or on the internets (at least
that I'm aware of), and it's really helpful to have someone look over
your shoulder and check your work, even if you could read what you need
to know. You've indicated that you're doing this at a
college/university with an EE program. Most have a crusty old tech
somewhere in the back (you'll frequently find him in the toolroom or
repair depot) who would be delighted to be freed of his imprisonment to
have a chance to teach some real world knowledge. As I've said before,
techs are usually much freer with advice if you offer them a six of
their favorite libation or some other similar inducement. Also check
out resources in the computer department, and any other likely
suspects.

<snip>

$200 USD plus shipping. Spec a good one with an emergency shutdown
circuit, and be sure to wire your reverse mousetrap safety
microswitch
into the circuit before you let 'em at it. Let them play, and be
creative and curious. Watch what they're doing. Help to keep them
moving. Ask good questions:

- Do you have an idea of voltage requirements yet?

I just e-mailed them again to see if they had decided on anything -
I'll
get back to you on that.

The point is that one practical experiment is sometimes worth more than
all the collective wisdom of the project group. Get a good power
supply and help them find out what they need.

- How about maximum current? (If this is an electrostatic
application, you probably won't need a tenth of the current you're
talking about. But you'll never know until you plug something in
and
check it out. It's your job to measure things.)

They want a max output of 20ma on each of the 10 outputs - so .2A
total
max output, though in reality it should be quite a good deal less
than
that.

Again, try it and you may, I say. Get a power supply, plug it in, and
find out.

- How much ripple voltage(many electrostatic applications go nuts
with
a superimposed AC ripple)?

I don't think there will be any ripple - but the load will be
changing,
if that will be a problem

To specify a DC power supply, you use many parameters. Some of them
are:
* Voltage Range (min/max)
* Current Range (max)
* Current limit/foldback/shutdown
* Voltage control or dual voltage/current control
* Line regulation
* Load regulation
* Ripple (Output AC voltage superimposed on the DC output, over line
and load range)

The last three are critical, because you're looking for a cheapie
solution, and should be hoping (I would) you can get away with an
unregulated supply (possibly controlled by Variac). You might even be
able to just cannibalize an old tube boat anchor and steal the filament
transformer, series filter choke or resistors and caps, and be done
with it for a miraculous low price. However, that isn't going to
happen if their application can't have more than 20mV p.p. of AC
ripple. That may also mean you require a linear power supply rather
than a HV switcher. Good to know.

<snip>

- Do they need operational control over the power supply? If so,
what
bandwidth? Or can you "make do" with a couple of relays if you
need a
bidirectional supply?

What exactly do you mean by operational control? As far as I know -
initially they'll need to set the voltage - and after that they'll
just
need to turn it on and off.

Imagine that you had a "super operational amplifier", that could be
programmed by feedback and/or a low voltage input, whose output went
from, say, 0 to 1000V with a current capability of 0 to 100 mA and a
power bandwidth in the audio range. That's an "operational" power
supply, a gift from the gods for research labs that are doing HV stuff.
Some are unidirectional (they can only source current), and some are
bidirectional (source or sink). Some will also be bipolar (with output
voltage range, say, from +400V to -400V, sourcing and sinking current
with positive or negative output). An operational power supply would
be ideal to rent, because you could simulate various amounts of ripple
and voltage sag to see if an unregulated or poorly regulated supply
would do the job, allowing you to home brew something quick & dirty and
saving money.

<snip>

Thanks,

-Michael

Well, Michael. This business of inserting replies in the middle of
copies of another post may work well in Outlook for email (where you
can use colors to distinguish replies) but it doesn't work well in
newsgroups. Also, it leads to time-consuming editing which might
distort meaning. Try to collect all your answers and post below the
copy of the prior response, please.

I'm serious about your finding somebody knowledgeable to stand in front
of you and give a few pointers about HV, and check your ideas and work.
Be safe first.

Luck is the residue of hard work. Good luck, Mike.
Chris

 

[Rich Grise]

On Sun, 20 Mar 2005 02:43:15 -0600, Michael Noone wrote:

Ken Moffett <KMoffet@mn.rr.com> wrote in news:Xns961DAB715231Cqwerty@

As I explained in another post - something to do with "magic molecules".
They want to apply high voltages at various points in channels
containing fluids of "magic molecules".

My best tool is, with my "helpful" questioning, to let them come up
with all the spec's. I frequently find that they are trying to do what
has already been done, by someone else. And, a little research will
give me the answers, or at least the right questions to put to my
customers.

I still can't believe they pay me to do this job.

Every week I prod them till they give me exact specs. Then the next week
- suddenly they decide that they were wrong and give me new specs - over
and over and over... This has been going on for a couple months now.

http://www.hammondmfg.com/700.htm
http://froogle.google.com/froogle?q=variac&num=50&hl=en&lr=lang_en&ie=UTF-8&safe=off&sa=N&tab=ff&oi=froogler

And some diodes - capacitor/choke are optional, as you characterize the
MMs for ripple amplitude.

Just put the transformer on the variac, rectify it, and optionally filter
it.

That's about the cheapest adjustable HV supply I can think of at the
moment.

Cheers!
Rich

 

[Michael Noone]

Rich Grise <richgrise@example.net> wrote in
newsan.2005.03.21.17.47.45.716067@example.net:

On Mon, 21 Mar 2005 12:47:03 +0000, Alan Turner wrote:

Hi,

On Mon, 21 Mar 2005 02:28:18 -0600, Michael Noone wrote:


Hmm - though it would be fairly practical - I expect my superiors
would shoot that idea down rather quickly. Though I must admit it'd
be damn fun to make a huge block of batteries... Though charging
would be a pain.

Well - if it's cheaper and it does the job - what have they got to
whinge about? Sometimes a pragmatic approach is the best

They are academics. They essentially have the opposite of a clue when
it comes to real-world stuff.

I still say Michael should take a stand: "Here is the power supply
that meets your specification - I'm sorry that it costs thirty
thousand dollars, but this is what it takes to meet your impossible
specifications. I _am_ the engineer here after all, and you obviously
have no idea what you need, which is why you hired me, no? This is the
one you need. Use it, pay me, and we can get on with our experiments."

Good Luck!
Rich

That's essentially my plan. I'm going to reccomend a Bertan supply and see
if they can stomach the price.

-Michael